Jake Powell Thesis (HSS18-HSS26)

Evaluation of Special Concentrically Braced Frames for

Improved Seismic Performance and Constructability

Jacob A. Powell

A thesis

submitted in partial fulfillment of the

requirements for the degree of

Master of Science in Engineering

University of Washington

2010

Program Authorized to Offer Degree:

Department of Civil and Environmental Engineering


Graduate School

This is to certify that I have examined this copy of a master’s thesis by

Jacob A. Powell

and have found that it is complete and satisfactory in all respects,

and that any and all revisions required by the final

examining committee have been made.

Committee Members:

Dr. Charles W. Roeder

Dr. Dawn E. Lehman

Dr. Greg Miller

Date:

In presenting this thesis in partial fulfillment of the requirements for a master’s degree at the

University of Washington, I agree that the Library shall make its copies freely available for

inspection. I further agree that extensive copying of this thesis is allowable for scholarly purposes,

consistent with “fair use” as prescribed in the U.S. Copyright Law. Any other reproduction for any

purposes or by any means shall not be allowed without my written permission.

Signature ___________________________________

Date ______________________________________


Abstract

Evaluation of Special Concentrically Braced Frames for Improved Seismic Performance and

Constructability

Jacob A. Powell

Chair of Supervisory Committee:

Dr. Charles W. Roeder

Department of Civil and Environmental Engineering

Special Concentrically Braced Frames (SCBFs) are a lateral force resisting system for steel

structures which have been widely implemented in recent years. Since the damage observed to

moment frames during the 1994 Northgate, CA and 1995 Kobe, Japan earthquakes, designers are

relying more on SCBFs to resist seismic loads for new construction of steel structures. Increased

initial stiffness and less material make SCBFs an economic and efficient choice but researchers and

design engineers admittedly do not fully understand the nonlinear demands to the entire system

when subjected to severe earthquake loading.

Significant past research has focused on the component behaviors for Braced Frames but little

research has been conducted to evaluate the behavior of the entire system, including brace, gusset

plates and framing elements. The research at UW beginning in 2004 and has worked to indentify

inadequacies with the current AISC design procedures for gusset plate design and develop an

alternative method to increase system ductility and improve seismic performance. The Balance

Design Approach for gusset plates is a variation of capacity based design that encourages yielding

beyond the brace while depressing undesirable failure modes to increase the ultimate drift capacity

of the system. Also, an elliptical clearance requirement at the brace end in lieu of the AISC 2t

linear clearance has resulted in more compact, economical gusset plates with increase rotational

capacity.

This thesis builds from the knowledge obtain from the previous work and the investigation of

additional design parameters including in-plane frame stiffness, requirement of net section

reinforcing, alternative brace sections and constructability with bolted connections. Nine full scale

single story, single bay specimens were tested within the UW Structures Lab and the results have

been analyzed to further develop the Balanced Design Approach as an alternative to current design

procedures for SCBF gusset plates to assure a more ductile response and improved seismic

performance.

Acknowledgement

This thesis would not have been possible without the guidance, support, and patience of my

advisors, Dr. Charles Roeder and Dr. Dawn Lehman. I am indebted to you both for granting me

this opportunity and pushing me to strive to limits beyond what I thought I was capable of.

During this time, you have help me grown as an engineer, as well as a person. I’d like to give

special thanks to my colleagues, KC, EL, JW, DB, AB, and IJ for your friendships, honesty, and

intuitiveness, making my experience at UW memorable on so many different levels.

Most of all, I would like to thank my family: My mother, Amy, for your unwavering love and

support, your belief in me and your example of strength in my life, and my brother, JP, for showing

me what it truly means to put your heart behind your work and to believe in yourself. This also

would not have happened if it wasn’t for the people throughout my life that have shaped who I am

today; PJ and AW.

This work is dedicated to the memory of Daniel M. Hamel. In everything I do, you are always in

my thoughts and in my heart.

i

Table of Contents

Table of Contents .................................................................................................................... i

List of Figures ........................................................................................................................ ix

List of Tables......................................................................................................................... xx

Chapter 1: Introduction ........................................................................................................... 1

1.1 Background ................................................................................................................................................ 1

1.2 SCBF Design Theory ............................................................................................................................... 2

1.3 Research Objectives .................................................................................................................................. 6

1.4 Overview of Report .................................................................................................................................. 7

Chapter 2: Literature Review .................................................................................................. 8

2.1 Introduction ............................................................................................................................................... 8

2.2 Brace Behavior........................................................................................................................................... 8

2.3 Gusset Plate Behavior ............................................................................................................................ 11

2.4 System Behavior ...................................................................................................................................... 15

2.5 Previous UW SCBF Research ............................................................................................................... 17

Chapter 3: Specimen Design ................................................................................................. 20

3.1 Introduction ............................................................................................................................................. 20

3.2 Balanced Design Procedure for Gusset Plates Connections ........................................................... 25

3.2.1 Brace Forces ........................................................................................................................................ 27

3.2.2 Brace to Gusset Plate Connection ................................................................................................... 28

3.2.3 Tension Limit States for Gusset Plates ........................................................................................... 29

3.2.4 Gusset Plate Geometry and Elliptical Clearance ........................................................................... 30

3.2.5 Gusset Plate Buckling ........................................................................................................................ 31

3.2.6 Net Section Reinforcement............................................................................................................... 33

3.2.7 Interface Weld Design ....................................................................................................................... 34

3.3 Design Parameters .................................................................................................................................. 35

ii

3.3.1 Bolted Shear Plate Beam-to-Column Connection Design ...........................................................35

3.3.2 Bolted Beam End Plate Connection Design ..................................................................................38

3.3.3 Wide-Flange Brace to Gusset Connection Design ........................................................................39

Chapter 4: Test Setup ............................................................................................................ 42

4.1 Introduction..............................................................................................................................................42

4.2 Test Configuration Overview ................................................................................................................42

4.3 Test Setup Components .........................................................................................................................47

4.3.1 Strong Floor and Strong Wall ...........................................................................................................47

4.3.2 Reaction Block and Actuator ............................................................................................................47

4.3.3 Load Transfer Beam ...........................................................................................................................49

4.3.4 Channel Assembly for Reactions ......................................................................................................49

4.3.4.1 Channel Assembly Modifications ............................................................................................51

4.3.5 Out-of-Plane Supports .......................................................................................................................52

4.3.6 Axial Load System ..............................................................................................................................55

4.3.7 Data Acquisition System and Storage ..............................................................................................56

4.4 Instrumentation .......................................................................................................................................56

4.4.1 Strain Gauges .......................................................................................................................................57

4.4.2 Potentiometers.....................................................................................................................................59

4.4.3 Visual Observations ............................................................................................................................62

4.5 Loading Protocol .....................................................................................................................................62

Chapter 5: Experimental Results........................................................................................... 65

5.1 Introduction..............................................................................................................................................65

5.2 Yield Mechanisms and Failure Modes .................................................................................................65

5.2.1 Plate and Frame Yielding ...................................................................................................................67

5.2.2 Brace Buckling and Brace Damage ..................................................................................................70

5.2.3 Local Buckling .....................................................................................................................................72

5.2.4 Weld Tearing and Fracture ................................................................................................................73

iii

5.2.5 Bolted Connection Slip ...................................................................................................................... 74

5.3 Specimen and Test Result Nomenclature ........................................................................................... 74

5.4 HSS-18: Thin Gusset Plate, Unwelded Frame Connection ............................................................. 76

5.4.1 Specimen Overview ........................................................................................................................... 76

5.4.2 Initial Drift Range: 0% to 1.25% ..................................................................................................... 79

5.4.3 Moderate Drift Range: 1.25% to 2.75% ......................................................................................... 80

5.4.4 Severe Drift Range: > 2.75% ............................................................................................................ 82

5.4.5 Specimen Summary ............................................................................................................................ 87

5.5 HSS-19: Bolted WT Brace Connection ............................................................................................... 88


5.5.2 Initial Drift Range: 0% to 1.25% ..................................................................................................... 91

5.5.3 Moderate Drift Range: 1.25% to 2.75% ......................................................................................... 94

5.5.4 Specimen Summary ............................................................................................................................ 95

5.6 HSS-21: 14 Bolt Beam End Plate Connection ................................................................................... 97


5.6.2 Initial Drift Range: 0% to 1.25% ................................................................................................... 100

5.6.3 Moderate Drift Range: 1.25% to 2.75% ....................................................................................... 106

5.6.4 Severe Drift Range: > 2.75% .......................................................................................................... 107

5.6.5 Specimen Summary .......................................................................................................................... 110

5.7 HSS-20: 18 Bolt Beam End Plate Connection ................................................................................. 113

5.7.1 Specimen Overview ......................................................................................................................... 113



5.7.4 Severe Drift Range: > 2.75% .......................................................................................................... 116

5.7.5 Specimen Summary .......................................................................................................................... 117

5.8 HSS-22: Tapered Gusset Plate, Unwelded Frame Connection ..................................................... 119


iv

5.8.2 Initial Drift Ranges: 0% to 1.25% ................................................................................................. 121

5.8.3 Moderate Drift Ranges: 1.25% to 2.75% ..................................................................................... 126

5.8.4 Severe Drift Ranges: > 2.75% ........................................................................................................ 128

5.8.5 Specimen Summary ......................................................................................................................... 134

5.9 WF-23: Wide-Flange Brace Section ................................................................................................... 136




5.9.4 Severe Drift Range: > 2.75% ......................................................................................................... 148

5.9.5 Specimen Summary ......................................................................................................................... 159

5.10 HSS-24: Welded Flange, Bolted Web Frame Connection ............................................................. 161

5.10.1 Specimen Overview .................................................................................................................... 161

5.10.2 Initial Drift Ranges: 0% to 1.25% ............................................................................................. 164

5.10.3 Moderate Drift Ranges: 1.25% to 2.75% ................................................................................. 167

5.10.4 Severe Drift Ranges: > 2.75% ................................................................................................... 172

5.10.5 Specimen Summary ..................................................................................................................... 179

5.11 HSS-25: Heavy Beam, No Net Section Reinforcing ....................................................................... 180




5.11.4 Severe Drift Ranges: > 2.75% ................................................................................................... 192


5.12 HSS-26: Heavy Beam, Near-Fault Drift History ............................................................................. 196





v

Chapter 6: Data Analysis ..................................................................................................... 204

6.1 Introduction ........................................................................................................................................... 204

6.2 System Response ................................................................................................................................... 204

6.2.1 Force vs. Drift Response ................................................................................................................. 207

6.2.2 System Stiffness ................................................................................................................................ 209

6.3 Brace and Gusset Plate Response ...................................................................................................... 214

6.3.1 Brace Response ................................................................................................................................. 215

6.3.2 Brace Buckling Capacity .................................................................................................................. 216

6.3.2.1 Nonlinear Brace Response ..................................................................................................... 218

6.3.3 OOP Response of Brace and Gusset Plates ................................................................................ 221

6.3.3.1 Brace Out-of-Plane Displacement ........................................................................................ 222

6.3.3.2 Gusset Plate Rotations ........................................................................................................... 224

6.3.3.3 Brace Buckled Shape and Curvature .................................................................................... 226

6.3.3.4 Weld Damage ........................................................................................................................... 230

6.3.4 Brace Behavior in Tension .............................................................................................................. 233

6.3.4.1 Brace Elongation ..................................................................................................................... 233

6.3.4.2 Gusset Plate Elongation ......................................................................................................... 235

6.4 Frame Response .................................................................................................................................... 237

6.4.1 Column Moments ............................................................................................................................. 237

6.4.2 Column Shears .................................................................................................................................. 240

6.4.3 Frame Hysteretic Response ............................................................................................................ 244

6.4.4 Shear Tab Connection Rotations ................................................................................................... 245

6.5 Distribution of System Resistance...................................................................................................... 247

6.5.1 Evaluation of Analysis Methods for Resistance .......................................................................... 247

6.5.2 Distribution of Resistance ............................................................................................................... 253

6.6 Energy Dissipation................................................................................................................................ 257

6.7 Performance Level Comparisons ....................................................................................................... 264

vi

Chapter 7: Comparison of Design Parameters .................................................................... 270

7.1 Introduction........................................................................................................................................... 270

7.1.1 Specimen HSS-01 Overview .......................................................................................................... 272



7.1.4 HSS-17 Specimen Overview .......................................................................................................... 278


7.2 Beam-to-Column Connections for Thin Rectangular Gusset Plates ........................................... 280

7.2.1 TRGP System Response Comparison .......................................................................................... 281

7.2.2 TRGP Brace and Gusset Plate Comparison ................................................................................ 284

7.2.2.1 TRGP Brace Force Comparison .......................................................................................... 284

7.2.2.2 TRGP Brace Displacement and Gusset Plate Rotation Comparisons........................... 288

7.2.2.3 TRGP Brace and Gusset Plate Elongation Comparisons ................................................ 291

7.2.3 TRGP Frame Response Comparison ........................................................................................... 293

7.2.4 TRGP Energy Dissipation .............................................................................................................. 296

7.2.5 TRGP Summary ............................................................................................................................... 298

7.3 Beam-to-Column Connections for Thin Tapered Gusset Plates ................................................. 299

7.3.1 TTGP System Response Comparison .......................................................................................... 300

7.3.2 TTGP Brace and Gusset Plate Comparison ................................................................................ 302

7.3.2.1 TTGP Brace Force Comparison .......................................................................................... 302

7.3.2.2 TTGP Brace Displacement and Gusset Plate Rotation Comparisons ........................... 304

7.3.2.3 TTGP Brace and Gusset Plate Elongation Comparisons ................................................ 308

7.3.3 TTGP Frame Response Comparison ........................................................................................... 310

7.3.4 TTGP Energy Dissipation .............................................................................................................. 312

7.3.5 TTGP Summary ............................................................................................................................... 313

7.4 Wide-Flange Verses HSS Tubular Brace Sections .......................................................................... 314

7.4.1 Wide-Flange vs. HSS System Response Comparison ................................................................ 314

vii

7.4.2 Wide-Flange vs. HSS Brace and Gusset Plate Response Comparison .................................... 316

7.4.2.1 Wide-Flange vs. HSS Brace Force Comparison ................................................................. 316

7.4.2.2 Wide-Flange vs. HSS Brace Displacement and Gusset Plate Rotation Comparison ... 317

7.4.2.3 Wide-Flange vs. HSS Brace and Gusset Plate Elongation Comparison ........................ 321

7.4.3 Wide-Flange vs. HSS Energy Dissipation Comparison ............................................................. 323

7.4.4 Wide-Flange Brace Summary ......................................................................................................... 324

7.5 Bolted Connections for SCBFs .......................................................................................................... 325

7.5.1 HSS-19: Bolted WT Brace to Gusset Plate Connection ............................................................ 325

7.5.2 HSS-20 and HSS-21: Bolted Beam End-plate Connection ....................................................... 326

7.5.3 HSS-24: Welded Flanges, Bolted Web Beam-to-Column Connection .................................... 328

7.6 Net Section Reinforcing Requirement ............................................................................................... 331

Chapter 8: Conclusions ....................................................................................................... 333

8.1 Introduction ........................................................................................................................................... 333

8.2 Research Summary ................................................................................................................................ 333

8.3 Conclusions ............................................................................................................................................ 333

8.3.1 Balance Design Procedure and Elliptical Clearance ................................................................... 336

8.3.2 Beam-to-Column Connection ........................................................................................................ 336

8.3.3 Bolted SCBF Connections .............................................................................................................. 336

8.3.4 Net Section Reinforcement Requirement..................................................................................... 337

8.3.5 Wide-flange Brace Sections ............................................................................................................. 337

8.4 Future Recommendation ..................................................................................................................... 337

Appendix A: Specimen Design Calculations ...................................................................... 339

A.1 General.................................................................................................................................................... 339

A.2 Specimen HSS-18 Design Calculation ............................................................................................... 339

A.2.1 Member Selection ........................................................................................................................ 339

A.2.2 Brace Forces ................................................................................................................................. 341

A.2.3 Brace to Gusset Plate Design ..................................................................................................... 342

viii

A.2.4 Gusset Plate Design .................................................................................................................... 343

A.2.5 Net Section Reinforcement ....................................................................................................... 347

A.2.6 Interface Weld Design Calculation ........................................................................................... 348

A.3 Specimen WF-23 Design ..................................................................................................................... 348

A.3.1 W6x25 Brace Section Check ...................................................................................................... 348

A.3.2 Wide-Flange Brace to Gusset Plate Design ............................................................................ 349

Appendix B: Design Drawings ............................................................................................ 352

B.1 Specimen Drawings .............................................................................................................................. 352

Appendix C: Analysis Plots ................................................................................................. 361

C.1 Brace and Gusset Plate Behavior ....................................................................................................... 361

C.2 Brace and Gusset Plate Behavior ....................................................................................................... 361

C.2.1 Out-of-Plane Displacement at Brace Center .......................................................................... 361

C.2.2 Gusset Plate Rotation ................................................................................................................. 363

C.2.3 Brace Elongation ......................................................................................................................... 367

C.3 Frame Response .................................................................................................................................... 368

C.3.1 Column Moments........................................................................................................................ 369

C.3.2 Column Shears ............................................................................................................................. 373

C.3.3 Beam Shear Tab Rotations ........................................................................................................ 377

ix

List of Figures

Figure 1.1.1: Response of Single Brace vs. Opposing Braces .................................................................... 2

Figure 1.2.1: Desired Hierarchy of Inelastic Behavior for SCBFs ............................................................ 5

Figure 1.2.2: Yield Mechanisms and Failure Modes for SCBFs ................................................................ 6

Figure 2.2.1: Shaback and Brown Test Specimen ....................................................................................... 9

Figure 2.3.1: Whitmore Effective Width ..................................................................................................... 11

Figure 2.3.2: Bjorhovde and Chakrabarti Test Specimen (1985) ........................................................... 12

Figure 2.3.3: Thornton Effective Length for Buckling Capacity ........................................................... 13

Figure 2.3.4: Cheng and Yam Test Setup (1993) ....................................................................................... 14

Figure 2.3.5: Astaneh-Asl et al. Test Specimen ......................................................................................... 15

Figure 2.4.1: Uriz and Mahin Test Setup (2004) ....................................................................................... 16

Figure 2.5.1: Roeder et al Test Specimen (2008) ...................................................................................... 18

Figure 2.5.2: Gusset Plate Elliptical Clearance .......................................................................................... 19

Figure 3.1.1: Prototype Specimen (Johnson 2005) .................................................................................... 20

Figure 3.2.1: Whitmore Effective Width ..................................................................................................... 29

Figure 3.2.2: Block Shear Area ...................................................................................................................... 30

Figure 3.2.3: Gusset Plate Geometry for Elliptical Clearance Requirement (Kotulka 2007) ............. 31

Figure 3.2.4: Thornton Method for Gusset Plate Buckling ..................................................................... 32

Figure 3.2.5: Reduced Area at Brace to Gusset Connection .................................................................... 33

Figure 3.3.1: Bolted Shear Plate Connection .............................................................................................. 35

Figure 3.3.2: Bolted Web CJP Welded Flange Connection Detail .......................................................... 36

Figure 3.3.3: Uniform Force Method .......................................................................................................... 37

Figure 3.3.4: UFM Resolved Forces ............................................................................................................ 37

Figure 3.3.5: Bolted Beam End Plate Connection ..................................................................................... 39

Figure 3.3.6: Wide-Flange Brace to Gusset Detail .................................................................................... 40

Figure 3.3.7: Wide-Flange Brace to Gusset Section-Cut .......................................................................... 41

Figure 4.2.1: Test Setup and Components .................................................................................................. 43

Figure 4.2.2: Test Setup Layout (Johnson 2005) ...................................................................................... 44

Figure 4.2.3: Actual Test Setup ..................................................................................................................... 45

Figure 4.3.1: Actuator and Reaction Block Photo .................................................................................... 47

Figure 4.3.2: Actuator and Reaction Block Assembly ............................................................................. 48

Figure 4.3.3: Load Beam ............................................................................................................................... 49

x

Figure 4.3.4: Channel Assembly Section Detail ........................................................................................ 50

Figure 4.3.5: Channel Assembly Plan Detail ............................................................................................. 50

Figure 4.3.6: Channel Assembly Modifications......................................................................................... 52

Figure 4.3.7: North Beam and West Column Out-of-Plane Supports .................................................. 53

Figure 4.3.8: Load Beam Out-of-Plane Support ....................................................................................... 54

Figure 4.3.9: Modified North Beam OOP Support ................................................................................. 54

Figure 4.3.10: Axial Load System Detail .................................................................................................... 55

Figure 4.3.11: Tension Rod Tie Downs ..................................................................................................... 56

Figure 4.4.1: Typical Strain Gauge Layout ................................................................................................. 58

Figure 4.4.2: Potentiometer Layout ............................................................................................................ 59

Figure 4.4.4: BEI Duncan Potentiometers at SW Gusset Plate ............................................................. 61

Figure 4.4.3: String Potentiometers at NE Gusset Plate ......................................................................... 61

Figure 4.5.1: Original Loading Protocol (Johnson 2005) ........................................................................ 63

Figure 4.5.2: Alternate Near-Fault Loading Protocol .............................................................................. 64

Figure 5.2.1: Yield Mechanisms and Failures Modes for Test Specimen .............................................. 66

Figure 5.2.2: Initial/Mild Yielding of the Gusset Plate (Y1) .................................................................... 68

Figure 5.2.3: Moderate Yielding of the Gusset Plate (Y3) ....................................................................... 69

Figure 5.2.4: Significant/Severe Yielding of the Gusset Plate (Y5) ........................................................ 69

Figure 5.2.5: Progression of Brace Buckling .............................................................................................. 71

Figure 5.2.6: Brace Damage at Plastic Hinge.............................................................................................. 72

Figure 5.2.7: Local Buckling of HSS-25 Columns ..................................................................................... 73

Figure 5.2.8: Examples of Weld Damage from WF-23 ............................................................................ 74

Figure 5.3.1: Specimen Component Notation (Johnson 2005) ............................................................... 75

Figure 5.3.2: Wide-Flange Area Designations related to Test Set-Up (Johnson 2005) ....................... 75

Figure 5.4.1: HSS-18 Connection Detail ..................................................................................................... 77

Figure 5.4.2: HSS-18 Drift History .............................................................................................................. 77

Figure 5.4.3: HSS-18 Hysteresis.................................................................................................................... 78

Figure 5.4.4: Y1 Initial Yielding of SW Gusset (0.11%) ........................................................................... 79

Figure 5.4.5: Initial Yielding (Y1) of Gusset Plate (0.21%) ...................................................................... 79

Figure 5.4.6: B1 Brace Buckling (-0.12%) ................................................................................................... 80

Figure 5.4.7: B2 Brace Buckling (-0.62%) ................................................................................................... 80

Figure 5.4.8: Moderate Yielding (Y3) of the NE Gusset plate (-0.94/0.41%) ...................................... 81

Figure 5.4.9: Initial Yielding of Brace Net Section (0.74%) ..................................................................... 82

Figure 5.4.10: Slight Edge Buckling (B1) of SW Gusset Plate (1.09%).................................................. 82

Figure 5.4.11: Initial Yielding (Y1) of NE Beam Web (1.10%) ............................................................... 83

xi

Figure 5.4.12: Severe Yielding (Y5) of Gusset Plate (1.42%) .................................................................. 83

Figure 5.4.14: Moderate Edge Buckling (B2) of SW Gusset Plate (1.60%)........................................... 84

Figure 5.4.13: Compressive Brace Failure (BC) (-2.09%) ......................................................................... 84

Figure 5.4.16: Moderate Yielding (Y3) of Northeast Beam (-2.54%) ..................................................... 85

Figure 5.4.15: Damage to SW Column (-2.54/1.60%).............................................................................. 85

Figure 5.4.17: Local Compressive Failure of Brace (-2.59%) .................................................................. 86

Figure 5.4.18: Brace Fracture at Plastic Hinge (1.31%) ............................................................................ 86

Figure 5.4.19: Gusset Plate Yielding at Completion of Test .................................................................... 87

Figure 5.4.20: SW Gusset Plate Edge Buckling at Completion of Test ................................................. 88

Figure 5.5.1: HSS-19 Detail ........................................................................................................................... 89


Figure 5.5.3: HSS-19 Hysteresis.................................................................................................................... 90

Figure 5.5.4: Moderate Yielding (Y3) of NE Splice Plate (-0.18%) ........................................................ 91

Figure 5.5.5: Initial Yielding (Y1) of the SW Splice Plate (-0.18%) ........................................................ 92

Figure 5.5.6: Severe Yielding (Y5) (-0.33%) ............................................................................................... 92

Figure 5.5.7: Southwest WT slip (BSLP) vs. Cycle .................................................................................... 93

Figure 5.5.8: Moderate Plate Buckling (B2) (-0.65%)................................................................................ 93

Figure 5.5.9: Binding of WT Stem on Brace .............................................................................................. 94

Figure 5.5.10: Buckled Shape of Brace (-0.72%)........................................................................................ 94

Figure 5.5.11: Northeast WT slip (BSLP) vs. Cycle .................................................................................. 95

Figure 5.5.12: Damage at Splice Plate Hinge before and after Fracture (PF) (0.28%) ........................ 95

Figure 5.5.13: Damage from WT and Tube Binding ................................................................................ 96

Figure 5.5.14: SW Gusset Plate at End of Test .......................................................................................... 97

Figure 5.5.15: Fractured Splice Plate at End of Test ................................................................................. 97

Figure 5.6.1: HSS-21 Connection Detail ..................................................................................................... 98


Figure 5.6.3: HSS-21 Hysteresis.................................................................................................................. 100

Figure 5.6.4: Initial Yielding (Y1) of NE Gusset Plate (0.07%) ............................................................ 101

Figure 5.6.5: Initial Yielding (Y1) of SW Beam/Col Connection (0.10%) .......................................... 101

Figure 5.6.6: Yield Lines Parallel to South Beam Interface Weld ......................................................... 102

Figure 5.6.7: Initial Yielding (Y1) of South Beam Web (0.14%) ........................................................... 102

Figure 5.6.8: Initial Yielding (Y1) of North Beam Web (0.17%) .......................................................... 103

Figure 5.6.9: Initial Yielding (Y1) at SW Gusset to Column Reentrant Corner (-0.34%) ................. 103

Figure 5.6.10: Y3 of the NE Gusset Plate (0.35%) ................................................................................. 104

Figure 5.6.11: Bolt Fracture of SW Beam End Plate Connection (0.36%) ......................................... 104

xii

Figure 5.6.12: Y3 of SW Gusset Plate (0.36%) ........................................................................................ 105

Figure 5.6.13: B2 Brace Buckling (-0.79%) ............................................................................................... 105

Figure 5.6.14: Y1 of NE Column Flange (0.56%) ................................................................................... 106

Figure 5.6.16: BC Damage at Brace Plastic Hinge (-2.13%) .................................................................. 107

Figure 5.6.15: Increased Yielding of Brace and NE Gusset Plate (0.86%) ......................................... 107

Figure 5.6.18: Y3 at NE Beam Web (1.60%) ........................................................................................... 108

Figure 5.6.17: Column Damage (-2.13/1.18%) ........................................................................................ 108

Figure 5.6.20: Tearing of Brace at Plastic Hinge (1.60%) ....................................................................... 109

Figure 5.6.19: Yielding at Bolt Holes of NE Column Flange (1.60%) ................................................. 109

Figure 5.6.21: B2 Edge Buckling of NE Gusset Plate (1.60%) ............................................................. 110

Figure 5.6.22: Brace Plastic Hinge Damage and Failure ......................................................................... 110

Figure 5.6.23: Gusset Plate Damage at End of Test ............................................................................... 111

Figure 5.6.25: Southwest End Plate Rotation at Column ....................................................................... 112

Figure 5.6.24: Southwest End Plate Prying............................................................................................... 112

Figure 5.7.1: HSS-20 Connection Detail ................................................................................................... 113

Figure 5.7.2: HSS-20 Drift History ............................................................................................................ 114


Figure 5.7.4: Y1 at NE Column Inside Flange (1.39%) .......................................................................... 117

Figure 5.7.5: Bolted End Plate Connection with Spacing Washers ...................................................... 118

Figure 5.8.1: HSS-22 Connection Detail ................................................................................................... 119

Figure 5.8.2: HSS-22 Drift History ............................................................................................................ 120


Figure 5.8.4: Y1 of NE Gusset Plate (0.11%) .......................................................................................... 122

Figure 5.8.5: Y1 of SW Gusset Plate (0.13%) .......................................................................................... 122

Figure 5.8.6: Increased Yielding of NE Gusset Plate (0.17%) .............................................................. 123

Figure 5.8.7: Y1 at NE Gusset to Column Reentrant Corner (0.35%) ................................................ 123

Figure 5.8.8: B1 Buckled Shape of Brace (-0.25%).................................................................................. 124

Figure 5.8.9: Yielding along Column Interface Welds (0.42%) ............................................................. 124

Figure 5.8.11: B2 Buckled Shape of Brace (-0.50%) ............................................................................... 125

Figure 5.8.10: Y3 at NE Gusset Plate (0.52%) ......................................................................................... 125

Figure 5.8.12: Y3 at SW Gusset Plate (0.79%) ......................................................................................... 126

Figure 5.8.13: Buckled Shape of Brace (-1.11%)...................................................................................... 127

Figure 5.8.14: WD at SW Column Reentrant Corner (-1.11%) ............................................................. 127

Figure 5.8.15: WD SW Gusset to Column Interface Weld Cracking (-1.46%) .................................. 128

Figure 5.8.16: Damage at SW Gusset Plate Welds (-1.86%) .................................................................. 128

xiii

Figure 5.8.17: Damage at Southwest Gusset Plate Welds (-1.86%) ...................................................... 129

Figure 5.8.18: Damage at NE Gusset Welds (-1.86%) ........................................................................... 129

Figure 5.8.20: Necking of Brace at Hinge (1.49%) .................................................................................. 130

Figure 5.8.19: Tearing of Brace at Plastic Hinge (1.49%) ....................................................................... 130

Figure 5.8.21: Y5 of Gusset Plate (1.48%) ................................................................................................ 131

Figure 5.8.22: Local Deformation (BC) of Brace (-2.48%) .................................................................... 131

Figure 5.8.23: SW Gusset to Column Interface Weld Crack (-2.48%) ................................................. 132

Figure 5.8.24: Damage at NE Gusset Welds (-2.48%) ........................................................................... 132

Figure 5.8.25: Y1 of SW Column (-2.48%) ............................................................................................... 133

Figure 5.8.26: BF of the Brace Center (1.38%) ........................................................................................ 134

Figure 5.8.27: Residual Deformation of Gusset Plate after Brace Fracture (1.38%) ......................... 134

Figure 5.8.28: NE Beam-to-Column Connection Rotation ................................................................... 135

Figure 5.8.29: NE Beam-to-Column Rotation (1.49%) .......................................................................... 135

Figure 5.8.30: NE Beam Torsion at Connection ..................................................................................... 136

Figure 5.9.1: WF-23 Connection Detail .................................................................................................... 137

Figure 5.9.2: WF-23 Displacement History .............................................................................................. 138

Figure 5.9.3: WF-23 Hysteresis ................................................................................................................... 138

Figure 5.9.4: Y1 at NE Gusset Plate (0.07%) ........................................................................................... 139

Figure 5.9.5: Y1 at Brace Flange Reduced Section in Tension (0.10%) ............................................... 140

Figure 5.9.6: Y1 of Brace Flange in Compression (-0.08%) .................................................................. 140

Figure 5.9.7: Yielding at NE Brace on Flange (0.12%) ........................................................................... 141

Figure 5.9.8: View of Yielding at NE of Brace (0.12%) ......................................................................... 141

Figure 5.9.10: Yielding at NE Gusset Plate (0.20%) ............................................................................... 142

Figure 5.9.9: Y1 on the Lower Edge of the Brace Flange (-0.18%) ..................................................... 142

Figure 5.9.12: Gusset Plate Yielding at Northeast Brace End (-0.27%) .............................................. 143

Figure 5.9.11: Brace Condition (-0.27%) ................................................................................................... 143

Figure 5.9.13: Y3 at Brace Center (-0.38%) .............................................................................................. 144

Figure 5.9.14: B2 Buckled Brace Shape (-0.49%) .................................................................................... 144

Figure 5.9.15: Y5 at Brace Center (-0.83%) .............................................................................................. 145

Figure 5.9.16: Y1 at NE Column and Beam Flanges (0.74%) ............................................................... 146

Figure 5.9.17: Y1 at NE Gusset Reentrant Corners (-1.16%) ............................................................... 146

Figure 5.9.18: Y3 at NE Gusset Plate (0.98%) ......................................................................................... 147

Figure 5.9.19: Y1 at SW Column (-1.48%) ............................................................................................... 147

Figure 5.9.20: Yielding at Brace Center (-1.48%) .................................................................................... 148

Figure 5.9.21: Y3 at SW Gusset Plate (1.23%) ......................................................................................... 148

xiv

Figure 5.9.22: Edge Buckling at SW Gusset Plate (1.23%) .................................................................... 149

Figure 5.9.23: SW Column Yielding (-1.80%) .......................................................................................... 149

Figure 5.9.25: Y1 at SW Column/Beam Connection (1.48%) .............................................................. 150

Figure 5.9.24: NE Frame Yielding (1.48%) .............................................................................................. 150

Figure 5.9.26: NE Column Web Yielding (1.48%) .................................................................................. 151

Figure 5.9.27: SW Beam Flange Yielding (1.48%) ................................................................................... 151

Figure 5.9.28: Y3 at SW Column (-2.15%) ............................................................................................... 152

Figure 5.9.29: B2 Edge Buckling at SW Gusset Plate (1.77%) .............................................................. 152

Figure 5.9.31: NE Column Deformation (1.77%) ................................................................................... 153

Figure 5.9.30: Column Web Yielding (1.77%) ......................................................................................... 153

Figure 5.9.32: Y3 at NE Column (-2.49%) ............................................................................................... 154

Figure 5.9.33: Y5 of Gusset Plate (2.05%) ................................................................................................ 154

Figure 5.9.34: SW Column Deformation (2.05%) ................................................................................... 155

Figure 5.9.35: Gusset Edge Plate Buckling (2.05%) ................................................................................ 155

Figure 5.9.37: Residual Deformation at Brace Center (2.35%) ............................................................. 156

Figure 5.9.36: Cracking (WD) at NE Gusset Plate Interface Welds (-2.86%) .................................... 156

Figure 5.9.39: SW Beam/Column Moment Connection Yielding (2.35%) ......................................... 157

Figure 5.9.38: SW Column Damage (2.35%)............................................................................................ 157

Figure 5.9.40: Y3 at NE Column Web (2.35%) ....................................................................................... 158

Figure 5.9.41: Severe Damage at NE Gusset to Beam Weld ................................................................. 158

Figure 5.9.42: System Failure Modes (2.32%) .......................................................................................... 159

Figure 5.9.43: Brace Center Damage (-3.21%) ......................................................................................... 160

Figure 5.9.44: Out-of-Plane Displacement at Brace Center .................................................................. 160

Figure 5.9.45: Bolt-Hole Elongation (Post Test) ..................................................................................... 161

Figure 5.9.46: Damage at SE Shear Tab Connection (Post Test) ......................................................... 161

Figure 5.10.1: HSS-24 Connection Detail ................................................................................................. 162

Figure 5.10.2: HSS-24 Displacement History ........................................................................................... 163

Figure 5.10.3: HSS-24 Hysteresis ............................................................................................................... 163

Figure 5.10.4: Y1 at NE Column at Gusset Reentrant Corner (0.27%) .............................................. 165

Figure 5.10.6: NE Column Flange Yielding at Beam/Column Connection (0.37%) ........................ 166

Figure 5.10.5: B1 Level Brace Out-of-Plane Displacement (-0.39%)................................................... 166

Figure 5.10.7: B2 Level Buckling of Brace (-0.63%) ............................................................................... 167

Figure 5.10.8: NE Column Flange Yielding (0.46%) .............................................................................. 168

Figure 5.10.9: Yielding at East Column Base (0.46%) ............................................................................ 168

Figure 5.10.10: Y3 at NE Gusset Plate (-0.81%) ..................................................................................... 169

xv

Figure 5.10.11: Y3 at NE Column Flange (0.68%) ................................................................................. 169

Figure 5.10.12: Buckled Shape of Brace (-1.17%) ................................................................................... 170

Figure 5.10.13: Increased Gusset Plate Yielding and Y1 at SW Column Flange (-1.17%) ............... 170

Figure 5.10.14: B1 at NE Column Flange (0.98%) .................................................................................. 171

Figure 5.10.15: Y1 of Beam Flanges at Reentrant Corners (0.87%) ..................................................... 171

Figure 5.10.17: Yielding at NE Beam/Column Connection (1.34%) .................................................. 172

Figure 5.10.16: Y3 at NE Column (1.34%) .............................................................................................. 172

Figure 5.10.18: Unbalanced Yielding at NE Gusset Plate (1.34%) ....................................................... 173

Figure 5.10.19: NE Column Web Yielding (-1.83%) .............................................................................. 173

Figure 5.10.20: Y3 Web Yielding and Flange Deformation at SW Column (1.71%) ........................ 174

Figure 5.10.21: B2 Gusset Plate Edge Deformation (1.71%) ................................................................ 174

Figure 5.10.23: Yielding of SW Column Flange (-2.14%) ...................................................................... 175

Figure 5.10.22: Brace Shape and BC Deformation (-2.14%) ................................................................. 175

Figure 5.10.24: Y5 at NE Gusset Plate (1.94%)....................................................................................... 176

Figure 5.10.26: Y3 at SW Column (1.94%) ............................................................................................... 177

Figure 5.10.25: Increased Yielding at NE Column (1.94%) ................................................................... 177

Figure 5.10.27: Brace Shape and Gusset Plate Rotation (-2.50%) ........................................................ 178

Figure 5.10.28: Column Yielding (-2.50%) ............................................................................................... 178

Figure 5.10.29: NE Gusset Plate to Beam Weld Crack (-2.50%) .......................................................... 179

Figure 5.10.30: NW Column Web Damage (Post Test) ......................................................................... 180

Figure 5.10.31: Gusset Plate Damage (Post Test) ................................................................................... 180




Figure 5.11.4: Brace Shape (-0.30%) .......................................................................................................... 184

Figure 5.11.5: B1 Brace Buckled Shape (-0.42%) .................................................................................... 185

Figure 5.11.6: Gusset Plate Yielding at Brace Ends (-0.53%) ................................................................ 185

Figure 5.11.7: Yielding of NE Column Flange (0.31%) ......................................................................... 186

Figure 5.11.8: Yielding at Reentrant Corner of NE Column Flange (0.31%) ..................................... 186

Figure 5.11.9: Increased Yielding of Gusset Plate (0.69%) .................................................................... 187

Figure 5.11.10: Increased Yielding at Column Flanges (0.42%) ............................................................ 187

Figure 5.11.11: Increased Yielding at SW Gusset Plate (-1.05%) .......................................................... 188

Figure 5.11.12: Y3 at NE Column Flange (0.54%) ................................................................................. 188

Figure 5.11.13: Crack Propagation at SW Brace Net Section (0.54%) ................................................. 189

Figure 5.11.14: Y3 at SW Gusset Plate (-1.50%) ..................................................................................... 190

xvi

Figure 5.11.15: Buckled Brace Shape (-1.50%) ........................................................................................ 190

Figure 5.11.16: Y3 at NE Gusset Plate (0.68%)....................................................................................... 191

Figure 5.11.17: Crack Propagation at Brace SW Net Section (0.68%) ................................................. 191

Figure 5.11.18: NE Column Damage (-1.94%) ........................................................................................ 192

Figure 5.11.19: NE Column Damage (0.88%) ......................................................................................... 192

Figure 5.11.20: Crack Propagation at Brace SW Net Section (0.88%) ................................................. 193

Figure 5.11.21: Brace Condition (-2.41%) ................................................................................................ 193

Figure 5.11.22: Y3 at NE Gusset Plate (-2.41%) ..................................................................................... 194

Figure 5.11.23: Yielding at Column Outside Flanges (-2.41%) ............................................................. 194

Figure 5.11.24: NE Column Damage (-2.41%) ........................................................................................ 195

Figure 5.11.25: Shear Tab Bolt Fracture (Post Test) ............................................................................... 195


Figure 5.12.2: Designed Near-Fault Drift History................................................................................... 197



Figure 5.12.5: Buckled Brace Shape (-0.40%) .......................................................................................... 199

Figure 5.12.6: Yielding at Brace Net Section Locations (0.74%) .......................................................... 200

Figure 5.12.7: Y3 at NE Column (0.74%) ................................................................................................ 200

Figure 5.12.8: NSF at SW Brace End (0.99%) ......................................................................................... 201

Figure 5.12.9: SW Column Damage (0.99%)............................................................................................ 202

Figure 5.12.10: NE Column Damage (0.99%) ......................................................................................... 202

Figure 6.2.1: Backbone Curve Description ............................................................................................... 210

Figure 6.3.1: Idealized Inelastic Brace Behavior Under Cyclic Loading (Kotulka 2007) .................. 215

Figure 6.3.2: Brace Strain Gauge Locations .............................................................................................. 216

Figure 6.3.3: Elastic-Plastic Stress/Strain Behavior ................................................................................ 218

Figure 6.3.4: Plasticity Model under Repeat Cyclic Loading .................................................................. 219

Figure 6.3.5: Shape and Curvature of Buckled Brace with Different End Conditions ..................... 222

Figure 6.3.6: Horizontal Drift Range to Brace OOP Displacement Relationship ............................. 223

Figure 6.3.7: Brace Out-of-Plane vs. Total Drift Range Comparison .................................................. 223

Figure 6.3.8: Gusset Plate Rotation ............................................................................................................ 224

Figure 6.3.9: Example of NE Gusset Plate Rotations ............................................................................ 225

Figure 6.3.10: Instrumentation for Determining Buckled Brace Shape ............................................... 226

Figure 6.3.11: Brace Shape Comparison at Initial Total Drift Range (Approx. 1.25%) .................... 227

Figure 6.3.12: Brace Shape Comparison at Moderate Total Drift Ranges (Approx. 2.75%) ........... 228

Figure 6.3.13: Buckled Brace Shape Immediately Prior to Failure ....................................................... 229

xvii

Figure 6.3.14: Additional Demands on Gusset Plate Interface Welds ................................................. 231

Figure 6.3.15: Brace Elongation vs. Total Drift Range ........................................................................... 234

Figure 6.3.16: Brace Axial Stress vs. Brace Strain .................................................................................... 235

Figure 6.3.17: Gusset Plate Elongation Comparison .............................................................................. 236

Figure 6.3.18: Brace Elongation verses Gusset Plate Elongation ......................................................... 237

Figure 6.4.1: Column Moment Calculation .............................................................................................. 238

Figure 6.4.2: Column Moment at Edge of NE Gusset Plate ................................................................. 239

Figure 6.4.3: Column Moment at Edge of SW Gusset Plate ................................................................. 240

Figure 6.4.4: Column Shear Calculations .................................................................................................. 241

Figure 6.4.5: Shear Tab Rotation Calculation ........................................................................................... 246

Figure 6.4.6: Northwest Shear Tab Rotation Comparison..................................................................... 246

Figure 6.4.7: Southeast Shear Tab Rotation Comparison ...................................................................... 247

Figure 6.6.1: Energy Dissipation (Kotulka 2007) .................................................................................... 258

Figure 6.6.2: Total Energy Dissipation Comparison ............................................................................... 259

Figure 6.6.3: Brace Energy Dissipation Comparison .............................................................................. 260

Figure 6.6.4: Frame Energy Dissipation Comparison ............................................................................. 261

Figure 6.6.5: WF-23 Energy Dissipation by Component ....................................................................... 264

Figure 7.1.1: Interface Weld Cracking at SW Gusset Plate (1.78% Total Drift Range) .................... 274

Figure 7.1.2: Weld Fracture at SW Gusset Plate (2.65% Total Drift Range) ...................................... 274

Figure 7.1.3: HSS-05 NE Gusset Yielding (4.96% Total Drift Range) ................................................ 277

Figure 7.1.4: Cracking of NE Gusset to Column Weld (4.96% Total Drift Range) .......................... 277

Figure 7.1.5: Severe Gusset Plate Yielding and Severe Weld Damage (-2.79%) ................................ 279

Figure 7.2.1: TRGP Load vs. Drift (Positive Envelop) .......................................................................... 283

Figure 7.2.2: TRGP Load vs. Drift (Negative Envelop) ........................................................................ 284

Figure 7.2.3: TRGP Brace Compression Degradation Comparison .................................................... 287

Figure 7.2.4: TRGP Brace Out-of-Plane Displacement Comparison .................................................. 288

Figure 7.2.5: TRGP NE Gusset Plate Rotation Comparison ................................................................ 289

Figure 7.2.6: TRGP Residual Brace OOP Displacement Comparison ................................................ 290

Figure 7.2.7: TRGP Residual NE Gusset Plate Rotation Comparison ................................................ 290

Figure 7.2.8: TRGP Brace Elongation Comparison................................................................................ 291

Figure 7.2.9: TRGP Gusset Plate Elongation Comparison ................................................................... 292

Figure 7.2.10: TRGP Brace vs. Gusset Plate Elongation ....................................................................... 292

Figure 7.2.11: TRGP Frame Resistance vs. Drift Comparison (+ Direction) .................................... 293

Figure 7.2.12: TRGP Frame Resistance vs. Drift Comparison (- Direction) ...................................... 294

Figure 7.3.1: TTGP Load vs. Drift (Positive Envelope) ........................................................................ 301

xviii

Figure 7.3.2: TTGP Load vs. Drift (Negative Envelope)....................................................................... 301

Figure 7.3.3: TTGP Compressive Degradation Comparison ................................................................ 304

Figure 7.3.4: TTGP Brace Out-of-Plane Displacement Comparison .................................................. 305

Figure 7.3.5: TTGP NE Gusset Rotation Comparison .......................................................................... 306

Figure 7.3.6: TTGP Residual Brace OOP Displacement Comparison ................................................ 307

Figure 7.3.7: TTGP Residual NE Gusset Plate Rotation Comparison ................................................ 307

Figure 7.3.8: TTGP Brace Elongation vs. Total Drift Range ................................................................ 308

Figure 7.3.9: TTGP Gusset Plate Elongation Comparison ................................................................... 309

Figure 7.3.10: TTGP Brace Strain vs. Gusset Plate Strain Comparison .............................................. 309

Figure 7.3.11: TTGP Frame Resistance vs. Drift Comparison (+ Direction) .................................... 310

Figure 7.3.12: TTGP Frame Resistance vs. Drift Comparison (- Direction) ...................................... 311

Figure 7.4.1: WF Backbone Curve Comparison ...................................................................................... 315

Figure 7.4.2: WF Brace Yielding at Plastic Hinge (-3.21%) ................................................................... 318

Figure 7.4.3: WF Brace OOP Displacement Comparison ..................................................................... 319

Figure 7.4.4: NE Gusset Plate Rotation Comparison for WF............................................................... 319

Figure 7.4.5: WF Residual Brace OOP Displacement Comparison ..................................................... 320

Figure 7.4.6: WF Residual NE Gusset Plate Rotation Comparison ..................................................... 321

Figure 7.4.7: Brace Elongation Comparison for WF .............................................................................. 321

Figure 7.4.8: Gusset Plate Elongation for WF ......................................................................................... 322

Figure 7.4.9: WF Brace Strain vs. Gusset Plate Strain Comparison ..................................................... 323

Figure 7.4.10: WF-23 Energy Dissipation ................................................................................................. 324

Figure 7.4.11: WF Brace to Gusset Plate Connection ............................................................................ 325

Figure 7.5.1: HSS-21 NE Gusset Plate at End of Test ........................................................................... 327

Figure 7.5.2: HSS-24 SW Gusset Plate Yielding at End of Test ........................................................... 329

Figure 7.5.3: HSS-24 Column Damage at Beam-to-Column Intersection (End of Test) ................. 330

Figure 7.6.1: HSS-25 Net Section Tearing (0.88% Drift Level) ............................................................ 331

Figure A.2.1: Gusset Plate Geometry for Elliptical Clearance Requirement (Kotulka 2007) .......... 344

Figure A.2.2: Buckling Lengths for HSS-18 ............................................................................................. 346

Figure B.1.1: Specimen HSS-18 .................................................................................................................. 352





Figure B.1.6: Specimen WF-23 ................................................................................................................... 357


xix



xx

List of Tables

Table 1.2.1: Performance Based Objectives for SCBFs (Johnson 2005) ................................................. 3

Table 3.1.1: Thesis 1 Specimens (Johnson 2005) ....................................................................................... 21

Table 3.1.2: Thesis 2 Specimen (Herman 2007)......................................................................................... 22

Table 3.1.3: Thesis 3 Specimens (Kotulka 2007) ....................................................................................... 23

Table 3.1.4: Thesis 4 Specimen (Powell 2010) ........................................................................................... 24

Table 5.2.1: Component Damage Nomenclature ...................................................................................... 67

Table 5.2.2: Damage State Shading for Performance Based Design ...................................................... 67

Table 5.4.1: HSS-18 Peak Results ................................................................................................................. 78

Table 5.5.1: HSS-19 Peak Results ................................................................................................................. 91

Table 5.6.1: HSS-21 Peak Results ............................................................................................................... 100



Table 5.9.1: WF-23 Peak Results ................................................................................................................ 139

Table 5.10.1: HSS-24 Peak Results ............................................................................................................ 164



Table 6.2.1: Global Performance Summary ............................................................................................ 206

Table 6.2.2: Hysteresis Comparison ........................................................................................................... 207

Table 6.2.3: Elastic and Tangent Stiffness Summary .............................................................................. 211

Table 6.2.4: Backbone Curve Comparison ............................................................................................... 212

Table 6.3.1: Experimental vs. AISC Brace Forces ................................................................................... 217

Table 6.3.2: Brace Hysteresis Comparison ............................................................................................... 219

Table 6.3.3: Brace Resistance Summary .................................................................................................... 221

Table 6.3.4: NE Gusset Plate Rotation Comparison Summary ............................................................ 225

Table 6.3.5: Drift Range and Brace Shape Comparison ......................................................................... 230

Table 6.3.6: Weld Damage Propagation Summary .................................................................................. 232

Table 6.4.1: Framing Element Material Properties .................................................................................. 239

Table 6.4.2: Enveloped Frame Resistance vs. Drift Ratio ...................................................................... 242

Table 6.4.3: Frame Stiffness and Peak Resistance Summary ................................................................. 243

Table 6.4.4: Frame Hysteresis Comparison .............................................................................................. 244

Table 6.5.1: HSS-18 Equilibrium Evaluation ........................................................................................... 249

xxi





Table 6.5.6: WF-23 Equilibrium Evaluation ............................................................................................. 251




Table 6.5.10: Distribution of System Resistance...................................................................................... 254

Table 6.5.11: Resistance Distribution Summary ...................................................................................... 257

Table 6.6.1: Percent Total Energy Dissipation Summary ...................................................................... 262

Table 6.6.2: Energy Dissipation of Brace vs. Frame ............................................................................... 263

Table 6.7.1: Brace Performance Comparison ........................................................................................... 265

Table 6.7.2: Gusset Plate Performance Comparison .............................................................................. 267

Table 6.7.3: Framing Element Performance Comparison...................................................................... 268

Table 7.1.1: Design Summary of Reference Specimens for Comparison ............................................ 271

Table 7.1.2: Peak Performance Summary for Reference Specimens .................................................... 272

Table 7.1.3: Specimen HSS-01 Hysteresis and Gusset Plate Detail ...................................................... 273





Table 7.2.1: Peak Performance Summary Comparison for TRGPs ..................................................... 281

Table 7.2.2: TRGP Experimental Buckling Capacity Comparison ....................................................... 285

Table 7.2.3: TRGP Brace Responses ......................................................................................................... 286

Table 7.2.4: TRGP Brace Compressive Capacity Degradation Ratio................................................... 287

Table 7.2.5: TRGP Distribution of Resistance ........................................................................................ 295

Table 7.2.6: Summary of TRGP Resistance Distribution ...................................................................... 296

Table 7.2.7: TRGP Energy Dissipation ..................................................................................................... 296

Table 7.2.8: Summary of Energy Dissipation for TRGP ....................................................................... 297

Table 7.3.1: Peak Performance Summary for TTGP .............................................................................. 300

Table 7.3.2: TTGP Experimental Buckling Capacity Comparison ....................................................... 303

Table 7.3.3: TTGP Brace Response ........................................................................................................... 303

Table 7.3.4: TTGP Brace Compression Capacity Degradation Ratio .................................................. 304

Table 7.3.5: TTGP Distribution of Resistance ........................................................................................ 311

xxii

Table 7.3.6: Summary of TTGP Resistance Distribution ...................................................................... 312

Table 7.3.7: TTGP Energy Dissipation Comparison .............................................................................. 313

Table 7.3.8: Summary of Energy Dissipation for TTGP ....................................................................... 313

Table 7.4.1: Peak Performance Summary for WF Comparison ............................................................ 315

Table 7.4.2: Wide-Flange vs. HSS Experimental Buckling Capacity Comparison ............................. 316

Table 7.4.3: Wide-Flange vs. HSS Brace Response ................................................................................. 317

Table 7.4.4: WF Energy Dissipation Comparison ................................................................................... 323

Table C.2.1: Brace OOP Displacement Hysteresis ................................................................................. 361

Table C.2.2: Enveloped Brace OOP Displacements .............................................................................. 363

Table C.2.3: NE and SW Gusset Plate Rotation Hysteresis .................................................................. 364

Table C.2.4: Enveloped NE and SW Gusset Plate Rotations ............................................................... 366

Table C.2.5: Enveloped Brace Elongation ................................................................................................ 367

Table C.3.1: Column Moments Hysteresis at Edge of Gusset Plate .................................................... 369

Table C.3.2: Enveloped Column Moments at Edge of Gusset Plate ................................................... 371

Table C.3.3: Column Shear Hysteresis ...................................................................................................... 373

Table C.3.4: Enveloped Column Shears.................................................................................................... 375

Table C.3.5: Beam-to-Column Shear Tab Rotations............................................................................... 377

Table C.3.6: Enveloped Shear Tab Rotations .......................................................................................... 379

1

Chapter 1: Introduction

1.1 Background

Special Concentrically Braced Frames (SCBFs) are lateral load resisting systems for steel structures

frequently used in high seismic regions of the US. A structure subjected to seismic ground motions

is relied upon to resist the inertial forces generated from the building’s dynamic response due to the

ground motion. SCBFs are considered highly effective during frequent less severe earthquakes

because of they provide large lateral resistance and elastic stiffness by resisting lateral load through

the axial stiffness of the bracing members, . However, during less frequent, more severe

earthquakes, it is uneconomical to design the system to remain elastic, and the structure must

sustain large inelastic displacements and dissipate the induced energy to assure life safety and

collapse prevention.

Since the damage observed to moment frames during the 1994 Northgate, CA and 1995 Kobe,

Japan earthquakes, engineers frequently use SCBFs to resist seismic loads for construction of steel

structures. The focus of this research is to better understand the nonlinear demands on SCBFs

subjected to severe seismic loading and to improve the performance of the system to achieve

ultimate drift capacities.

The inelastic response of SCBFs consists of tensile elongation of the brace and buckling and post-

buckling deformation of the brace. The response of a braced frame with a single diagonal is

unsymmetrical and therefore, by design, opposing braces are implemented to assure equal

resistance and ductility in both directions. The lateral load verses horizontal story drift hysteretic

responses of a single diagonal brace and of a frame with opposing brace is illustrated in Figure

1.1.1.

2

1.2 SCBF Design Theory

Concentrically brace frames (CBFs) essentially act as vertical trusses incorporated into steel

structures providing lateral resistance to horizontal loads from wind or earthquakes. Lateral loads

are delivered to the lines of bracing through the structural floor, typically composite concrete slabs

on metal decking connected to the beams and girder with steel shear connectors. The loads are

collected from the floor diaphragms and pass through the braces down to the building foundations

and out of the structure. The term “special” in SCBF is added to refer to additional design and

detailing requirements for concentrically based frames used in seismic regions to assure an inelastic

response of the system.

Performance Based Design (PBD) criteria have been proposed to evaluate the full nonlinear

response of SCBFs to better predict post-buckling behavior and improve performance for various

Single Brace Response

Opposing Brace Response

Figure 1.1.1: Response of Single Brace vs. Opposing Braces

3

intensity seismic events. This is a similar approach to what has been done to improve the seismic

performance of steel moment frames (Roeder 2002), which allow specific levels of damage to occur

during the response based on the probability of occurrence of the earthquake intensity. The three

performance levels within Performance Based Design are Immediate Occupancy (IO), Life Safety

(LS), and Collapse Prevention (CP). Table 1.2.1 illustrates the three performance levels and the

damage considered acceptable for SCBFs.

To meet the Immediate Occupancy performance level, the resistance and functionality of the

system cannot be jeopardized and the structure needs to be able to achieve a similar response to

aftershocks or a similar magnitude event. The amount of acceptable damage to the system

increases for less frequently occurring earthquakes.

Under the Life Safety performance level, large buckling displacements and tensile yielding of the

brace would be expected, as well as significant inelastic rotations of the gusset plate. However, the

damage should not approach system failure. Life Safety also relates to nonstructural damage that

could result in injury or death such as falling architectural components or utilities. The level of out-

of-plane displacement of the brace in buckling, especially if immediately adjacent to the exterior of

the building, could cause damage to the façade and send debris falling on people below. For this

reason, out-of-plane displacement is a relevant performance parameter for SCBFs.

For the most severe expected earthquakes, severe and irreparable damage to the lateral force

resisting system is anticipated but the goal is to avoid partial or total collapse of the structure.

Large inelastic strains due to hinging and large tensile yielding of the brace and significant gusset

Table 1.2.1: Performance Based Objectives for SCBFs (Johnson 2005)

4

connection yielding would be expected but brace fracture should be delayed as long as possible.

This research focuses on the gusset plate connection and adjacent beam to column connections to

delay the large inelastic strains associated with brace plastic hinging and extend the life of the brace

to achieve large drift ranges expected during severe seismic events.

The design of SCBFs is a capacity based design approach with the brace as the controlling

component. Brace connections and adjacent framing elements are design to provide the required

strength and ductility to resist the maximum expected forces developed in the brace and to allow

the brace to buckle. The AISC Seismic Provisions define the maximum expected brace forces with

the brace in tension and compression based on Equation (1.2.1) and Equation (1.2.2), respectively.

(1.2.1)

(1.2.2)

The maximum expected brace force in tension is defined using the brace cross sectional area, ,

the specified yield stress of the brace material, , and the ratio of expected yield stress to the

specified yield stress, , based on the variability of material properties. The maximum expected

compressive force uses the nominal buckling capacity, , of the brace as calculated using AISC

Design Specification (Equation E3-1).

The brace force is typically transferred from the brace to the framing elements using gusset plate

connections. For braces buckling out-of-plane, gusset plates connections consist of flat plates

connected in-plane to the framing elements. The brace is connected directly to the gusset plate,

and the current AISC Design Specification and Seismic Provision require gusset plates to be

designed so that the maximum expected brace forces is less than for all failure modes. The

gusset plates are sized that Equation (1.2.3) is satisfied.

(1.2.3)

This presents a fundamental flaw in the design of SCBFs. Using a strength design for gusset plate

connections has no relation to the required ductility of the connection. By designing the gusset

plates for only strength and not for more realistic parameters considering a ductile response, the

AISC design procedure assures the brace with buckle and yield in tension but does not address the

behavior beyond that.

An alternative design procedure for gusset plate connection is discussed within this thesis that

considers a hierarchy of yielding to provide a more ductile response. Further discussion and detail

5

of the development of the BDP is available in the previous theses (Johnson 2005, Kotulka 2007).

The Balanced Design Procedure (BDP) is used to design the connections to achieve additional

ductility from desired yield mechanism while suppressing the occurrence of less desirable failure

modes (Johnson 2005). Figure 1.2.1 shows the hierarchy of inelastic behavior for SCBF systems

desired to increase the total ductility and energy dissipation.

Figure 1.2.1: Desired Hierarchy of Inelastic Behavior for SCBFs

The BDP addresses yield mechanism and failure modes similar to AISC strength design by

assigning balance factors, β, to the nominal resistance of the connection. Unlike AISC strength

design which evaluates demand verses resistance to yield mechanism and failure modes individually,

the BDP considers the resistance of the connection with regards to each interdependently to assure

yielding occurs where desired. The mean yield resistance of the connection using the Balances

Design Procedure is represented is Equation (1.2.4). The resistance to failure modes is shown in

Equation (1.2.5).

(1.2.4)

(1.2.5)

Yield mechanisms are associated with inelastic component behavior that result in yielding or

deformation but create minimal risk of strength loss. Failure modes, on the other hand, are

associated with inelastic behavior and brittle failure where strength of the component is lost

suddenly and cannot be recovered. The yield mechanisms and failure modes for SCBFs are

illustrated in Figure 1.2.2a and Figure 1.2.2b.

Brace Buckling Tensile BraceYielding

Gusset PlateYielding

Beam andColumn Yielding

Brace Fracture

6

Figure 1.2.2: Yield Mechanisms and Failure Modes for SCBFs

Larger ultimate drift ranges for SCBFs can be achieved by extending the ductile behavior beyond

buckling and tensile yielding of the brace and into the gusset plates and framing elements. The

experimental results from the previous test series have been used to develop the current β factors

recommended within the BDP and have resulted ultimate drift capacities larger than what has been

seen for specimens design following AISC design specification and recommendations for SCBFs.

1.3 Research Objectives

The research within this thesis is part of the National Science Foundation program CMS-0619161,

International Hybrid Simulation of Tomorrow’s Braced Frame Systems. The specific objectives of this

research are to build upon the knowledge gained from the previous experimental results and

analytical work to further improve the performance of SCBF systems. The selection of test

parameters for investigation and the design of specimens was also greatly influence by

representatives with AISC to study SCBFs with bolted connection to increase constructability while

improving seismic performance. A list of specific objectives for this research is provided below.

Evaluate the balanced design procedure for gusset plate connections with alternative brace

sections

Evaluate bolted connection for SCBF designed with the balanced design approach

Investigate the influence of beam-to-column stiffness on SCBF system

Investigate additional failure modes and AISC requirements for SCBF detailing with

regards to net section fracture

Inelastic ShorteningDue to Post-buckling

Behavior of Brace

Yielding of Beamsand Columns at

Gusset Plate Edge

Tensile Yieldingof Brace

Yielding ofGusset Plate

Bolt HoleElongation

Gusset Plate NetSection Fracture

Brace NetSection Fracture

BlockShear

Severe Weld TearingBolt

Fracture Gusset PlateBuckling

b) Failure Modesa) Yield Mechanisms

7

1.4 Overview of Report

This thesis documents nine experimental tests completed at the University of Washington

Structures Laboratory between January 11, 2007 and July 10, 2008. A review of previous braced

frame research and their findings are discussed in the Chapter 2: Literature Review. Design issues

and the design procedures for the nine tests within this series are included in Chapter 3: Design.

Chapter 4: Experimental Test Setup describes the physical test setup and instrumentation used to

test each specimen and record data for analysis. The events of each test are described in Chapter

5: Experimental Results, including the methods and nomenclature for describing and documenting

damage as it occurred. Chapter 6: Performance Analysis describes the interpretation and analysis

used to compare performance of the specimens. Comparisons of these nine tests are made with

selected test from the previous test series to evaluate relevant test parameters in Chapter 7:

Comparison of Test Parameters. Finally, conclusion, design recommendations and a discussion of

future work are included in Chapter 8: Conclusions. Appendices are included for Design

Calculations (A), Construction Drawings (B), Instrumentation Drawings (C), and Data Reduction

(D).

8

Chapter 2: Literature Review

2.1 Introduction

After the 1994 Northridge Earthquake, insecurity regarding the reliability of special moment

frames has prompted increased usage of SCBFs as the primary seismic resisting system for low

to mid rise steel structures (Roeder 2002). Unfortunately, engineers do not fully understand the

inelastic response demanded of SCBFs under severe, infrequent seismic events. This section

discusses the progression of experimental research that has influenced and contributed to trends

in design philosophy and the 2005 AISC Design Specification and Seismic Provisions. The

majority of past experimental research has focused on brace and gusset plate behaviors and their

failure modes, but little has been done to study the response of the entire system as a whole,

including realistic and varying framing element parameters. This chapter is divided into previous

research related to brace behavior (Section 3.1), gusset plate behavior (Section 3.2), and system

behavior (Section 3.3). The 9 tests included in this thesis are an extension of the related work

completed at UW by Johnson (2005), Herman (2006), and Kotulka (2007). A brief discussion of

their finding is included in Section 3.4.

2.2 Brace Behavior

The behavior of the brace is the most influential to the overall performance of the SCBF

systems. Over the last four decades, extensive experimental research has been completed

evaluating the factors that contribute to improving energy dissipation, achieving larger drifts and

extending fracture life for various brace sections under quasi-static cyclic loading. The current

AISC Seismic Provisions place limitations on brace slenderness and section width-to-thickness

ratios based on the finding, such as Astaneh-Asl (1982), Aslani and Goel (1989), Gugerli and

Goel (1982) and Walpole (1996).

Further work evaluating cold formed hollow steel sections (HSS) as bracing members are

summarized in Shaback and Brown (2003) and Trembley (2002). HSS braces have been

common in recent construction of low to mid rise SCBF structures because of their simple

connections and efficient buckling resistance to area. Their higher radii of gyration and

resistance to local buckling make them a favorable choice to design engineers over other rolled

sections with equivalent area, such as angles or channels.

9

Shaback and Brown (2003) tested 9 square HSS brace sections of varying sizes and lengths with

slotted gusset plate connections. The main parameters for this test sequence were effective

slenderness ratios, width-to-thickness ratios, and, to a degree, end connection fixity. Slenderness

ratios, KL/r, ranged between 52.3 and 65.8, width-to-thickness ratios, b/t, ranged from 8.93 to

14.60 and the gusset free length varied 1.26 and 2.09 times the gusset plate thickness. Figure

2.2.1 illustrates the schematic setup of the typical test specimen.

Figure 2.2.1: Shaback and Brown Test Specimen

Similar trends in hysteretic behavior were noted for all specimens including degradation of

compressive capacity observed after initial buckling occurred. Loss of axial compression

stiffness and residual axial displacement were also consistently observed regardless of test

parameters. The effective slenderness ratio had the most influence on the hysteretic behavior

and energy dissipation capabilities. Specimens with a higher KL/r showed greater degradation in

compressive capacity and a more pinched compressive loops, and therefore dissipated less

energy. The specimens with lower effective slenderness degraded less after initial buckling and

reloaded at a greater stiffness in tension which created fuller hysteretic loops. Slightly higher

tensile resistances were also observed at the larger drifts.

The fracture life of the brace was most influenced by the width-to-depth ratio. Specimens with

lower b/t sustained more cycles and reached larger drift ranges under this specific loading

protocol. Yield strength and effective slenderness ratio showed minimal effect on improving

fracture life.

guss

et w

idth

gusset free length

total gusset length

L/2

HSS brace

gusset plate

Elevation View

Plan View

10

Trembley (2002) completed a much more comprehensive and thorough evaluation into the

inelastic response of steel bracing members under cyclic loading. The study spanned 9 test

programs and 76 individual test specimen, 41 of which were hollow steel sections (HSS). The

intent of this research was to further evaluate the implications of KL/r and b/t and their affects

on hysteretic behavior, energy dissipation and fracture life, as well as developed equations that

could be used by designer to assure that adequate performance levels are met.

Current design methods using AISC Seismic Provisions for beams and columns in SCBF system

account for the combination of the maximum expected brace forces in addition to gravity loads.

Accurate predictions of the maximum forces in tension and compression are necessary, but also

knowing the minimum compressive capacity of the brace post buckling is needed to determine

when the unbalanced loading condition is at its worst depending on the bracing configuration.

An example where more accurate minimum compressive strengths of the buckled brace is

necessary is in the design of the beams at the intersection of chevron braces toward maximum

expected drifts.

Trembley proposed methods that more accurately predict the maximum compressive capacity,

the minimum compressive strength at increasing ductility levels capturing the nonlinear

regression and the expected achievable drifts before fracture of the brace occurs. An interesting

conclusion from the test data was that fracture life of HSS braces were strongly dependent on

the slenderness ratio and much less on width-to-thickness or induced displacement history,

which contradicts the conclusions of Shaback and Brown (2003). More slender braces were able

to achieve larger drift levels before fracture because as slenderness increases, the strain demand

at the plastic hinge is reduced. It was recommended that minimum slenderness ratios be

established and more rigorous criteria for width-to-thickness ratios should be placed on braces

with smaller slenderness ratios in order to achieve a desire drift level.

Out-of-plane displacement as the brace buckles has been recognized as a potential life-safety

issue where damage to adjacent partition walls or exterior cladding could result in loss of life

during a seismic event (Trembley 2002). Knowing the maximum out-of-plane displacement of

the brace for the expected drift level would allow designers to provide adequate space for the

brace to buckle freely within the wall cavity. An equation to accurately predict the displacement

depending on the expected maximum drift was also presented.

11

2.3 Gusset Plate Behavior

Gusset plates are used to connect the ends of the brace to the framing elements and are

depended upon to transfer the axial compressive and tensile loads as well as provide adequate

rotation as the brace buckles out-of-plane. Research into the behavior and potential failure

mechanisms of gusset plates by Bjorhovde and Chakrabarti (1985), Cheng et al (2000), Grondin

et al (1999), Nast et al (1999) Whitmore (1952) has influenced the current AISC design

specification by addressing the strength and ductility demands for SCBF systems.

Gusset plate geometry can be rectangular or tapered and connections to the framing elements

can be shared by both the beams and columns, or to either the beam or the column only. Brace

axial forces are transferred through an effective width in the gusset plate established by

Whitmore (1952) and referred to as the “Whitmore effective width”. The connection length

between the brace and gusset plate is used to determine the Whitmore effective width by

extending lines 30o from the start to the end of the connection as shown in Figure 2.3.1.

Additional experimental investigation and finite element analysis by Hardin (1958), Davis (1967)

and Varsarelyi (1971) confirmed that in general using the Whitmore width is a suitable approach

to determine the location of peak elastic stresses.

Figure 2.3.1: Whitmore Effective Width

Investigation into the inelastic behavior of gusset plates in tension completed by Bjorhovde and

Chakrabarti (1985) identified additional potential failure mechanism. 6 full-scale gusset plate

specimens were tested with bolted brace to gusset connections at 30o, 45o and 60o angles. The

Whitm

ore Effective

Width

Lengt

hCon

nect

ion

Brace Location

Gusset Plate

30°

12

typical test specimen is shown in Figure 2.3.2. Hydraulics rams were used to load the brace in

tension until failure of the system or until reaching the maximum capacity of the loading

apparatus. The results from this experimental research showed the potential for tearing across

the bottom bolt row of the brace to gusset connection or tearing from the bottom holes to the

edge of the gusset plate perpendicular to the brace length. This further validated the criteria for

designing maximum stresses across the Whitmore effective width. Hardash and Bjorhovde

(1985) proposed new design methods for gusset plates in tension and included a method for

calculating the block shear capacity based on tensile yielding across the bottom bolt row and

shearing along the length of the bolt configuration to the free edge of the gusset plate.

Figure 2.3.2: Bjorhovde and Chakrabarti Test Specimen (1985)

The method to predict the elastic buckling capacity of the gusset plate was developed by

Thornton (1984) and investigated through experimental and analytical research. The proposed

method calculated capacity by considering idealized column strips extending from the Whitmore

effective width at 3 locations, one at each end and at the brace center, labeled L1, L2 and L3 in

Figure 2.3.3. An effective length factor of 0.65 was used with the maximum strip length to

calculate a lower bound buckling capacity.

13

Figure 2.3.3: Thornton Effective Length for Buckling Capacity

Experimental work consisting of 13 tests was conducted by Cheng and Yam (1993) to evaluate

the uni-axial compressive capacity of gusset plates. The typical test setup is shown in Figure

2.3.4. The test parameters under investigation were plate thickness, plate size, angle of the brace,

out-of-plane brace restraint and moments in the framing elements. Plate thickness was the most

influential factor for gusset plate buckling capacity while brace angle and moments in the framing

elements were the parameters that showed minimal impact. An additional analytical study was

performed on the 13 specimen using finite element modeling program ABAQUS by Yam et al

(1998). A comparison of the experimental and analytical results was summarized in a paper by

Cheng et al (2000). It was determined that the Thornton method for calculating the buckling

capacity was overly conservative and did not include effects for redistribution of load as yielding

occurred. The modified Thornton method utilized a 45o angle from the start of the brace to

gusset connection and results in a wider effective width for calculating the maximum

compressive stresses. The buckling predictions using the modified Thornton method more

closely correlate with the experimental and analytical results.

L1

L2

L3

Whitm

ore Width

Brace Location

Gusset Plate

14

Figure 2.3.4: Cheng and Yam Test Setup (1993)

Research into the behavior of gusset plates subjected to cyclic loading has also been valuable to

develop current design specifications for SCBF systems. Jain et al (1978) began by investigating

the contribution of gusset plate bending stiffness on brace behavior by evaluating the

performance of 18 brace configuration with gusset plate connections. Three gusset plate

geometries were used and the clear length from the brace end to the beam to gusset connection

varied as one of the test parameters. It was found that increased stiffness of the gusset plates

reduces the slenderness ration of the brace and therefore improved the energy dissipation

capabilities of the system, similar to decreasing the slenderness ratio of the brace.

Astaneh-Asl (1982) tested 17 full-scale double-angle brace sections with gusset plate connection

in an idealized beam-column frame under cyclic loading. The gusset plates were connected only

to the beams and the framing elements were designed to keep the response elastic for repeat

tests. The typical test specimen is shown in Figure 2.3.5. It was observed that the behavior of

the gusset plates depended greatly on the direction of buckling in the brace. If the brace buckled

in-plane, plastic hinges formed at 3 locations; the midpoint of brace and in the brace just outside

of each gusset plate. For braces that buckled out-of-plane, the plastic hinge still formed at the

brace midpoint, but the two end hinges formed within the gusset plates.

15

Figure 2.3.5: Astaneh-Asl et al. Test Specimen

From the findings, it was recommended that gusset plates be designed with sufficient area at the

brace ends to allow for the formation of plastic hinges as the brace buckles out-of-plane. It was

seen that fracture of the gusset plate could possibly occur after only a few cycles if adequate

clearance to rotate are not provided. Astaneh-Asl (1986) recommended a 2t linear clearance

from the end of brace to a line from the closest re-entrant corner perpendicular to the brace

length, with t being the thickness of the gusset plate.

Aslani and Goel (1989) continued investigating double angle braces within the Astaneh-Asl test

setup shown above but also evaluated additional brace end conditions. Flexible as well as fully

fixed brace end connections were incorporated into specimen design. Increasing the end

connection fixity decreasing the slenderness ratio directly affects the energy dissipation and

buckling capacity of the brace. Adversely, this also limits the fracture life of the brace and could

lead to early failure.

2.4 System Behavior

There has been limited research investigating of the overall system behavior of SCBFs. Few full

scale experimental programs utilized test specimens that consist of brace, gusset plates and

realistically designed framing elements. Current AISC Seismic Provision do have design criteria

to account for the demands on columns and beams from maximum brace forces and brace

configuration but additional investigation is required into the inelastic demands on all

components as a complete system. The response of concentric braced frames cannot be

16

accurately characterized by simple pin-pin truss action. The interaction between gusset plates

and framing elements is difficult to represent in analysis without highly developed FE models.

Flexural demands on the framing elements adjacent to gusset plates and additional stresses on

corner gusset plates due to the opening and closing of beam to column connection at large frame

drifts add to the complex inelastic behavior of SCBF systems.

Uriz and Mahin (2004) evaluated a full scale two story SCBF test specimens with chevron brace

configuration subjected to pseudo-static cyclic loading. The design followed 1997 AISC Design

Specification and Seismic Provisions including 2t linear clearance for gusset plates. The objective

of the research was to obtain experimental data to better understand the relationship between

overall system, component and connection behaviors, validate and improve FE models, and to

evaluate the current design procedures and analysis methods. The test specimen was pseudo-

statically loaded at the upper beam only and subjected to a displacement controlled symmetric

protocol with increasing amplitude based on an induced drift corresponding to the critical

buckling force in the brace. Figure 2.4.1 illustrates the two story chevron SCBF test specimen

test at University of California Berkeley.

Figure 2.4.1: Uriz and Mahin Test Setup (2004)

The majority of inelastic response occurred at the first story with braces buckling in compression

and yielding in tension, while the second story remained relatively elastic. Plastic hinges formed

at the desired 3 locations over the brace length; at the midpoint and in the 2t region of the gusset

17

plates. Fracture eventually occurred at the brace plastic hinges for both first story braces. The

test continued after brace fracture in order to evaluate the remaining resistance of system.

Fracture of both first story columns occurred just below the intermediate beam with tearing

straight through the inside flanges and into the webs.

This test substantiated the concerns for a potential soft story forming once buckling occurs for

multi-story SCBFs. Even though the response of the first story braces closely match those seen

in previous research with simple brace and gusset plate test specimen, the global response and

resulting drift capacity were poor. Addition research was recommended into varying brace

configurations for multi-story test specimen, further consideration into the fixity of beam to

column connections to improve frame action post brace fracture, and the development of more

efficient gusset plate design.

2.5 Previous UW SCBF Research

The objective of the SCBF research at the University of Washington beginning in 2004 was to

perform a comprehensive parameter study for corner gusset plates in a realistically sized frame

representative of a low-to-mid rise steel structure. AISC Design Specification and Seismic

Provisions recognize the brace as the primary inelastic component and rely on yielding in tension

and buckling to dissipate energy and achieve adequate story drifts. The designs of gusset plates

are to meet the strength demands of the brace and incorporate the 2t linear clearance for out-of-

plane rotation as the brace buckles but designs can result in large, overly stiff connections. It is

thought that current design procedures for gusset plates could potentially lead to soft stories,

unexpected failures modes and early brace fracture (Roeder et al. 2004).

Three previous SCBF experimental test programs by three graduate students (Johnson 2005,

Herman 2007, and Kotulka 2007) consisted of full scale single story, single bay test specimen

with a diagonal brace section subject to a pseudo-static cyclic loading protocol. A total of 17

tests were conducted. The typical specimen consisted of a single HSS5x5x5/8 brace connected

to corner gusset plates welded to a 12’ by 12’ beam and column frame sized to reflect the actual

gravity and lateral demands of a low-to-mid-rise structure. Designs were intended to investigate

various test parameters including gusset plate thickness, gusset plate shape, interface weld design,

alternate clearance requirements for out-of-plane rotation of gusset plates, framing element

stiffness, and beam to column connection detailing. The typical UW test specimen is shown in

Figure 2.5.1.

18

Figure 2.5.1: Roeder et al Test Specimen (2008)

All of the conclusions and recommendations are available in the three theses (Johnson 2005,

Herman 2007, and Kotulka 2007). Some of the key conclusions are discussed below. Results

indicated that more compact and economical gusset plates can be designed using an elliptical

clearance at the brace end and still improve the system ductility, increasing the drift capacity and

delaying fracture of the brace. Figure 2.5.2 is an example illustrating the elliptical clearance

defined by N*tg, with 8*tg showing the most promising results. Tapered gusset plates also

showed better out-of-plane rotation capabilities but subjected the interface welds to greater

demands and higher potential for cracking. More importantly, gusset plate to framing element

interface welds should be designed for the full plastic capacity of the plate and not for the

demands from the brace as calculated by the Uniform Force Method or KISS Method (“keep it

simple stupid”). Also, the size and stiffness of the beam element has considerable impact on the

global response and total drift capacity of the system.

19

Figure 2.5.2: Gusset Plate Elliptical Clearance

One of the ongoing objectives of this research has been the development of the Balance Design

Approach for SCBF gusset plate design, which is the method of balancing yield mechanisms and

preventing undesirable failure modes with β factors. A hierarchy of yielding is established to

maximize the ductility of the system while also suppressing early system failure. This will be

discussed in detail in the following chapter.

In conjunction with the experimental work at UW was the development of a highly sophisticated

FE modeling method using ANSYS to predict the local and global response of SCBFs (Yoo

2006). This analytical research also included an approach to predict brace fracture based on an

equivalent plastic strain at the brace plastic hinge. The results from the experimental tests were

used to improve accuracy of the model in capturing the global and local responses of the system.

As research continues into the holistic behavior of SCBFs as an integrated system of

components, the economical and practical means to experimentally evaluate multi-story test

specimen with varying test parameters become increasingly difficult and accurate analytical tools

emerge as increasing valuable.

20

Chapter 3: Specimen Design

3.1 Introduction

The specimen design is discussed in this chapter. The single story single bay specimen is

representative of a bottom story of a low-rise, or an intermediate story of a mid-rise SCBF

structure. Figure 3.1.1 shows the typical specimen with member sizes. The original specimen

prototype was designed by Shawn Johnson with influence from the current AISC codes, the SAC

building model, and technical advice from practicing engineers. Further discussion regarding

member selection and original details are available in his thesis (Johnson 2005). The Balance

Design Procedure for gusset plate design discussed in Section 3.2 including the procedure for

gusset plate design used for the nine specimens within this test series. Specimen design parameters

having special considerations are described in Section 3.3. Examples of specimen designs are

presented in Appendix A.

Figure 3.1.1: Prototype Specimen (Johnson 2005)

The design procedure and specimen selection for this test series built off the results from previous

tests within this program. A brief summary of the design parameters evaluated for each of the

previous 17 tests are presented in the tables below. Table 3.1.1 shows the five specimens first

tested by Shawn Johnson. Table 3.1.2 summarizes the designs for the six specimen tested by David

21

Herman. The six specimen tested by Brandon Kotulka are shown in Table 3.1.3. The designs for

the nine specimens tested within this test series are presented in Table 3.1.4

Table 3.1.1: Thesis 1 Specimens (Johnson 2005)

22

Table 3.1.2: Thesis 2 Specimen (Herman 2007)

23

Table 3.1.3: Thesis 3 Specimens (Kotulka 2007)

24

Table 3.1.4: Thesis 4 Specimen (Powell 2010)

HS

S-1

8 Specimen was designed to evaluate

performance with the bolted shear

plate beam-to-column connection

with rectangular gusset plates

HS

S-1


bolted brace to gusset plate

connections for improved

constructability

HS

S-2

0

Specimen was designed with a

bolted beam end plate connection

for improved constructability

HS

S-2

1

Specimen was designed with a

bolted beam end plate connection

for improved constructability and

reduced bolt configuration

HS

S-2


performance with the bolted shear

plate beam-to-column connection

with tapered gusset plates

1'-

9"

1'-23

4"

2'-1"

12" PLATE

1" A490X BOLTS

38" GUSSETPLATE

5/16"

3/8"

5/16"

1/2" GUSSETPLATE

1'-23

4"

3/4" x 12" PLATE

WT4x17.5 (BOTH SIDES)

1-1/8"Ø A490 BOLTS

1/2"

5/16"1'-0"

31

2"

4 34 "

1'-03

4"

1'-11"

1'-

99 16"

1'-

9"

1'-23

4"

2'-0"

4"

38" GUSSETPLATE

3/8"

5/16"

18- 34" A490X BOLTS

1" END PLATE

1'-

9" 3

8" GUSSETPLATE

3/8"

1'-23

4"

2'-0"

5/16"

4"

14- 34" A490X BOLTS

1" END PLATE

1'-23

4"

1'-

41 4

"

3/8"

1'-61116

"

5/16"

12" PLATE

1" A490X BOLTS

38" GUSSETPLATE

5/16"

25

WF

-23 Specimen was designed to evaluate

performance with a more ductile

wide-flange brace section

compared to the HSS tube

H

SS

-24 Specimen was designed to evaluate

performance with welded beam

flanges and bolted web at the

beam-to-column connections

HS

S-2

5

Specimen was designed to increase

likelihood of net section fracture at

brace to gusset connection

HS

S-2

6 Specimen was designed to increase

likelihood of net section fracture at

brace to gusset connection

identical to HSS-25

3.2 Balanced Design Procedure for Gusset Plates Connections

This section describes the balance design procedure (BDP) used to design the gusset plate

connections for the nine specimens within this test series. The BDP is an alternate to the current

AISC resistance design which addresses strength alone and can neglectfully result in gusset plate

connections overly stiff and prone to brittle failure modes. The current design procedure

recommended in the AISC Specifications utilizes resistance factors, φ, to assure adequate strength

to statically extreme loading conditions. However, this could be at the expense of system ductility

3 12 "

1'-

8"

38" GUSSETPLATE

3/8"

1'-1

"

1'-978"

5/16"ALL AROUND

38" WEB PLATEEA. SIDE

5/16"12" PLATE

1" A490X BOLTS

1'-

9" 3

8" GUSSETPLATE

3/8"

1'-23

4"

2'-1"

5/16"

BACKGOUGE2'-

01 2

"

78 " GUSSETPLATE

1'-23

4"

2'-412"

5/16"

W16x89

BACKGOUGE

2'-

01 2

"

78 " GUSSETPLATE

1'-23

4"

2'-412"

5/16"

W16x89

26

and a more ductile connection can provide adequate resistance while improving the overall

performance of the system.

The fundamental theory of the BDP is to balance resistance to desired yielding mechanism while

suppressing undesirable failure modes to increase the overall ductility of the system and control

behavior at severe drift ranges. Resistances to limit states using the BDP address both strength and

ductility in order to establish a hierarchy of yielding. The desired ductile response is to extend

yielding beyond the brace and into the gusset plates and framing elements. Greater detail into the

development of the BDP is available in the theses of Johnson (2005), Herman (2007) and Kotulka

(2007).

The BDP uses balance factors, β, applied to yield mechanism and failure modes to encourage

yielding when desired while delaying the occurrence of undesirable failure modes from controlling

the response of the system. Similar to φ factors in resistance design, β factors are no larger than 1.0

and lower the value, the more conservative the designed resistance with respect to the calculated

nominal resistance, . The balance equation for determining the mean yield resistance of the

connection is shown in Equation X.

(3.2.1)

The use of the material overstrength factors, , in the equation is to account for actual stresses of

the connection material to assure desired yielding occurs. The approach to balance resistance to

failure modes is similar, except that material overstrength factors are not included as shown in

Equation Y.

(3.2.2)

This section discusses the general steps used to design gusset plate connection using the BDP with

commentary and equations. The nominal resistances to limit states were calculated using

procedures within the AISC Design Specification and replace φ factors with β factors. Some

variations, such as interface weld designs, are describe in detail. The values for β factors are based

on the experimental results from the previous test series and follow the design recommendations

within Kotulka’s thesis (2007).

1. Determine brace sizes and forces from seismic analysis of system (Section 3.2.1).

27

2. Design brace to gusset plate connection and length (Section 3.2.2).

3. Determine preliminary gusset plate thickness based on whitmore yielding, whitmore

fracture and block shear (Section 3.2.3).

4. Determine preliminary gusset plate geometry using Elliptical clearance requirement

(Section 3.2.4).

5. Check gusset plate buckling using Thornton Method with preliminary thickness and

geometry (Section 3.2.5). Re-determine gusset plate thickness if required.

6. Design net section reinforcement based on actual slot thickness (Section 3.2.6).

7. Design interface welds for full capacity of gusset plate (Section 3.2.7).

3.2.1 Brace Forces

Brace sections used for SCBFs must meet seismic compactness criteria within the AISC Seismic

Design Manual in Table I-8-1 for resistance to local buckling. HSS tubular brace sections must

have a width-thickness ratio, or , that is less than Equation (3.2.3). Wide-flange brace

sections must have a flange width-thickness ratio, , less than Equation (3.2.4) and a web

width-thickness ratio, , less than Equation (3.2.5).

(3.2.3)

(3.2.4)

(3.2.5)

The limiting width-thickness ratios, , are used to determine if a brace (or column) section is

seismically compact and can be used in the design of an SCBF. The variable is Young’s modulus

of elasticity and is the minimum yield stress of the brace material.

The gusset plate connections are design to provide the required tensile and compressive resistances

to brace forces calculated using Equations (3.2.6) and (3.2.7) from AISC Seismic Design Manual

(Section 13.3). The term is the ratio of the expected yield stress to the minimum specified yield

stress, , accounting for variability in the properties of the brace material. The maximum expected

force in tension, , is the expected yield stress, , multiplied by the gross area of the brace

28

cross section, , and is used to determine required strength of tensile limit states. The maximum

expected brace force in compression is times the nominal compressive strength, .

Although the design manual does not specify the method for determining the effective length of

brace for SCBFs, using an effective length factor, , of 1.0 and the actual brace length for

calculating the nominal compressive strength was recommended as an acceptable method by

practicing engineers.

(3.2.6)

(3.2.7)

3.2.2 Brace to Gusset Plate Connection

The connection between the brace end and the gusset plate can be either welded or bolted. HSS

tubular sections are typically slotted and welded directly to the gusset plate with 4 fillet welds. This

is the connection method typically used to connection the HSS5x5x3/8 brace to the gusset plates

and is described in this section. The resistance of the connection must be greater than the

maximum expected brace force in tension as determined in Equation (3.2.6). This is determined as

the lesser of the shear capacity of the welds and the shear rupture capacity of the brace base

material as calculated using Equations (3.2.8) and (3.2.9), respectively. Similar to the specified AISC

resistance factor for both resistances of the welds and of the base material, a balance factor, β, of

0.75 is recommended.

(3.2.8)

(3.2.9)

The variables and are the nominal strength of the weld material and the effective throat

thickness of the fillet weld. The wall thickness of the brace is given as . The number of welds

or the number planes considered is given as . The total length of the splice in both equations is

Using a shorter splice length will reduce gusset plate geometry and in the case of A500 B/C

tubular brace section with nominal yield strength of 46ksi connection to A992 plate steel, the

rupture capacity of the base material controls. The required splice length can be determined from

Equation (3.2.9) and used to size the welds.

29

3.2.3 Tension Limit States for Gusset Plates

The geometric parameters to determine a minimum gusset plate thickness is available once the

brace section and splice length are determined. The preliminary thickness of the gusset plate is

determined based on the resistances to tension limit states: whitmore yielding, whitmore rupture

and block shear. A fundamental part of the BDP is to encourage gusset plate yielding over the

whitmore effective width. This was achieved by using a balance factor, β, of 1.0 for whitmore

yielding. Figure 3.2.1 shows the method for determining the whitmore effective width, , by

extending 30o lines from the start of the brace splice welds to the brace end.

Figure 3.2.1: Whitmore Effective Width

The resistance to whitmore yielding is calculated using Equation (3.2.10), where is the plate

thickness. Similarity, resistance to shear rupture is calculated using Equation (3.2.11), where β is

0.85 and is minimum specified tensile strength of the plate.

(3.2.10)

(3.2.11)

Block shear is the tensile limit state where the rectangular shape of the gusset plate bound within

the brace to gusset connection ruptures and separates from the remainder of the plate. Resistance

to block shear is determined as rupture along the net shear path, , and along the net tensile

path, . Figure 3.2.2 illustrates the area used to calculate block shear resistance for gusset plates.

The resistance to block shear rupture is calculated using Equation (3.2.12) where the factor

Whitm

ore Effective

Width

Lengt

hCon

nect

ion

Brace Location

Gusset Plate

30°

30

accounting for distribution of tension stress, , is equal to 1.0 assuming tensile stress is uniform.

The recommended β factor for block shear is 0.85.

Figure 3.2.2: Block Shear Area

(3.2.12)

3.2.4 Gusset Plate Geometry and Elliptical Clearance

The dimension of the gusset plate are determine by providing the elliptical clearance line that

follows the pattern of yielding observed as the gusset plate rotates out-of-plane. This is

accomplished using the preliminary gusset plate thickness determined above and the general

geometry of the adjacent beam to column connection and the angle of the brace. This method,

which was used to design the gusset plates within this test series, was purposed and described in

detail within the thesis design recommendation of Kotulka (2007). The elliptical clearance for

corner gusset plates recommended from the previous experimental results is between 6tp and 8tp

(Johnson 2005 and Kotulka 2007). Figure 3.2.3 shows the gusset plate with variables used to

calculate plate dimensions described in the procedure below. The process is iterative and was

automated during the design of the specimen using Microsoft Excel.

Lnv

Brace Location

Gusset PlateLnt

31

Figure 3.2.3: Gusset Plate Geometry for Elliptical Clearance Requirement (Kotulka 2007)

If using the AISC Uniform Force Method to determine the forces on the interface welds, the

height to width ratio of the gusset should be proportioned as to not result in moments acting on

the welds. This is achieved by using dimensions that meet the requirement in Equation (3.2.13)

from the AISC Manual in Chapter 13, where and are the depth of the beam and column,

respectively.

(3.2.13)

This adherently sets the variable equal to . The purposed method suggests making the free

edges of the plate intersect at the center line of the brace, or setting equal to 0. Since the

interface welds were not designed using the Uniform Force Method and were instead designed for

the full capacity of the plate as described in Section 3.2.7, it is considered acceptable to set equal

to 0. This also results in a slightly more compact geometry. However, to keep the gusset plate

design consistent with previous specimens, all specimens with the exception of WF-23 used gusset

plate geometry with variable equal to .

3.2.5 Gusset Plate Buckling

The area of the gusset plate between the brace end and the framing elements was designed to resist

maximum expected brace force in compression as calculated in Equation (3.2.7). The buckling

length was determined using the Thornton Method, which sets three lengths extending parallel to

32

the brace center-line from the whitmore effective width to the flanges of the framing elements.

This is illustrated in Figure 3.2.4.

Figure 3.2.4: Thornton Method for Gusset Plate Buckling

The gusset plate resistance to buckling, , is calculated using Equation (3.2.14) with a

recommended factor equal to 0.90. The nominal buckling strength, , was calculated following

the AISC Specifications Chapter E using equal to the average of the three lengths, and the

effective length factor, , equal to 0.5. The buckling area, , is the whitmore width, ,

multiplied by the plate thickness, .

(3.2.14)

The procedure to calculating the critical buckling stress, , is as follows:

(3.2.15)

Where is Young’s modulus of the plate material and is the radius of gyration of the whitmore

cross section. When , Equation Y is used to calculate .

(3.2.16)

If , is calculated using Equation (3.2.17).

L1

L2

L3

Whitm

ore Width

Brace Location

Gusset Plate

33

(3.2.17)

If the calculated resistance to buckling was less than the demand, the preliminary gusset plate

thickness would be deemed insufficient and a thicker plate would be required. Gusset plate

geometry would need to be re-determined in order to meet the designated elliptical clearance

requirement. However, gusset plate resistance to tension limit states would be adequate by

inspection.

3.2.6 Net Section Reinforcement

The slot in the brace that is slid over the gusset plates to connect the brace creates a reduced area

of the brace cross section as shown in Figure 3.2.5. The necessity of net section reinforcement has

been a design parameter evaluated within the previous test series as well as while in the thesis.

Experimental results have shown that specimens with gusset plate design using the BDP and tested

without additional reinforcing plate at the brace to gusset connection were capable of developing

the full capacity of the brace and ultimately failed due to brace fracture at the plastic hinge (Kotulka

2007). The specimens with typical HSS tubular brace sections designed within this thesis were

checked for resistance to net section fracture but reinforcement plates were not included during

fabrication. However, this section does illustrate the procedure for designing net section

reinforcement using a recommended β factor of 0.90. The calculated resistance is shown in

Equation (3.2.18).

Figure 3.2.5: Reduced Area at Brace to Gusset Connection

(3.2.18)

Gusset Platehole diameter 14" largerthan gusset plate thickness

Reduced BraceSection

34

The term is the ratio of expected tensile stress to the specified tensile stress, , for the brace

material. The reduced net section of the brace is . The tensile stress and the cross sectional

area of the reinforcement plate are represented as and . The term is a shear lag factor

from AISC Table D3.1 that is applied to account for shear lag effect, or concentration of shear

stress over the length of the connection, . The term is calculated in Equation (3.2.19) and

uses , the distance from the center of gravity of the brace to the center of gravity to the brace

section split by the slot including the net section area. The values Table D3.1 for are

approximations and actual values can be calculated using the reinforcing plate dimensions and slot

width iteratively. This resulted in slightly smaller plate area required for reinforcement.

(3.2.19)

3.2.7 Interface Weld Design

The designs of the interface welds connecting the gusset plates to the framing elements for the

specimen within this series are different than what has been recommended by the current AISC

procedure. The stress concentrations that occurs at the gusset plate reentrant corners from both

in-plane rotation of the beam-to-column connection and out-of-plane deformation of the gusset

plate is not considered when sizing the weld for the vertical and horizontal components of the

brace force based on the AISC Uniform Force Method. Sizing the weld for the full plastic capacity

of the gusset plate was recommended as part of the BDP to increase resistance to weld fracture.

Equation (3.2.20) was used to size the welds and was presented in the Kotulka thesis design

recommendation (2007).

(3.2.20)

The balance factor, β, used to design the interface welds is equal to 0.65, which is greater than the

resistance factor, φ, used for fillet welds. The right side of the equation showing the full plastic

capacity of the gusset plate does consider the material overstrength factor, . The on the left

of the equation is to account for additional weld capacity in tension. Although the applied brace

force is a combination of shear and tension, the stresses that result in tearing of the interface welds

a tensile stress due to the large deformation demand on the gusset plate.

Specimen HSS-25 and HSS-26 both utilized complete joint penetration weld to connection the

gusset plates to the framing elements. These also develop the full capacity of the gusset plates and

achieve the same effect as the fillet weld. From a design stand point, greater weld material is

35

required to fully develop the plastic capacity of the plate with fillet welds, but CJP welds require

additional consideration for preparing the plate edges and providing backing bars.

3.3 Design Parameters

This section discusses the design parameters that required special consideration during the design

of each specimen. The design of the bolted shear plate connection adjacent to the gusset plates

utilized in specimens HSS-18 and HSS-22 is described in Section 3.3.1. The design of the bolted

end plate connections for HSS-20 and HSS-21 is described in Section 3.3.2. Design consideration

specific for the wide-flange brace to gusset connection of WF-23 is provided in Section 3.3.3.

3.3.1 Bolted Shear Plate Beam-to-Column Connection Design

The designs of specimen HSS-18 and HSS-22 incorporated a bolted shear plate connection rather

than the CJP welded beam-to-column connection adjacent to the gusset plates. These two test

specimens are intended to evaluate modifying the beam-to-column connection method for

specimens with thin rectangular gusset plates and thin tapered gusset plates to assess the effect on

system performance. Specimen HSS-24 also assesses the beam-to-column connection by utilizing a

bolted web and CJP welded flanges. The bolted shear plate connection is shown in Figure 3.3.1

and the bolted web CJP welded flange connection is shown in Figure 3.3.2.

Figure 3.3.1: Bolted Shear Plate Connection

36

Figure 3.3.2: Bolted Web CJP Welded Flange Connection Detail

Specimens HSS-18 and HSS-24 utilized the identical brace and gusset plate design as HSS-05 while

HSS-22 used the tapered gusset plate design identical to HSS-17. The beam-to-column connection

was designed to resist the resolved forces from the brace acting on the beam to gusset plate

connection using the AISC Uniform Force Method calculated in Equations (3.3.1) through (3.3.5).

(3.3.1)

(3.3.2)

(3.3.3)

(3.3.4)

(3.3.5)

The variable is the maximum expected brace force in tension, , calculated in Equation (3.2.6).

The variables , , , and are shown in Figure 3.3.3 and the resolved design forces are

shown in Figure 3.3.4.

37

Figure 3.3.3: Uniform Force Method

Figure 3.3.4: UFM Resolved Forces

The vertical and horizontal brace forces acting on the beam interface welds were combined into a

single force applied at an angle using Equation (3.3.6). The angle, , was calculated using

Equation (3.3.7). The force on the shear tab, , is applied eccentrically to the bolt group

the length of , determine as the horizontal distance between the bolt group center of gravity and

the center of gravity of the beam-to-gusset weld. The capacity of the bolts was determined using

AISC Construction Manual Table 7-7.

eceb

W.P.

e

Pu

Vub

Hub

Huc

Vuc

38

(3.3.6)

(3.3.7)

AISC Design Specifications were used to determine the connection resistance including φ factors

for the relevant failure modes.

The shear plate was designed to resist shearing across the reduced area with the bolt holes

and shearing across the gross area.

The fillet welds connecting the plate to the column flange were designed to resist

following AISC Design Specifications Table J2.5 for the nominal resistance of

welds loaded at an angle.

The capacities of the W16x45 beam section were also checked for adequate resistance to

web shearing and block shear where the web area is reduced, based on AISC Design

Specification J5.2 and J5.3.

3.3.2 Bolted Beam End Plate Connection Design

Specimens HSS-20 and HSS-21 utilized bolted beam end plate connections to the columns adjacent

to the gusset plates. These design parameters evaluate the performance of the specimen with

similar gusset plate design to HSS-05 with an alternate bolted beam-to-column connection. The

beam section of the connection is shown in Figure 3.3.5.

39

Figure 3.3.5: Bolted Beam End Plate Connection

The bolted connections were designed to meet strength requirements based on the maximum

capacity of the connecting elements. The required rotational resistance of the bolted connection

was the nominal plastic moment, , of the W16x45 beam section. The required strength of the

bolted connection was design to resist the combined shear and tension resulting from the vertical

and horizontal components of the maximum expected brace force, , from Equation (3.2.6). The

flexural capacities of the bolted configurations were calculated following the AISC Manual Part 7

Design Considerations for Bolts using the instantaneous center of rotation method.

The beam web was fillet welded on each side directly to the 1”x10” plate. The fillet weld was

designed for the nominal shear capacity of the W16x45 beam, , determined in the AISC Design

Specifications Section G.

3.3.3 Wide-Flange Brace to Gusset Connection Design

The connection between the W6x25 wide-flange brace section and the gusset plates was more

intricate than the simple slotted brace of the typical HSS tube and required additional consideration

during the design. To minimize that amount of additional connection plates and welding, it was

decided to slot and weld the flanges and connect the web using plates. The wide-flange brace to

gusset plate connection is shown in Figure 3.3.6. The method for designing the gusset plates and

framing element connections for specimen WF-23 was the same as previous specimens with the

1'-518"

1'-

9"

2'-0"

29 16"

3'-

21 8"

4"

512"

10"

1'-

21 8"

1" PLATE

40

typical HSS tubular brace. A design example of the wide-flange brace to gusset connection is

included in Appendix A.

Figure 3.3.6: Wide-Flange Brace to Gusset Detail

The required splice length was determined to resist the maximum expected brace force calculated in

Equation (3.2.6) for the W6x25 brace section. The length was controlled by the rupture capacity of

the welds or the flange base material shown in Equations (3.3.8) and (3.3.9) using a balance factor,

β, of 0.75 for both calculations.

(3.3.8)

(3.3.9)

In both equations, the length of the splice connection was given as . The strength of the welds

here was calculated similarly to those for the HSS brace to gusset splice in Equation (3.2.8). The

strength of the base material calculated in Equation (3.3.9) used the flange thickness, , and the

number of shear planes, , taken as four for the slotted wide-flange. The actual splice length in the

design of specimen WF-23 was 13 in., which was larger than what was required by design in order

to result in gusset plate geometry comparable to previous specimens with HSS braces.

The web plates provided continuity between the web and the gusset plate and act as net section

reinforcement from the reduced brace area from the slotted flanges. The plates were sized similarly

as net section reinforcement for the reduced section for the HSS brace in Section 3.2.6. The

variable names were modified for the web plates and shown in Equation (3.3.10). The total area of

the web plates is given as and the total reduced flange area as . A balance

factor, β, of 0.9 was used.

1'-1

"

38" GUSSET PLATE

5/16"

1" 3 12 "

61

2"

61

2"

SEE SECT A3 SIDESTYPICAL

38" GUSSET PLATE BRACE WEB

(.32" THICK)

SECTION A

1"

1/4"

1/2"

3 8"

5 8"

TYP.

38"x 31

2"x 14"WEB PLATEEACH SIDE

38"x 31

2"x 14"WEB PLATEEACH SIDE

14" SHIM PLATES

41

(3.3.10)

The shear lag factor was calculated following AISC Design Specifications Table D3.1 in

Equation (3.3.11). The splice length is and the distance between the gusset plate face and the

center of gravity of half the splice connection, , is shown in Figure 3.3.7. This distance accounted

for the unsymmetrical connection because of the shim plates used to induce an initial eccentricity in

the brace to control the direction of buckling.

(3.3.11)

Figure 3.3.7: Wide-Flange Brace to Gusset Section-Cut

The length of the web plates was determined based on the length of weld required to develop the

plastic capacity of the plate. This is calculated using Equation (3.3.12) with equal to the number

of longitudinal welds and is the length of weld connecting the web plate to the brace web or

gusset plate, and is the cross sectional area of the web plate. The actual length of the plate was

two times plus one inch clear between the gusset plate and brace weld.

(3.3.12)

38"x 31

2"x 14" WEBPLATE EACH SIDE

CENTERLINEOF BRACE

c.g.

x

BRACEFLANGE

CENTERLINE OFGUSSET PLATE

38"GUSSETPLATE 1

4"SHIM PLATE

42

Chapter 4: Test Setup

4.1 Introduction

This chapter summarizes the experimental test setup used to test the nine specimens within this

test program at the University of Washington Structures Laboratory. The test setup was

designed and constructed for the first set of one story, one bay SCBF tests in 2005, and reused

for this test program (Johnson 2005). This chapter outlines the general concept of the test setup

(Section 4.2), the major components including modifications made to improve the experimental

boundary conditions (Section 4.3) and the instrumentation used to capture the response during

the tests (Section 4.4). Description of the two drift histories is also included (Section 4.5).

4.2 Test Configuration Overview

Accurately matching actual field conditions in the laboratory can be difficult and expensive when

testing full scale structural steel experiments. The objective of the test setup is to successfully

test the single story, single bay system subjected to pseudo-static loading with the focus on the

diagonal brace and gusset plate connections. This testing configuration, which tests the

specimens parallel to the floor, was designed to simulate as closely as possible the boundary

conditions and degrees of freedom of an actual SCBF for gusset plate connections, while taking

into account the space, construction, and financial limitations of the tests. Figure 4.2.1 shows a

plan view of the test setup with labels for the major test components. Dimensions for the layout

of each component are shown in Figure 4.2.2. Photos of the actual test setup are shown in

Figure 4.2.3.

43

Figure 4.2.1: Test Setup and Components

Threaded Rods tostress (G) to (A)

(A) Strong Wall(B) Strong Floor(C) Reaction Block(D) Out-of-Plane Supports(E) Actuator and Swivel(F) Load Beam(G) Channel Assembly(H) Gravity Load System(I) Test Specimen

(H)

(C)

(E)

Swivel

Base ShearConnection

Load Beam

(B)

(D)SW

(D)West

(D)LoadBeam

(D)North

(D)NE

(D)East

(D)SE

Threaded Rodsstress (E) to (C)

(A)

(I)

(G)

44

Figure 4.2.2: Test Setup Layout (Johnson 2005)

45

Data Acquisition Center

Gravity Load System

Channel Assembly

Out-of-Plane Supports

Strong Wall

Strong Floor

Load Beam

Actuator

Out-of-Plane Supports

Figure 4.2.3: Actual Test Setup

46

The horizontal and vertical load paths that result during testing can be described using Figure

4.2.1 to reference the components within the test setup. Horizontal load is applied to the

northeast corner of the specimen through a single hydraulic actuator and load beam, labeled (E)

and (F), respectively. The actuator is supported by the reaction block, (C), which is tensioned to

the strong floor, (B), to resist the applied load delivered to the system. The applied horizontal

force, or story shear, exits the specimen through the bolted shear connection between the south

beam and the channel assembly, (G). The channel assembly is connected and tensioned to the

strong wall, (A). Vertical load, representative of gravity loads from the theoretical structure

above, is applied to the columns by the gravity load system, (H), through high-strength tension

rods between the north ends of the columns and the channel assembly. Vertical overturning

forces at the column bases, compression and tension, are resisted by direct bearing between the

base of the columns and the channel assembly, and tension in the gravity load system.

The bolted connections between the north and south beams of the specimen and the test setup

provide the boundary conditions for the system. These are intended to be representative of an

actual SCBF, including the composite slab which transfers lateral load to the system and restrain

the system in-plane, however, including a section of composite slab to the test specimen was

determined too costly and the specimen were constructed and tested without a composite slab.

The focus of this research is on the brace and gusset plate behaviors and on the framing

members immediately adjacent to the gusset plates. It is thought that this test setup was realistic

enough to provide accurate results for SCBF systems in the field and meets our testing

objectives. The test boundary conditions and the elimination of the composite slab can create

behaviors that in some cases did not reflect actual field behavior, and therefore must be

recognized as unique to this test setup.

The test specimen is restrained in-plane using a system of out-of-plane supports, (D) in Figure

4.2.1, which are tied to the strong floor, (B), using threaded rod. The supports restrict out-of-

plane movement of the beams and columns while allowing freedom for the entire frame to move

in-plane. The brace and gusset plates are not restrained and free to displace out-of-plane as

expected in compression.

The quasi-static loading protocol was input into the system through the actuator controller in the

form of input displacement, rate of loading, and type of loading. Instrumentation was placed

over the specimen to capture the global response as well as local responses of the components.

Strain gauges were used to continuously record the strains at strategic locations on the beams,

columns, and brace sections. The actuator and instrumentation data was recorded by the data

47

acquisition system throughout the test. Visual and verbal documentation of observations were

also used to record the response of the specimen.

4.3 Test Setup Components

4.3.1 Strong Floor and Strong Wall

The strong floor and strong wall, also referred to as a reaction wall, are permanent components

of the structures lab and shown in Error! Reference source not found. as (A) and (B). A

easonable assumption in the test setup is that the strong floor and strong wall provide rigid

support for attaching components and do not move during testing. The strong floor is a 30 inch

thick prestressed concrete floor system with embedded threaded tie down anchors spaced in

each direction at 36 inches. The L shaped strong wall stands 13’-6” above the strong floor and

also consists of 30” thick prestressed concrete. The wall provides conduits spaced at 18 in. on

center for through bolting and stressing of horizontally mounted test component.

4.3.2 Reaction Block and Actuator

The MTS hydraulic actuator used to apply the load to the specimens and the Reaction Block are

shown in Figure 4.3.1 and Figure 4.3.2. Displacement of the actuator is controlled using the

MTS controller. The actuator has a stroke range of ±10” and a maximum capacity of 470k

pushing and 330k pulling at 3000 psi hydraulic pressure. The specimens were constructed to

have the actuator pushing while the brace was in tension so that the reduced capacity coincided

with the smaller brace capacity in compression.

2” ø Anchor Rods

Reaction Block

1-1/8” ø Threaded Rods

Hydraulic Hoses

Load Cell

Spiral Washers

Actuator

Swivel

Figure 4.3.1: Actuator and Reaction Block Photo

48

Figure 4.3.2: Actuator and Reaction Block Assembly

The base of the actuator was mounted to a 6’x8’x4’ concrete reaction block and tensioned to the

strong floor using six 2” diameter Williams rods stressed to 220k each. Hydro-Stone was used

to provide uniform bearing between the surfaces at the bottom of the block and the strong floor

and the top of the block and the tension rod washer plates. The actuator and the block were

connected by six 1-1/8” diameter B-7 threaded rods stressed for a total of 360k. A 4 in. thick

adapter plate and a 2 in. thick elastomeric bearing pad were compressed between the two. The

“load train” between the reaction block and the load beam was as follows: actuator, load cell,

spiral washers, swivel, and load beam. The load cell was used to determine the load in the

actuator and has been periodically calibrated over the course of this test sequence for accuracy.

The swivel at the end of the actuator acts to allow rotational freedom between the actuator and

load beam assuring only horizontal shear was applied to the specimen. An additional ¾ in. plate

was added between the swivel head and load beam for HSS24 thru HSS26 to position the start

point of the actuator closer to the “zero point” of the stroke length, giving full range in both

directions.

The Load Cell and the LVDT within the actuator provided a continuous loop of information to

the MTS controller during operation. The actuator was set to “displacement control” and values

for the induced LVDT displacement and the rate of loading were manually input for each

movement of the actuator. More information on the loading protocol is included in Section 4.5.

118"Ø B-7 Threaded

Rods (6 total)

2"Ø Williams Anchor

Rods (6 total)

2"x 1'-2"x 1'-2"thick ElastometricBearing Pad

4" Adaptor Plate

Load Cell

Spiral Washers

Swivel Head andSide Plate

Load Beam

Actuator34"x 1'-0"x 1'-95

8"Spacer Plate

Reaction Block

49

4.3.3 Load Transfer Beam

A built-up W21x62 beam section was used to transfer the load imposed by the actuator to the

specimen and is shown in Figure 4.3.3. The beam was designed to minimize local yielding in the

specimen and transfer the lateral load over 40% of the beam length. The beam was mounted to

the actuator swivel head and bolted to the flange of the north beam with ten 1 in. diameter A490

bolts. The swivel between the actuator and load beam allows for rotation in three directions and

acted to apply the actuator force as a horizontal load to the specimen in bolt shear. The

eccentric location of the load beam resulted in a significant torque on the end of the load beam

unique to the test setup which does not reflect true field conditions. A web doubler welded to

the north beam at this location to reduce local deformation and web buckling was included in

the design of the test specimens.

Figure 4.3.3: Load Beam

4.3.4 Channel Assembly for Reactions

The channel assembly was design and fabricated by Shawn Johnson for the first SCBF test

program (Johnson 2005). The cross section detail through the assembly is shown in Figure 4.3.4

and the plan view of the corner and the shear connection are shown in Figure 4.3.5.It consisted

of two C15x50 channel sections built up with plate steel. The assembly was mounted and post-

tensioned to the strong wall and provided shear transfer for the south beam of the specimen and

distributed overturning column forces to the strong wall.

Actuator Swivel Head

1"Ø Hex-head

Bolts (4 total)

Test Specimen

10-1"Ø A490 Bolts at

Shear Connection

12" TriangularStiff Plates

2" End Plate

34" Spacer Plate

2" Stiffener Plates(Top and Bottom)

12 " Stiffener Plates(Top and Bottom)

W21x62

Load Beam

12" Flange Plate

38"x 1'-2"x 1'-111

2" Doubler

Plate (Bottom Only)

5'-234"

712"

8"

11"

5"

1'-

3"

1'-

95 8"

Cutouts forGravity LoadSystem

1078"

50

Figure 4.3.4: Channel Assembly Section Detail

Figure 4.3.5: Channel Assembly Plan Detail

The nominal shear capacity of the connection, 589.3 kip, in Figure 4.3.5 was designed to resist

the maximum horizontal force that could be applied by the actuator, 470 kip. Ten 1” diameter

A490 bolts connected the south flange of the south beam to the 1-1/2 in. thick shear transfer

plate of the channel assembly. The minimum pretension force of 64 kip per bolts was applied,

although base shear was transferred through bolt bearing during testing.

2-C15x50

12" Bottom Plate

12" Top Plate

112" Plate at

Shear Connection

C15x50

Threaded Rods stresses

Channels to Strong Wall

Strong Wall

Strong Floor

Hydro-Stone

Strong Wall

10-1"Ø A490 Bolts at

Shear Connection

158"Ø Threaded rods stresses

Channels to Strong Wall in

both directions (5 total)

Channel Assembly atShear Connection

Channel Assemblyat Corner

Test Specimen

Tension Rod

Tie-down

Tension Rod

1"Ø Threaded rods stresses

Channels to Strong Wall in

both directions (7 total)

112" Plate at

Shear Connection

12" Bearing

Plate

Hydro-Stone

51

Five 1-3/4 in. diameter Williams Form high-strength tension rods were tensioned to 220 kip

each and seven 1 in. diameter threaded rods were tensioned to 60 kips to attach the channel

assembly to the strong wall. The surface between the assembly and the concrete wall was filled

with Hydro-Stone to assure uniform bearing and shear transfer in friction. The assembly

connected to both legs of the “L” shaped strong wall at the corner transferring the applied shear

from the specimen to the wall through bearing and tension/compression of the east wall and

friction at the south wall. A coefficient of friction of 0.2 was used for the steel to concrete

surface in the calculations for the assembly design (Johnson 2005). The overturning forces in the

columns were resisted through direct bearing and tension of the gravity system tension rods,

which were anchored into the channel assembly. More regarding the tension rods and anchors

is discussed under Section 4.3.6 to follow.

4.3.4.1 Channel Assembly Modifications

Part of the modifications to the test setup in December 2007 after specimens HSS23 and before

HSS24 was the replacement of the bearing plates at the location where the specimen column

bases contact the channel assembly. The ½ in. bearing plates that were part of the original

configuration showed considerable local deformation as a result of repeat testing and were

replaced with new ½ in. bearing plates. Deformation was also noticed in the front plate of the

channel assembly at the 1-3/4 in. diameter tension rod locations. Additional ½ in. plates were

added to better distribute the force of the tension rods. Figure 4.3.6 shows the modifications to

the channel assembly front face.

52

Figure 4.3.6: Channel Assembly Modifications

4.3.5 Out-of-Plane Supports

Out-of-plane (OOP) supports were used to restrain the specimen from substantial overall

movement while permitting lateral movement. Photos of the out-of-plane support

configurations for the north beam and west column are shown in Figure 4.3.7. The out-of-plane

support at the load beam is shown in Figure 4.3.8. The locations of the out-of-plane supports

can be seen in Figure 4.2.1. Each OOP support consisted of wide-flange beam sections secured

to the strong floor with 1 in. diameter rods. These assemblies were also used to prop up the

specimen to the testing height during installation. Threaded rods were used to connect the

bottom support to the top wide-flange section. Metal shims were placed under the bottom

section to level the support and top and bottom bolts holding the top sections in place were only

snug tightened.

12" Bearing Plate at

East Column

12" Bearing Plate at

West Column

New 12" Bearing Plate

at East Column

New 12" Bearing Plate

at West Column

Original Face of Support Channel

Modified Face of Support Channel

12" Plate atTension Rod

12" Plates atTension Rods

12" Face Plate of

Channel Assembly

Location ofSpecimen Column

Location ofSpecimen Column

1

1

1

Plan View ofTest Setup

Holes forGravity LoadSystem Rods

Holes for Thru Rodsto Strong Wall

53

The frictionless contact surfaces between the specimen and the supports consisted of nylon

bumpers coated with silicone lubricant that were notched and placed over the flange edges of the

columns (top only), north beam and load beam at the support locations. Polished stainless steel

plates were placed between the wide-flange support sections and the bumpers. The lubricant

was also applied to the stainless steel plates to cover the expected full range of movement during

the test. A slight variation to this method was used at the underside of each column. Flat pads

of nylon were placed on the bottom out-of-plane support and coated with the silicone lubricant.

Stainless steel plates with edges bent upward were placed between the column section and the

nylon pads and would moved with the columns in-plane.

North Beam West Column

Nylon Bumpers

Stainless

Steel Plates

Stainless

Steel Plates

Nylon Pad

Figure 4.3.7: North Beam and West Column Out-of-Plane Supports

Specimen Specimen

54

Figure 4.3.8: Load Beam Out-of-Plane Support

Specimens HSS25 and HSS26 utilized a heavier W16x89 beam section in the frame as one of the

test parameters. The dimensions of the larger beam section required a modification to the north

beam out-of-plane support. The modified configuration of the north beam support is shown in

Figure 4.3.9. The bottom wide-flange support was rotated 90o and propped with W4 sections to

reach the appropriate height. The threaded rods anchoring the support to the strong floor were

attached through the web, rather than the flanges of the bottom support.

Figure 4.3.9: Modified North Beam OOP Support

Load Beam

Specimen

Load Beam

55

4.3.6 Axial Load System

Column axial loads were applied to simulate the gravity load from the theoretical structure above

and to resist the overturning forces. Details of the gravity load system with labeled components

are shown in Figure 4.3.10. Similar to the previous test series, a force of 350 kip was applied to

each column for specimens HSS18 through HSS23 through two 1-3/8 in. diameter (150ksi)

Williams Form rods tensioned symmetrically on each side of the web. Heavier rods were used

for specimens HSS24 through HSS26, and a force of 450 kip was applied to each column using

two 1-3/4 in. diameter rods. The larger rods were used to reduce uplift of column from the

channel assembly.

Figure 4.3.10: Axial Load System Detail

The column axial load was applied by tensioning the Williams Form rods using a short-stroke

hydraulic ram to the appropriate pre-tensioned force. 4 inch thick cap plates distributed the load

to the load beam at the north of the west column and directly to the west column. The rods

passed along either side of the columns and were anchored into the channel assembly by passing

through HSS4x4x1/2 tubes welded to the channel webs. The ends of the rods are fastened with

Williams Form spherical hex nuts and dish plates. The surface between the hex nut dome and

dish plates was coated with machine greased to allow rotation as the rods moves with the frame.

Increasing the pre-tensioned column force 350 kip to 450 kip was a primary modification of the

test setup in December of 2007. The rod diameter was increased from 1-3/8 in. to 1-3/4 in. and

the design force increased from 175 kips per rod to approximately 225 kips. In addition, the

modifications to the channel assembly were also required. The lengths of the HSS4x4 tubes

were shortened from 9-3/8 in. to 8-3/8 in. to allow for larger spherical hex nuts and dish plates.

138" or 13

4" Ø Williams Form

Tension Rods (top and bottom)

Test Specimen

Channel Assembly

Load Beam

4" thick Cap Plate

Spherical Hex

Nut with 1"

thick Dish Plate

Spherical Hex

Nut with 1"

thick Dish

Plate

HSS4x4x12 (top

and bottom)

West Column

Strong Wall

56

¾ in. thick plates were added to each side of the tubes and welded to the channel webs to

increase the capacity of the connection. A photo of the original tension rod tie down

configuration for specimens HSS18 through HSS23 and of the modified connection used for

HSS24 through HSS26 are shown in Figure 4.3.11.

4.3.7 Data Acquisition System and Storage

The data acquisition system was controlled through a personal computer using the Windows

based software LabVIEW verson 7.1. Calibration factors were assigned to each channel to

convert the measured changes in voltage to physical quantities of strain, rotation, displacement,

or force. The available data channels were organized and labeled within the program and stored

for each test. Readings were taken and recorded at each second and stored in a tab delimited

data file. This output file was then processed and analyzed with MatLab and Microsoft Excel.

National Instrument hardware was used for the data acquisition system and consisted of SCXI

1001 Chassis, SCXI 1100 and SCXI 1300 modules for potentiometers, and SCXI 1121 and SCXI

1321 modules for strain gauges. The constant 10 volts required for the instrumentation was

provided by a Hewlett Packard E3611A DC Power supply.

4.4 Instrumentation

The instrumentation used to measure the global and local responses of the system are described

in the following sub sections. The general instrumentation scheme was developed for the

HSS4x4x1/2

Spherical Hex Nut and Dish Plate

1-3/8” Williams Rod

Spherical Hex

Nut and Dish

Plate

Add’l ¾” Plates

1-3/4” Williams Rod

Figure 4.3.11: Tension Rod Tie Downs

Modified Original

57

previous three test series and used for this series. Some adjustments and additions were made to

capture specific behavior based on test parameter and to improve the overall scheme.

The types of instrumentation used are strain gauges, potentiometers, and visual observations.

Visual observations refer to whitewashing the specimen, still photographs, continuous video

taken from atop the strong wall, and the detail description of where and when specific events

occurred during the test.

4.4.1 Strain Gauges

The strains in the brace and framing elements were recorded at locations shown in Figure 4.4.1.

The strain gauges used for this test series were the Tokyo Sokki Kenkyuho Co. Ltd. FLA-6-11-

5L model. These are a uni-axial stain gauge having a 6 mm gauge length and nominal gauge

factor of 2.12. Their accuracy is reliable for strain ranges of ± 5%. The stress-strain curves were

obtain through coupon tests to determine that actual material properties of the members. Using

these relations, stresses, axial forces and moments can then be calculated at the location of

interest. Strain gauges were placed to measure:

Brace axial load (strain gauges 13, 14, 15 and 16)

Brace in-plane moment at location (strain gauges 14 and 16)

Brace out-of-plane moment at location (strain gauges 13 and 15)

Beam moments at location (strain gauges 5, 6, 11 and 12)

Column moments at location (strain gauges 1, 2, 3, 4, 7, 8, 9, and 10)

Column shears (strain gauges 1, 2, 3, 4, 7, 8, 9, and 10)

58

Figure 4.4.1: Typical Strain Gauge Layout

The procedure for attaching the strain gauges was as follows. Each location was marked and a

belt sander was used to expose clean steel with no rust or mill scale present. The area is then

sanded by hand with a fine grit paper until smooth. Acid and base was applied alternately until

each wipe clean of any dirt or residue. The strain gauge was set on strain gauge tape and placed

into the desired location. Once the gauge cross hairs were in-line with the local axes of the steel

section, Tokyo Sokki Kenkyuho Co. Ltd. strain gauge glue was applied to the underside of the

gauge head. The gauge was bonded to the steel was sustained pressure for 3 minutes. The two

wires were directly connected or soldered to the appropriate channel in the modules. The actual

initial resistance of each gauge was recorded and input into LabView. A factor was also input

into LabView to bridge the initial voltage which reduced the amplitude of error to 10-6,

essentially cleaning the sinusoidal “chatter” typical of strain gauge readings.

W16x45

W1

2x

72

W1

2x

72

W16x45

9

7

1

3

2

N

10

12

11

5

6

14

16

13 (top)15 (bot)

E

S

W

LOAD BEAM

4

CHANNEL ASSEMBLY

x5

x7

x6

x4

x3

x2

x1

2'-4"

3' 3-1/2"

1'-10"

5' 2-7/8"

5' 2-7/8"

1'-1"

1'-4"

Dimension

= Strain Gauge

59

4.4.2 Potentiometers

Potentiometers monitored both the global and local movements of the specimen and the test

setup by measuring linear displacement between the point of support, or base, and the point of

contact to the specimen. In some cases, potentiometers measured the relative displacements

between two points on the specimen, such as the movement/rotation between the beam web

and column flange at the shear tab connection. For other applications, pots measured

movement of the specimen from a stationary point, such as brace out-of-plane displacement at

the midpoint which was measured from an instrumentation block located on the strong floor.

The values obtained from the potentiometers in some cases required corrections to convert to

the potentiometer readings to actual displacements for plots and figures. The methods and

equations used to extract the values are summarized in Appendix D. The general layout of

potentiometers is illustrated in Figure 4.4.2 with designations to the type of pots at each

location.

Figure 4.4.2: Potentiometer Layout

Potentiometers were placed to capture the following responses from the test specimen:

60

Horizontal displacement of the frame (#36)

Diagonal displacement of the frame from the NE to the SW work points (#42)

Brace elongation and shortening (#41)

Brace out-of-plane displacement at mid-point (#s 53 and 54)

Buckled Brace Shape (#s 32, 1, 15, 9, 53, 54, 5, and 10)

Gusset Plate out-of-plane rotation (#s 35, 6, 0, and 24)

Rotation at plastic hinge region of beams and columns adjacent to gusset plates (#s 12,

14, 23, 26, 2, 11, 19 and 20)

Rotations at beam to column shear tab connections (#s 27, 28, 44 and 45)

It was also necessary to monitor the test setup for movement to more accurately determine the

response of the specimen. The following is a list of movements monitored.

Uplift and slip between the channel assembly and the strong wall (#s 30, 43 and 47

Uplift and slip between the columns and the channel assembly (#s 40, 29, 49 and 3)

Horizontal slip at the south beam shear connection (#25)

Horizontal slip at the load beam connection (#22)

Vertical movement of the specimen at the corner work points (#s 7, 13, 21 and 33)

Movement between the actuator reaction block and the strong floor (#s 34, 51 and 52)

Three types of potentiometers were used for this test series. UniMeasure model P510 string

potentiometers were used to measure displacement larger than approximately 3 in. Because the

out-of-plane movement of the northeast gusset plate was measured relative to the overhead

crane, string potentiometers were attached to the gusset plate with piano wire and screws were

tapped into the gusset plate steel. This setup allowed a large distance between the support and

the contact point on the specimen permitting (because of small angles) the in-plane movement

of the gusset plate to be neglected. Figure 4.4.3 demonstrates the method to attach the string

pots and wires to the crane beam overhead.

61

Two types of BEI Duncan linear conductive potentiometers were used. The 9600 Series was

used for small displacements, ± ¼” to 2”. The BEI Duncan 600 Series were used to measure

medium range displacements, ± 2” to 5”. Stands were built from available material as required

to support the 600 Series pots. Figure 4.4.4 shows the instrumentation for measuring the

southwest gusset plate out-of-plane displacement with 600 Series pots.

Figure 4.4.4: BEI Duncan Potentiometers at SW Gusset Plate

Additional pots or modifications to the previous layout were necessary in order to capture

specific behavior depending on the changing specimen design and connection configurations.

All of the potentiometer setups with dimensions for each test specimen are included in

Appendix D.

Figure 4.4.3: String Potentiometers at NE Gusset Plate

UniMeasure

P510 String Pots

Crane Beam

Wires to

Tapped Screws

BEI Duncan

600 Series

Stand

62

4.4.3 Visual Observations

Visual observations and photographs were a valuable way to document and record damage as it

occurred during the tests. A whitewash mixture was applied to the wide-flange framing elements

over the mill-scale. When yielding occurs in the steel, flaking of the mill-scale and of the

whitewash made the patterns and location of yielding clearly visible. The whitewash also acted as

an effective contrast to the darker yield lines in photographs. The simple mixture was 3 parts

water to 1 part standard Plaster-of-Paris powder and applied with a paint brush. The cold

formed HSS brace section are not prepared with mill-scale so whitewashing would not have been

useful for showing yielding of the brace. The video camera placed on the top of the strong wall

focused on the overall response of the system. The camera was turned on and record

continuously once buckling of the brace could be seen.

A tablet was photographed before each change in the test protocol noting the cycle and the

induced drift that was used to organize the photos after the test. Observers walked around the

specimen during the test looking for initial yielding, changes in behavior, and noting progression

of damage as the drift ranges increased. A recorder would write the observations as they were

called out into a file that referenced when in the test protocol the event occurred.

4.5 Loading Protocol

The symmetric cyclic loading protocol with increasing amplitude used for Specimen HSS18

through HSS25 was adopted from the previous test series. The protocol was developed based

on recommendations from the ATC-24 Protocol and the SAC Steel Project (Johnson 2005). A

plot of the protocol is shown in Figure 4.5.1. The key parameter, θy, is based on the interstory

drift angle that corresponds to the onset of yielding or buckling of the brace section. The value

for θy was originally based on the results from an idealized computer model and verified through

the experimental results from earlier test specimens. The value used for the loading protocol for

this series of tests relates to a horizontal frame displacement at the top of the frame of

approximately 0.625” (Kotulka 2007).

63

Figure 4.5.1: Original Loading Protocol (Johnson 2005)

Considerable differences can be seen between the induced displacement and the actual story

displacement for each test. This was due to losses in the system because of uplift at the column

bases and slip in the bolts at the load beam to north beam connection and at the south beam to

shear connection in the channel assembly. The affect of the losses was most significant with the

brace in tension because of the greater stiffness in the frame and larger force in the actuator.

The actual displacement history for each specimen is included in Chapter 5.

The actuator was pushed to the maximum displacement of the cycle and held at “peak” while

damage was photographed and observation taken. The specimen was then pulled back through

zero to the maximum negative displacement and held at “valley”. After time for observations

and photographs, the specimen was returned to zero and the appropriate adjustment to the MTS

controller settings made before attempting the next cycle. The total rate of loading for each full

cycle was also based on the ATC recommendations and increased as the displacement increased.

For 0 ≤ θ < 1.0θy → 60 seconds cycles

For 1.0θy ≤ θ < 2.0θy → 80 seconds cycles

For 2.0θy ≤ θ < 4.0θy → 120 seconds cycles

For θ ≥ 4.0θy → 160 seconds cycles

An alternate tension dominated near-fault loading protocol was introduced and used for

specimen HSS-26. At the time of this test series, there had not been an established near-fault

loading protocol for SCBF systems. The manner for loading the specimen was taken from the

SAC recommended Near-Fault Drift History for Special Moment Frames set forth in FEMA

64

355D. The amplitude of the first large tension push was based on the maximum expected tensile

displacement of the brace from cyclically loaded SCBF tests and subsequent cycles were scaled

based on that value. The tension dominated near-fault loading protocol for HSS-26 is shown in

Figure 4.5.2.

Figure 4.5.2: Alternate Near-Fault Loading Protocol

HSS26 Near Fault Loading History

-1.0

-0.5

0.0

0.5

1.0

1.5

2.0

2.5

3.0

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16

Cycle

Dri

ft,

%

-0.787%

-1.206"

2.121%

3.251"

2.536%

3.888"

1.705%

2.614"

0.875%

1.341"

0.459%

0.704"

1.290%

1.978"

0.875%

1.341"

-0.984%

-1.508"

-0.044%

-0.068"-0.371%

-0.569"

2.536%

3.888"

-0.787%

-1.206"

Drift, %

Displacement, inches

65

Chapter 5: Experimental Results

5.1 Introduction

The experimental results for HSS-18 through HSS-26 are presented in this chapter. Detail

descriptions of the events of each test are provided including photographs documenting

component behavior and damage as it occurred. Section 5.2 describes potential yield

mechanisms and failure modes of SCBFs. This also includes descriptions of the method and

nomenclature used for the initiation and progression of damage as it occurred during the tests.

Section 5.3 outlines the components of the test and the approach to describe where on the

specimen damage occurred. A section is provided for each specimen describing in detail the

events during the test. The specimens are described and an explanation is provided for testing

each specific design and test parameters. The behavior of the specimen during the tests is

described in terms of yield mechanisms and failure modes. These behaviors are related to the

performance of the specimen and described at three frame drift ranges: intermediate, moderate,

and severe. Damage is noted at maximum and minimum drift ratios and applied horizontal force

from the actuator. A summary reiterating important observations are then included at the end of

each section.

Methods used to improve the overall performance have repeatedly been supported through the

three previous test programs at UW since the original base line test following AISC design

procedures, HSS-01. A number of the test specimens from this series were chosen based on

discussions with consultants from AISC to better improve the constructability of SCBFs in the

field while maintaining improved performance. A table of general information for all the

specimens tested from HSS-01 to HSS-26 is shown in Table 5.0.1 with study parameters

highlighted.

5.2 Yield Mechanisms and Failure Modes

This section outlines the yield mechanisms and failure modes observed during the tests. Figure

5.2.1 shows the locations of potential yield mechanisms and system failure modes. Damage of

specimen components is documented throughout the test. The term damage can be deceiving

when describing specimen behavior because some types of damage is desired and expected. The

balance design approach discussed in Chapter 4 depends on the desirable yield mechanisms of

specific components in order to improve the overall system performance.

66

Figure 5.2.1: Yield Mechanisms and Failures Modes for Test Specimen

Table 5.1.1 outlines the nomenclature used in the results section of each test to describe the type

and level of damage observed. The table defines the damage levels used in this research study.

However, the severity suggested by individual damage levels does not necessarily coincide with

actual performance design damage states. For example, moderate or severe yielding of a framing

element would not necessarily be considered moderate or severe system damage, since it will

often not reduce strength or limit the system’s ability to perform in future events. Instead, many

of these damage states are relative indicators of progress of damage and progression of yield

deformation. Table 5.1.2 is included to relate each damage state to their potential consequences

related to performance levels: Immediate Occupancy (IO), Life Safety (LS) and Collapse

Prevention (CP). The shading used to denote the severity of damage is used in the tables

summarizing the damage states as they occurred during each test. Darker shading indicates more

severe damage.

67

Table 5.2.1: Component Damage Nomenclature

Table 5.2.2: Damage State Shading for Performance Based Design

5.2.1 Plate and Frame Yielding

Yielding of plates and of beams and columns is referenced at three performance levels:

initial/mild, moderate, and significant/severe. Determining the level of yielding of components

Brace Gusset Plates Beams/Columns Performance Based Description

Minor

Moderate

Severe

Failure

Fracture of component. Major loss of strength and

potential for system failure or collaspe. Beyond LS

and possibly CP.

Y5

BF; BS; PF;

NSFWF; WFB BS

Y5; BS; BC;

PC; NSD

Y5; SWD;

SWDB

SCBF Damage State Shading for Performance Based Design

Superficial damage not effecting stength of system.

Repairs not required for IO.

Some obvious yielding or residual deformation.

Minimal loss of strength. Repairs possibly required

for IO perception.

Severe visual yielding and residual deformations.

Some loss of strength but okay for LS and CP.

Component would have to be replacd for IO.

Y3; B2; NSDY3; B2; B3;

WD; WDB

Y1; B1;

BSLPY1; B1 Y1; B1

Y3; B2

68

was done by visual inspection during the test. Whitewash is applied over the mill scale of hot-

rolled steel sections and plates to flake as the steel yields. This visually enhances the yielding for

observations and photographs.

Gusset plate yielding is an important performance characteristic of gusset plate connections.

The gusset plates may yield when the brace is in tension as well as in compression due to the out-

of-plane rotation from the brace buckling. The yield lines vary depending on the type and

direction of the stresses applied to the plate. Concentrated yielding can occur at highly stressed

areas at the brace ends or at reentrant corners. Figure 5.1.2 shows initial or mild yielding of the

NE gusset plate at the brace end for HSS-24. This level of yielding is referred to as Y1 in the

individual test sections. Moderate yielding is defined when the yield lines extend beyond half of

the depth or width of the member being observed. Figure 5.1.3 is an example of moderate

yielding (Y3) covering approximately half of the NE gusset plate area during HSS-24.

Figure 5.2.2: Initial/Mild Yielding of the Gusset Plate (Y1)

69

Figure 5.2.3: Moderate Yielding of the Gusset Plate (Y3)

The Y5 gusset plate yielding damage state is described as significant or severe yielding covering

most of the section or surface being observed. Figure 5.2.4 shows significant or severe yielding

of the NE gusset plate during HSS-24.

Figure 5.2.4: Significant/Severe Yielding of the Gusset Plate (Y5)

Yielding of the frame elements is also referenced by this same damage scale. For wide-flange

sections, the damage might occur on the web but not necessarily the flanges. As a result, the

location of yielding is also noted for wide-flange sections. Yielding of the HSS brace was more

difficult to visually observe during testing because these sections do not have mill scale. As a

result, measured brace elongation can be used to determine when tensile yielding occurs as

illustrated in the equation below. Yielding of the brace was defined as occurring when the strain

70

reaches 0.2%. The original length of brace, , is taken as the brace length from

clear edge of gusset to clear edge of gusset.

5.2.2 Brace Buckling and Brace Damage

Figure 5.1.5 shows the progression of brace buckling during the test HSS-24. Three distinct

damage states are defined for classifying the brace out-of-plane movement. Figure 5.2.5a shows

the first damage state, B1, which is initial buckling of the brace. This is achieved when the

vertical out-of-plane movement reaches 2% of the brace length, Lb. In this case the entire brace

length, , is defined as the actual length of the wide-flange or HSS tube brace. Figure 5.2.5b

shows the B2 damage state and it is determined when the vertical out-of-plane displacement

exceeds the member depth. The brace continues to buckle until a plastic hinge and localization

of damage formed at the center as shown in Figure 5.2.5e and this behavior characterizes damage

state BC. Typically cupping of the underside and bulging of the two adjacent sides of HSS

tubular braces is observed as seen in Figure 5.2.5f.

71

Fracture of the brace hinge is the anticipated failure mode of the system. Typically micro-cracks

begin to develop on the underside of the brace and shortly there after, the brace begins to tear in

tension. The designation BF for brace fracture is assigned to the point when the brace begins to

tear and the load resisting capability of the system is lost. Figure 5.2.6a from HSS-18 is an

example of micro-cracks and a larger partial tear through the center of the brace. Figure 5.2.6b

a) b)

c) d)

e) f)

Figure 5.2.5: Progression of Brace Buckling

72

shows a fractured brace from HSS-22.

Another potential failure mode of the system is net section fracture of the brace at the reduced

section located at the bored hole at the end of the gusset plate slot. AISC design specifications

required net section reinforcing by design. The designation used for cracking or tearing of the

brace material at the net section in Table 5.1.1 is NSD. The designation for fracture of the brace

at the net section is NSF.

5.2.3 Local Buckling

Initialization and progression of local buckling of components was visually observed during the

tests and correlated to frame drift and various damage states. Similar designations are given for

local bucking of plates, beams and columns. B1 is defined when initial buckling deformation

becomes visible. Figure 5.2.7a is an example of B1 flange buckling of the column at flange at the

beam column connection during test HSS-25. Damage state B2 describes moderate local

buckling and is again achieved when the local deformation exceeds the thickness of the buckled

element. Figure 5.2.7b shows B2 flange buckling of the HSS-25 column adjacent to the gusset

plate. Damage state B3 occurs when local buckling progresses to a point that local deformation

simulates cupping and bulging of the plastic hinge of the brace (designated as BC) and illustrated

in Figure 5.2.5f.

a) BF Brace Fracture b) Total Brace Fracture

Figure 5.2.6: Brace Damage at Plastic Hinge

73

5.2.4 Weld Tearing and Fracture

The designations used for describing crack initiation, crack propagation and fracture depend on

where the crack is observed. Mild ductile tearing of gusset plate welds progresses after crack

initiation and can have a favorable effect in achieving greater drifts while maintaining total frame

resistance. As cracks continue to propagate, the risk for complete weld fracture increases as the

controlling failure mechanism of the system.

Crack initiation is designated at WD, noting the first observable weld damage. The WD damage

state is used for weld damage up 5% of the total weld length including both the beam and

column interface weld lengths. Crack lengths greater than 5% and less than 10% of the total

weld length is referred to as MWD, moderate weld damage. Crack lengths greater than 10%

designate severe weld damage, SWD. The undesirable failure mechanism of weld fracture is

referred to as WF. Weld crack initiation can be seen directly in the weld material or in the heat

affected zone (HAZ) of the base metal. Weld damage located in the based material is indicated

with a “B” at the end of the damage state designation. Figure 5.2.8a is an example of the

MWDB damage state at the NE interface weld of WF-23. Figure 5.2.8b shows SWD weld

damage that initiated in the base material and propagated into the weld material extending nearly

the full length of the gusset-to-beam weld.

b) B2 initial buckling a) B1 initial buckling

Figure 5.2.7: Local Buckling of HSS-25 Columns

74

5.2.5 Bolted Connection Slip

When bolted connections were designed to transfer shear forces, slip critical bolts were used.

The action of the shear forces overcoming the bolt resistance and slipping is referred to as BSLP.

The bolts would then act in bearing to transfer the load. Bolt fracture at shear tab connections

or anywhere else on the specimen is referred to as BS. Specimen HSS-19 implemented a bolted

brace to gusset connection.

5.3 Specimen and Test Result Nomenclature

Each section summarizing the test result presents will follow a similar format. Plots for drift

history and the hysteretic behavior are provided. The actual frame drift history varies from the

induced drift history due to losses from column uplift, and bolt slip between the north beam and

the actuator load beam. A table is provided relating frame drift ratio, horizontal frame force, and

the performance of each specimen. The performance of the specimen is explained using the

damage states and severity shading as described in the previous section. Columns within the

tables are sectioned for brace, gusset plates, and frame behavior. Abbreviated designations for

damage states are added denoting the level of damage and where and when it occurred. This

allows the reader to review the table and quickly gain an impression of test specimen’s

performance.

Abbreviated nomenclature is used to minimize the amount of writing necessary to describe what

and where damage was observed during the tests. Figure 5.3.1 outlines the letter designations

pertaining to different components of the typical specimen. The column and beam designations

a) Initial Cracking in Base Material (WDB) a) Severe Weld Cracking (SWD)

Figure 5.2.8: Examples of Weld Damage from WF-23

75

can refer to only the web for flange of the section by and “W” or “F”, such as SWBF refers to

the south west beam flange. The flanges of the framing elements are referred to as inner and

outer. Inner flanges are toward the inside of the frame while outside flanges are to the outside.

The manner used to refer to specific areas of the wide-flange frame elements related to the test

set-up is described in Figure 5.3.2.

Figure 5.3.1: Specimen Component Notation (Johnson 2005)

Figure 5.3.2: Wide-Flange Area Designations related to Test Set-Up (Johnson 2005)

Specific verbiage is use for consistency when discussing welds. The welds connecting the brace

ends to the gusset plates are referred to as brace splice welds and the welds connecting the gusset to

the frame elements are gusset interface welds. Complete joint penetration welds between the beam

and column flanges are referred to as beam/column flange welds.

76

Three ranges of frame drift ratio are used to evaluate the test results, initial drift range, moderate

drift range, and severe drift range (Kotulka 2007). Drift range is total between the maximum

positive and minimum negative drift. Initial drift range is between zero and 1.25%. Moderate

drift range is between 1.25% and 2.75% and severe is greater than 2.75%. Based on the

orientation of the test specimen, positive drift results in brace tension and negative drift result in

brace compression. The frame must retain resistant to consider successfully reaching a drift.

The value for maximum or minimum drift is taken up to the point of system failure. If the

system failure is not sudden and obvious, resistance is considered lost when the applied

horizontal force drops below 70% of the yield force.

5.4 HSS-18: Thin Gusset Plate, Unwelded Frame Connection

5.4.1 Specimen Overview

The necessity of a fully welded beam/column moment connections for SCBFs is an economic

deterrent and a topic of interest to engineers. The current AISC Seismic Provisions do not

require a full moment beam-to-column connection adjacent to the gusset plates, but these

connections have been used by design engineers in high seismic hazard areas of the United

States.

Specimen HSS-18 evaluated the affect unwelded flanges and a shear plate beam-to-column

connection has to the overall frame performance. Shear tab connections designed to account for

the additional vertical force component from the UFM were implemented at the beam column

connections adjacent to the gusset plates. The shear tab was ½ in. thick and connected to the

beam with 4-1 in. A490 bolts. The gusset plate and brace were identical to HSS-05. The 3/8 in.

thick gusset plate used an 8t elliptical clearance and fillet interface welds were sized to develop

the full capacity of the plate. The detail for specimen HSS-18 is shown in Figure 5.4.1

77

Figure 5.4.1: HSS-18 Connection Detail

HSS-18 was tested January 1, 2007 and the specimen ultimately failed in the 35th cycle due to

brace fracture at the plastic hinge. Maximum and minimum drift ratios of 1.60% and -2.59% for

a range of 4.19% were achieved. The maximum story shear forces resisted in the frame with the

brace in tension was 323.8 kip and -152.1 kip in compression for a range of 475.9 kip. Figure

5.4.2 is the frame drift history. Figure 5.4.3 shows the hysteretic behavior of the specimen.

Figure 5.4.2: HSS-18 Drift History

78

Figure 5.4.3: HSS-18 Hysteresis

Table 5.4.1 is a summary of results during the test including drift ratios, lateral load, and

performance of specimen components. The level of gusset plate yielding is especially notable for

this test. Y5 level damage was observed for both the NE and SW gusset plates at the higher

drifts.

Table 5.4.1: HSS-18 Peak Results

79

5.4.2 Initial Drift Range: 0% to 1.25%

Slight initial tensile yielding (Y1) of the southwest gusset plate at the brace end was observed at

0.11% drift. Yield lines diagonal to the longitudinal axis of the brace extend approximately 2 in.

from the brace end as shown in Figure 5.4.4. Similarly, Y1 yielding at the brace end of the

northeast gusset plate was observed at a drift of 0.14%. Figure 5.4.5a and Figure 5.4.5b show

initial yielding of the column reentrant corner for the southwest gusset plate and the

beam/column intersection corner of the northeast gusset plate at 0.21% drift, respectively.

Figure 5.4.4: Y1 Initial Yielding of SW Gusset (0.11%)

a) Y1 Yielding of SWG Column Corner b) Y1 Yielding of NEG Beam/Col Intersection

Figure 5.4.5: Initial Yielding (Y1) of Gusset Plate (0.21%)

80

B1 buckling of the brace out-of-plane exceeding 2% of the brace length, 3 1/8 in., was achieved at

-0.31% drift. Figure 5.4.6 shows the B1 buckled shape of the brace. Out-of-plane displacement

reached B2 level buckling, 5 in., at -0.62% drift and is shown in Figure 5.4.7.

Figure 5.4.6: B1 Brace Buckling (-0.12%)


5.4.3 Moderate Drift Range: 1.25% to 2.75%

Yielding of both gussets plates quickly increased to the Y3 damage state due to a combination of

both the tensile stresses at 0.41% drift and stresses from the out-of-plane rotation in

compression at -0.94% drift. Long slashing yield lines from the end of the brace extend to the

81

gusset edges and moderate yielding was observed between the brace splice weld and the framing

elements. Figure 5.4.8 shows the extent of the yielding for the northeast gusset plate.

Figure 5.4.8: Moderate Yielding (Y3) of the NE Gusset plate (-0.94/0.41%)

Initial yielding was observed on the top flange of the north beam at the east end of the load

beam connection. This type of yielding is commonly seen in previous test series and is a result

of the local couple created by the eccentricity of the actuator load beam to the work point of the

north beam. This unique condition of the testing load path has been discussed further in

Chapter 4.

The brace elongation exceeded 0.2% and is considered to have initial Y1 yield over the entire

length of the brace between the ends of the gusset plates at 0.73% drift. Initial yielding of the

southeast side of the brace at the net section at the gusset plate slot was observed at 0.74% drift

as shown in Figure 5.4.9. Initial yielding was also observed on the outside flange of the

southwest column at 0.74% drift.

82

Figure 5.4.9: Initial Yielding of Brace Net Section (0.74%)

5.4.4 Severe Drift Range: > 2.75%

Slight edge buckling (B1) of both the northeast and southwest gusset was observed at 1.09%

drift. Figure 5.4.10 shows this behavior at the column edge of the southwest gusset plate. Initial

yielding (Y1) of the north beam flange at the gusset reentrant corner was also seen at this drift.

Yielding of the northeast beam at the beam/column connection was observed extending out

from the bolts shear tab. This initial yielding (Y1) at 1.10% drift is shown in Figure 5.4.11.

Figure 5.4.10: Slight Edge Buckling (B1) of SW Gusset Plate (1.09%)

83

Figure 5.4.11: Initial Yielding (Y1) of NE Beam Web (1.10%)

Gusset plate yielding has continued to increase. Severe yielding (Y5) is observed at a drift of

1.42% across the northeast and southwest plates thoroughly extending from edge to edge.

Figure 5.4.12a and Figure 5.4.12b show severe yielding of the northeast and southwest gussets

respectively.

A plastic hinge has formed at the center of the brace. Cupping of the underside and bulging of

the two vertical sides of the brace were observed at -2.09% drift. Figure 5.4.13 shows the start

of the compressive brace failure (BC) at the hinge location and the buckled shape of the brace.

a) NE Gusset Plate Yielding b) SW Gusset Plate Yielding

Figure 5.4.12: Severe Yielding (Y5) of Gusset Plate (1.42%)

84

Gusset plate edge buckling has progressed to more than the thickness of the plate (B2) at 1.60%

drift. This is seen at both the beam and column edges of the northeast and southwest gussets

and shown in Figure 5.4.14. This moderate edge buckling is likely due to a combination of the

increased opening and closing rotation at the beam/column connections and inelastic

lengthening of the gussets in tension.

Figure 5.4.14: Moderate Edge Buckling (B2) of SW Gusset Plate (1.60%)

Local buckling (B1) and moderate yielding (Y3) were observed on the outside flange of the

southwest column at -2.54/1.60% drift. Figure 5.4.15 shows the concentration of damage which

is across from the gusset plate free edge and can be attributed to the high moment demand as

Figure 5.4.13: Compressive Brace Failure (BC) (-2.09%)

85

the frame opens and closes. Moderate yielding (Y3) was also observed on the north beam web at

-2.54% drift with yield lines across the entire depth of the web and highly concentrated at the

center. This is shown in Figure 5.4.16.

Figure 5.4.16: Moderate Yielding (Y3) of Northeast Beam (-2.54%)

Tiny cracks could be seen on the underside of the brace plastic hinge while in tension at 1.59%

drift. A minimum drift of -2.59% was achieved before fracture of the brace at the hinge location

during the following cycle. The brace center displaced out-of-plane a maximum value of 15.89

in. Figure 5.4.17 shows the level of local failure, cupping and bulging, at the hinge during this

a) Moderate Yielding (Y3) b) Local Buckling (B1)

Figure 5.4.15: Damage to SW Column (-2.54/1.60%)

86

final compression cycle. The brace fractured (BF) at 1.31% drift and a lateral load of 207.4 kip

applied to the frame, approximately 64% of the maximum lateral load the specimen resisted.

Figure 5.4.18a shows the brace necking at the hinge location and tearing from the bottom up.

Figure 5.4.18b is of the fracture brace.

Figure 5.4.17: Local Compressive Failure of Brace (-2.59%)

a) Brace Tearing at Failure (BF) b) Brace Fracture

Figure 5.4.18: Brace Fracture at Plastic Hinge (1.31%)

87

5.4.5 Specimen Summary

HSS-18 was designed with a thin gusset plates, 8t elliptical clearance and interface weld designed

to develop the full plastic capacity of the gusset plate, similar to HSS-05, the best performing

specimen from previous test series. The major test parameter was the elimination of the CJP

welded moment connection between the beams and columns adjacent to the gusset connection

and the usage of a bolted shear plate connection. The bolted shear plate beam-to-column placed

significantly larger demands on the gusset plates than previous tests. Figure 5.4.19a and Figure

5.4.19b show the northeast gusset plate and the southwest gusset plate at the completion of the

test, respectively. The amount of yielding is complete and widespread compared to those of

pervious tests. This increase demand on the gusset plates is also seen in the resulting edge

buckling seen in Figure 5.4.20.

a) NE Gusset Plate a) SW Gusset Plate

Figure 5.4.19: Gusset Plate Yielding at Completion of Test

88

Figure 5.4.20: SW Gusset Plate Edge Buckling at Completion of Test

There was also significantly less damage to the framing elements during this test than seen

previously. Further, no weld cracking was observed in this test. These factors may also be

attributed to the observation that by eliminating the CJP welded moment beam-to-column

connection, the specimen exhibits less frame action which would reduce the moment demands

on the beam and columns as the frame opens and closes. This minimizes inelastic behavior

within the framing elements which could increase the overall ductility of the system.

5.5 HSS-19: Bolted WT Brace Connection


Specimen HSS-19 was a bolted brace to gusset plate connection as shown in Figure 5.5.1 and

was developed based on consultation with representatives with AISC and design engineers.

HSS-16 (Kotulka 2007) used a large single plate to connect the brace to the gusset with 15 1-1/8

in. A490 bolts in order to eliminate the need for net section reinforcing and to evaluate a

connection that can be bolted in the field. The specimen achieved very large drifts, 5.89% total

drift range, but was significantly less stiff and failed at the extension plate connecting the brace to

the southwest gusset plate. The specimen failed in the connection at a resistance significantly

lower than the buckling resistance of the brace.

HSS-19 looked to improve on this performance by using two WTs to connect the brace to the

gusset plate. The WT was chosen because of its out-of-plane stiffness and connectability

through the flanges. Figure 5.5.1 shows a detail of the gusset to brace connection. The ½ in.

89

gusset plate was design with a 6.8t elliptical clearance and the interface welds were designed to

develop the full plastic capacity of the plate.

Figure 5.5.1: HSS-19 Detail

Specimen HSS-19 was tested March 21, 2007 and ultimately failed through fracture of the splice

plate connecting the WTs to the brace during the 25th cycle. The connection did not develop

buckling of the brace in compression and the frame reached only moderate drift levels. Figure

5.5.2 shows the frame drift history. The maximum and minimum drifts achieved were 0.31%

and -1.01% for a total drift range of 1.32%. The maximum and minimum lateral forces resisted

by the frame are 224.3 kip with the brace in tension and -101.9 kip in compression. Figure 5.5.3

shows the hysteretic behavior of the specimen.

90



The summary of the test results are included in Table 5.5.1. As stated before, the specimen only

achieved moderate drift ranges and fractured prematurely at the 7/8 in. splice plate between the

brace and the WTs at the southwest gusset connection. Minimal yielding of specimen

components was observed prior to failure.

91



Initial downward buckling was observed at -0.09% drift. At -0.18% drift, it became clear that a

hinge was forming at both the northeast and southwest in the 7/8 in. splice plates between brace

end and the WTs. Moderate yielding (Y3) and slight buckling (B1) of the northeast splice plate

can be seen in Figure 5.5.4. Initial yielding (Y1) was observed at the similar location on the

southwest splice plate as shown in Figure 5.5.5.

Figure 5.5.4: Moderate Yielding (Y3) of NE Splice Plate (-0.18%)

92

Figure 5.5.5: Initial Yielding (Y1) of the SW Splice Plate (-0.18%)

Moderate yielding (Y3) and initial buckling (B1) was observed at the southwest splice plate at -

0.33% drift. Severe yielding (Y5) of both the northeast and southwest splice plates is achieved

during a repeat cycle at -0.33% as shown in Figure 5.5.6a and Figure 5.5.6b, respectively.

Bolt slip occurred at the southwest brace to gusset plate connection. Potentiometers were place

to capture potential bolt slippage and shown in Figure 5.5.7. The plot shows southwest WT slip

verses cycle and a spike can be seen at cycle 17, pertaining to 0.21% drift and 177.6 kip of lateral

force resisted by the frame.

a) NE WT Splice Plate b) SW WT Splice Plate

Figure 5.5.6: Severe Yielding (Y5) (-0.33%)

93

Figure 5.5.7: Southwest WT slip (BSLP) vs. Cycle

A clear hinge has formed at the southwest brace splice plate. Moderate buckling is observed at -

0.65% drift as shown in Figure 5.5.8. As the brace kinks at the hinge, the stem of the WT is

binding on the HSS brace and causing severe local deformation of the tube. This can be seen in

Figure 5.5.9.

Figure 5.5.8: Moderate Plate Buckling (B2) (-0.65%)

94

Figure 5.5.9: Binding of WT Stem on Brace

Initial Yielding of the southwest gusset plate beam reentrant corner was observed at -0.72% drift.

This is not surprising considering the large rotational demand on the southwest gusset bending

out-of-plane as the brace hinges at the splice plate. Figure 5.5.10 shows the deformed shape of

the gusset with the brace in compression.

Figure 5.5.10: Buckled Shape of Brace (-0.72%)


Bolts slip of the northeast gusset connection occurred at a frame story drift and lateral load of

0.28% and 224.3 kip at cycle 23. A plot of the northeast WT slip vs. cycle shown in Figure

5.5.11 shows jump at cycle 23 as bolt slip occurred. The hinge at the southwest brace splice plate

95

yielded over the entire area and exhibited severe buckling (B3) at -1.01% drift. Initial yielding

was also observed on the southwest gusset at the column reentrant corner. Figure 5.5.12 shows

the splice plate immediately before and after fracture. Fracture occurred at 0.28% drift with the

brace in tension.

Figure 5.5.11: Northeast WT slip (BSLP) vs. Cycle


This specimen came well short of performance expectations for drift, strength and desired failure

mode. From a performance based design stand point, this specimen would not have achieved

collapsed prevention or life safety performance levels for a moderate seismic event. This type of

connection should not be used in a seismic region without further study.

Figure 5.5.12: Damage at Splice Plate Hinge before and after Fracture (PF) (0.28%)

96

This brace configuration relied on multiple elements and wasn’t able to buckle over the full

length. Initial the brace buckled over the full length but the splice plates acted as a weak point

resulting is early lose of compressive strength and hinging. An alternate design could maintain

out-of-plane stiffness over this transition in order to develop buckling in the brace. This could

possibly be done by shop welding a vertical fin to the brace that would then connect to the WT

stem with slip critical bolts in the field. This would also eliminate the binding damage between

the WT and tube and the plate hinges shown in Figure 5.5.13.

Figure 5.5.13: Damage from WT and Tube Binding

There was virtually no damage to the framing elements at the completion of this test. Only the

southwest gusset plates saw very minimal yielding at the reentrant corners as shown in Figure

5.5.14. These results completely contrast what this research is trying to accomplish in utilizing

the total ductility provided by tensile yield and post-buckling deformation of the brace as well as

limited yield deformation of the gusset plates and frame elements in order to increase system

performance. All of the inelastic behavior was concentrated in to one 7/8 in. plate which not

surprisingly fractured very early. Figure 5.5.15 shows the fracture splice plate at the completion

of the test.

97

Figure 5.5.14: SW Gusset Plate at End of Test

Figure 5.5.15: Fractured Splice Plate at End of Test

5.6 HSS-21: 14 Bolt Beam End Plate Connection


Contractors, erectors and fabricators of steel structures have at times expressed preference for

bolted connections rather than welded connections to speed construction and reduce labor costs.

The connection detail chosen for the next two test specimen was proposed by Tom Schafly of

AISC and Tim Fraser of CANRON to evaluate a bolted beam end plate connection and its

affect on the system performance. Specimen HSS-21 and HSS-20 both use this alternate beam

to column connection which consists of 1 in. thick plates shop welded beam web and gusset and

98

then bolted to the columns. The specimens were both fabricated by CANRON, since geometric

control of the fabrication is more critical with this connection geometry. The results of HSS-21

are discussed first followed by HSS-20 in the proceeding section. This is because the test photos

from HSS-20 were inadvertently deleted and the photos from HSS-21 will be reference when

needed for clarity. The behavior and patterns of yielding on the gusset plates and frame

elements was very similar for both tests.

Two identical sets of beams and columns were fabricated by Canron Western Constructors in

Portland, OR. Because this configuration would be professionally shop fabricated, the interface

welds connecting the gusset plate to the beam and the end plate would be welded in a controlled

environment which typically results in fewer imperfections, which is more desirable to the

designers.

The 3/8 in. gusset plates were designed with a 6.8t elliptical clearance and 7/16 in. fillet interface

welds to develop the plastic capacity of the gusset plate. The HSS5x5 brace section was

fabricated at UW and the frame was assembled on site in the lab. HSS-21 uses a 14 – 3/4 in.

A490 bolt configuration. The bolts were pretensioned to the required minimum bolt pretension

given in the AISC Specification, Table J3.1, of 35 kip each. Direct tension indicator (DTI)

washers were used to assure the correct pretensioning. Figure 5.6.1 is of HSS-21 which has

empty bolt hole at the 3rd and 5th rows from the top.


The beam web and gusset plate were welded to the end plate with fillet welds on both sides but

the beam flanges are not weld to the gusset. They are flush to the face of the end plate and

99

should develop bearing of the compression flange at the end plate during the opening and

closing of the frame as it cycles through the test. This is different from HSS-18, which used a

shear tab connection instead of a fully welded moment connection at this location but allowed

adequate room for the beam to rotate to avoid bearing on the column flange. The HSS-20 and

21 connections are expected to exhibit stiffness somewhere in between the rigidity of a fully

welded moment connection and the simple shear tab connection evaluated in HSS-18.

Additional instrumentation was added for this test to monitor rotation or prying of the end

plates from the columns.

HSS-21 was tested June 15, 2007. The specimen ultimately failed on the 35th cycle by fracturing

the plastic hinge of the brace. A total drift range of 4.14% was achieved, 1.60% with the brace in

tension and -2.55% in compression. The maximum and minimum lateral load resisted by the

frame was 347.6 kip and -163.7 kip for a total force range of 511.3 kip. The increase in drift and

loss of strength in the brace compression direction could be attributed to reduced stiffness of the

connection by using bolts. Figure 5.6.2 shows the HSS-21 frame drift history and HSS-21

hysteresis is shown in Figure 5.6.3. The peak results including observed damage are shown in

Table 5.6.1.


100




Initial yielding at the brace ends of both the southwest and northeast gusset plates was observed

at 0.07% drift and 73.2 kip resisted by the frame. Figure 5.6.4 shows the short yield lines

protruding from the end of the brace to gusset welds of the northeast connection. At 0.10%

drift, loud concussions could be heard toward the northeast and southwest of the frame and

were suspected to be slip of the beam to column connection bolts. Initial yielding of the

southwest column flange at the southern most bolt row was also observed at this drift. This is

understandable considering the highest tensile force occurs at this outside bolt row from the

101

rotational moment as the frame closes. Figure 5.6.5 shows the yielding southwest column flange

at the beam outside flange of the beam-to-column connection.

Figure 5.6.4: Initial Yielding (Y1) of NE Gusset Plate (0.07%)

Figure 5.6.5: Initial Yielding (Y1) of SW Beam/Col Connection (0.10%)

Yielding continues to increase at both northeast and southwest gusset plates. Long yield lines

running parallel to the south beam to gusset weld were observed at -0.22% drift as shown in

Figure 5.6.6. Figure 5.6.7 shows initial yielding (Y1) of the south beam web which is seen the

following cycle at 0.14% drift.

102

Figure 5.6.6: Yield Lines Parallel to South Beam Interface Weld

Figure 5.6.7: Initial Yielding (Y1) of South Beam Web (0.14%)

At 0.17% drift, initial yielding (Y1) of the north beam web was observed running parallel to the

gusset interface welds and extending approximately 5 in. as shown in Figure 5.6.8. Initial yielding

(Y1) of the northeast and southwest gusset plate reentrant corners at the column was observed at

-0.34% drift. Yielding at the southwest gusset reentrant corner adjacent to the end plate

connected to the column is shown in Figure 5.6.9. Out-of-plane displacement of the brace

center in buckling exceeded the 2% of the total brace length (B1) at a drift of -0.40%.

103

Figure 5.6.8: Initial Yielding (Y1) of North Beam Web (0.17%)

Figure 5.6.9: Initial Yielding (Y1) at SW Gusset to Column Reentrant Corner (-0.34%)

Moderate yielding (Y3) of the northeast gusset plate could be observed at 0.35% drift. The

gusset plate has exhibited yielding across most of its surface and yield lines can be seen extending

from behind the brace end to the free edges in Figure 5.6.10. At drift -0.79%, initial yielding

(Y1) at the northeast gusset plate beam reentrant corner was observed.

104

Figure 5.6.10: Y3 of the NE Gusset Plate (0.35%)

The top bolt from the middle row of the southwest connection and the bottom bolt from the

middle row of the northeast connection fractured (BS) in tension at 0.36% drift. Figure 5.6.11a

shows the location of the bolt fracture for the southwest end plate connection. The fracture

occurred in the bolt rows that most closely follow the line of the brace as it intersects the

beam/column work point. Figure 5.6.11b shows the failure plane of the bolts in the reduced

section of the threads with no shear deformation which is typical of a bolt tension failure. While

tensile bolt fracture is clearly a failure mode of the connection, this bolt fracture did not limit the

resistance and deformation capacity of the frame and the connection. The bolts were not

replaced, and the test was continued.

a) Bolt Fracture Location b) Tensile Fracture of Bolts

Figure 5.6.11: Bolt Fracture of SW Beam End Plate Connection (0.36%)

Beam

Gusset

Plate

105

Also at 0.36% drift, the total brace elongation has exceeded 0.2% signifying moderate yielding

(Y3) over its entire length. Figure 5.6.12 shows yielding of the southwest gusset plate which has

also increased to the moderate level covering most of the surface at this drift. Out-of-plane

displacement of the brace center at -0.79% drift had increased greater than 5 in. (B2) as shown in

Figure 5.6.13. Initial yielding (Y1) of the southwest gusset plate beam reentrant corner was also

observed at this negative drift level.

Figure 5.6.12: Y3 of SW Gusset Plate (0.36%)


106


Initial yielding of the southwest beam flange was observed at -1.06 drift along the gusset

interface weld. Initial yielding of the north beam flange at the southern extreme bolt rows was

observed at 0.56% drift. This is shown in Figure 5.6.14. Yielding of the south beam web at the

south edge of the beam to end plate weld was also observed at this drift.

Figure 5.6.14: Y1 of NE Column Flange (0.56%)

A number of damage observations were documented at 0.86% drift. Figure 5.6.15a and Figure

5.6.15b show the concentration of yielding at the brace center resulting in flaking of the mill

scale and severe yielding (Y5) at the northeast gusset plate, respectively. Moderate yielding was

observed at the north beam flange at the end of the load beam. Yielding was also observed on

the flanges of both the east and west columns at their supports.

107


Cupping and bulging of the plastic hinge could be observed at -2.13% drift and is shown in

Figure 5.6.16. The opening and closing moments also increased the yielding for the northeast

and southwest columns. The inside flange of the northeast column is shown yielded at 1.18%

drift in Figure 5.6.17a. The moderate yielding (Y3) of the southwest column outside flange is

shown in Figure 5.6.17b.

Figure 5.6.16: BC Damage at Brace Plastic Hinge (-2.13%)

a) Yielding at Brace Center b) Y5 at NE Gusset Plate

Figure 5.6.15: Increased Yielding of Brace and NE Gusset Plate (0.86%)

108

At 1.60% drift, yielding of the north beam flange adjacent to the gusset plate has increased to a

moderate level (Y3) as shown in Figure 5.6.18. Yielding has also increased at the northeast

column flange bolt holes as a result of the concentrated tensile forces from the extreme bolts as

the frame closes. The concentration of yielding can be seen in Figure 5.6.19.

Figure 5.6.18: Y3 at NE Beam Web (1.60%)

a) Yielding of NE Inside Column Flange b) Yielding of SW Outside Column Flange

Figure 5.6.17: Column Damage (-2.13/1.18%)

109

Slight tearing of the brace at the plastic hinge could be seen in Figure 5.6.20 at the maximum

positive drift ratio achieved of 1.60%. A larger hole was present at the original location of the

tapped screw used to connect the out-of-plane displacement potentiometer. B3 level edge

buckling was observed at both the northeast and southwest gusset plates. Figure 5.6.21 shows

edge buckling of the northeast gusset plate.

Figure 5.6.20: Tearing of Brace at Plastic Hinge (1.60%)

Figure 5.6.19: Yielding at Bolt Holes of NE Column Flange (1.60%)

110

Figure 5.6.21: B2 Edge Buckling of NE Gusset Plate (1.60%)

The level of deformation, cupping and bulging, of the plastic hinge at the minimum drift ratio

achieved of -2.55% is shown in Figure 5.6.22a. The brace fractured the next tension cycle at a

drift of 1.48%. The tear started at the bottom of the plastic hinge (at the tapped screw hole) and

propagated up and around the section. Figure 5.6.22b shows the fractured brace.


The overall performance of this specimen was positive in that the frame reached high drift levels,

4.14% total, and that the controlling failure mode was fracture of the brace at the plastic hinge.

The gusset plates both experience severe yielding over their entire surface equal to or great than

what was seen in previous test. But, no cracking of the interface welds was observed as seen

a) Plastic Hinge prior to Fracture (-2.55%) b) Brace Fracture (1.48%)

Figure 5.6.22: Brace Plastic Hinge Damage and Failure

111

before. Figure 5.6.23a and Figure 5.6.23b show the condition of the northeast and southwest

gusset plates at the end of the test. This level of yielding is more than what was typically seen in

previous test with 3/8 in. rectangular gusset plates.

The frame elements, on the other hand, exhibited much less damage than previous tests.

Moderate yielding was observed at the north beam web but only slight yielding was observed at

the south beam web and at either beam’s flanges. The southwest column saw the most yielding

but still experienced no inelastic deformation as typically seen at the completion of prior tests.

By not transferring yielding from the gusset to the frame elements, the 1 in. thick end plate

connection is not maximizing the full potential ductility of the beam and columns in order to

increase the total drift capacity of the system.

Potentiometers were places at the southwest connection to monitor movement between the

beam end plate and the face of the column. The prying at the north and south of the plate is

plotted over drift and shown in Figure 5.6.24a and Figure 5.6.24b, respectively. The rotation of

the end plate with respect to the column face is plotted in Figure 5.6.25. This is not the same as

the rotation of the entire beam to column connection because it does not account for rotation of

the beam related to the end plate.

a) NE Gusset Plate b) SW Gusset Plate

Figure 5.6.23: Gusset Plate Damage at End of Test

112

Figure 5.6.25: Southwest End Plate Rotation at Column

The 14 – ¾ in. A490 bolts used for the beam end plate connections was a reduced configuration

and did not effectively withstand the demands from the brace forces and the opening and closing

moments on the frame. Significant consideration was given to the demands of the extreme line

of bolts resisting the connection moment during the design but the tension force in the brace

was assumed to be distributed evenly over all the bolts. This is a simplified approach and did not

correlate with what resulted during the test. The fracture of the bolts in the center row at both

the northeast and southwest connections shows greater demand on the bolts closets to the

center line of the brace to transfer the axial load to the frame elements. Accurately modeling

bolt group behavior is a difficult engineering problem and would be helpful in further

understanding the stress and strain demands for this type of bolted connection.

a) Prying at North Edge of SW End Plate b) Prying at South Edge of SW End Plate

Figure 5.6.24: Southwest End Plate Prying

113

5.7 HSS-20: 18 Bolt Beam End Plate Connection


Specimen HSS-20 utilized an 18 bolt configuration of ¾ in. A490 bolts for the beam end plate

connection to the columns. Figure 5.7.1 shows the detail of the beam to column connection.

The members were fabricated by Canron Western Constructors in Portland, OR and erected at

the UW Structures Lab. As stated in the previous section, this test was completed before HSS-

21 but is discuss in a later section in order to be able to reference photos of similar damage and

behavior when necessary.


The design was proposed by Tom Schafly with AISC and Tim Fraser of CANRON to evaluate

the performance of the connection and its influence to the overall performance of the system.

The same 3/8 in. thick gusset plate with 6.8t elliptical clearance was used connected to the 1 in.

thick end plate and the beam with 7/16 in. fillet welds. The beam is welded to the end plate only

at the web and not the flange.

HSS-20 was tested April 20, 2007 and failed during the 35th cycle. The specimen ultimately failed

due to brace fracture at the plastic hinge. A maximum and minimum drift of 1.69% and -2.28

were achieved for a total range of 3.97%. The maximum and minimum lateral loads resisted by

the frame were 321.5 kip and -187.7 kip for total range of 509.2 kip. Figure 5.7.2 shows the

HSS-20 frame drift history during the test and Figure 5.7.3 is a plot of the hysteretic behavior.

114



Table 5.7.1 shows peak results from the HSS-20. The performance section was created using the

written observations from the test and the recorded data. Photo documentation has always been

one of the most effective and important instrumentation tools in recording the behavior of the

test specimen. Without it, relating the drift ranges to exactly when damage occurred loses its

accuracy. Photos are also extremely valuable for reexamining the tests at a later date which

115

almost always results in the reviewer noticing behavior or damage that was not documented on

the day of the test. However, photographic data is not available for this test, because the files

containing the photos were not back up and inadvertently loss with a crashed hard drive.



Initial yielding (Y1) of the southwest gusset plate at the brace end was observed at 0.09% drift.

The lateral force resisted in the frame was 72.3 kip. Slight buckling upward of the brace was

noticed during the first compression cycle at a drift -0.06%. Initial yielding (Y1) of the northeast

gusset plate at the brace end occurred at a drift of 0.15% and the yielding of the southwest gusset

plate increased. Yielding of the southwest column flange (Y1) at the extreme southern bolt hole

was also observed at this drift level. A concussion was heard through the frame at -0.08% drift

but the exact location was unknown. The concussion could possibly be from slip at either bolted

end plate connection as the shear force in the beam to column connection exceeds the frictional

resistance of the pretensioned bolts.

Slight yielding (Y1) of the southwest gusset plate column reentrant corner was observed at -

0.13% drift. Initial yield lines were observed running parallel to the gusset interface weld on the

southwest beam web at 0.34% drift.

At 0.38% drift, yielding of both gusset plates has increased slightly and yield line were observed

at all of the southwest column flange bolts holes for the southwest end plate connection. Initial

yielding (Y1) the north beam web was also seen. Out-of-plane displacement of the brace center

From To Min Max Range Min Max Comp Tens Comp Tens Comp Tens Comp Tens

1 6 -0.07 0.09 0.15 -73.9 72.3 Y1-SWGE

7 8 -0.08 0.15 0.23 -114.0 103.7 YI-NEGE Y1-SWCF

9 10 -0.09 0.22 0.30 -145.7 127.0

11 16 -0.10 0.28 0.38 -170.7 152.2

17 18 -0.13 0.34 0.47 -181.2 179.2 Y1-SWGC Y1-SWBW

19 20 -0.22 0.38 0.60 -181.9 203.9 B1 Y1-NEGC Y1-NEBW

21 22 -0.31 0.41 0.72 -179.0 221.0

23 24 -0.60 0.48 1.08 -176.7 256.4 B2 Y3

Y1-SWGB,

NEGB

Y3-NEG,

SWG

25 26 -0.78 0.53 1.32 -175.8 276.3

Y1-NEBF,

SWBF

27 28 -1.20 0.76 1.96 -183.5 304.5 Y1-NLB

29 30 -1.56 1.07 2.63 -187.3 318.8 Y1-NECF

31 32 -1.94 1.39 3.34 -187.7 321.5 BC

Y5-NEG,

SWG; B1-

NEGB Y1-NECW

33 34 -2.28 1.69 3.97 -182.7 320.5

B2-NEG,

SWG B1-NECF

Y3-SWCF;

B1-SWCF

35 35 - 1.89 - - 164.2 BF

Gusset Plates Beams Columns

HSS-20

Performance

Cycle Drift Ratio Load (kips) Brace

Initial

Drift

Range

Moderate

Severe

116

in buckling exceeded the 2% of the brace length (B1) at -0.22% drift as well as initial yielding

(Y1) of the northeast gusset plate column reentrant corner.

Moderate yielding (Y3) of the brace in tension was calculated as occurring at 0.48% when the

brace elongation exceeded 0.2% of the length. Moderate yielding (Y3) of the northeast and

southwest gusset plate was achieved at this drift. Documentation states that yield lines are

extending from the brace end to the area between the brace and the interface welds, covering the

majority of the gusset plate surfaces. The brace buckled out-of-plane further than 5 in. (B2) the

ensuing compression cycle at a drift of -0.60%. The large displacement as brace buckles

adherently resulted in larger rotational demands of the gusset connections and high

concentration of stress at the reentrant corners. Initial yielding (Y1) of the northeast and

southwest gusset plate beam reentrant corners was also observed at this negative drift.


Slight yielding (Y1) was observed at -0.78% drift on the beam flanges at the ends of the

northeast and southwest beam interface welds. Initial yielding of the north beam flange at the

end of the load beam from the couple created by the eccentricity between the actuator and the

north beam center lines was observed at 0.76% drift. Yielding was also observed at the column

supports for the east and west column bases. Local yielding of the column flanges is occurring

as the axial load in the columns increases due to overturning.


The yielding of the gusset plates continued to increase. Yield lines are running parallel and

perpendicular to the column and beam interface welds and concentrating along the beam

interface welds. Initial yielding of the northeast column flanges occurred around the two

furthest north and south end plate bolts holes at 1.07% drift. The buckled shape of the brace

has become more linear and is clear hinge is forming at the center.

Gusset plate yielding at northeast and southwest connections has increase to severe (Y5) at a

drift of 1.39%. Edge plate buckling of the northeast gusset was observed perpendicular to the

north beam flange. Yielding occurred at the northeast column at the south of the end plate

connection with yield lines flaring out from the edge of the end plate on the inside flange face

and extending into the web. Similar yielding is shown in Figure 5.7.4. Local deformation of the

plastic hinge (BC) could be seen in the form of cupping at the underside and bulging of vertical

sides of the brace section at a drift of -2.13%.

117

Figure 5.7.4: Y1 at NE Column Inside Flange (1.39%)

Moderate yielding (Y3) was observed at the outside flange of the southwest column at 1.69%

drift. Yield line covered the entire flange and has extended into the web. Significant edge plate

buckling (B2) was also observed at the northeast and southwest gusset connections deflecting

greater than the thickness of the plate. Local deformation was could be seen on the inside

flanges of the northeast and southwest columns at a drift of -2.28/1.69% in line with the edge of

the end plates. The 1 in. thick end plates are significantly stiffer than the column flanges and the

edges appeared to be acting as a leverage points during the opening and closing of the frame and

resulting in local damage in this area.

The local deformation of the plastic hinge was extremely severe at the minimum drift of -2.28%.

The surface of the tube was actually warm to the touch from the level of yielding exhibited by

the section. The brace fracture at the plastic hinge the following cycle. Necking of the brace

section could be observed immediately before tearing started from the underside of the section

and propagated up and around.


The specimen performed well in that the controlling failure mode was fracture at the plastic

hinge and yielding was distributed through the gusset plate and into the framing elements. Also,

no cracking of the gusset interface welds was observed. Specimen HSS-20 achieved slightly

lower drift levels than HSS-21, 3.97% compared to 4.14%. Although the total range of

resistance for HSS-20 and HSS-21 were approximately the same, 509.2 and 511.3 kip

118

respectively, frame resistance varied directionally between the two tests. HSS-20 minimum

resistance while the brace was in compression was -187.7 kip while for HSS-20, the specimen

resistance was significantly less (12.8%) at -163.7 kip. The maximum frame resistance with the

brace in tension was larger (7.5%) for HSS-21, 347.6 kip compared to 321.5 kip. Further

comparison of the two tests is included in Chapter 6.

The stiffness of the 1 in. beam end plate limited the frames ability to transfer yielding from the

gusset plate to the web of the column. Although some yielding of the columns was observed,

this connection does not capitalize on the potential ductility of column as seen on the beams.

This connection shows excellent potential for constructability and performance. It would be

worthwhile to look deeper into bolted end plate connections analytically to better understand the

behavior in this type of SCBF. Test parameters related directly to this connection including end

plate thickness, other bolt configurations, and the use of washers as spacers between the end

plate and the column flange face would be useful in determining the maximum achievable

performance for frames with bolted end plate connections. The use of washers as spacers is

shown in Figure 5.7.5 and is thought to help reduce local damage plate edge by creating a gap

and eliminating a pinching point. Yielding would be more evenly distributed over the column

from the opening and closing moments in the frame.

Figure 5.7.5: Bolted End Plate Connection with Spacing Washers

119

5.8 HSS-22: Tapered Gusset Plate, Unwelded Frame Connection


This specimen is similar to HSS-18 in that is uses simple shear tab connection rather than fully

welded beam to column connection adjacent to the gusset plate. Significant material and labor is

required for the fully welded beam to column moment connections used for SCBFs. The AISC

Seismic Provisions does not require a moment connection adjacent to the gusset plate but it has

become the trend of design engineers in major seismic hazard zone to incorporate this

connection for increased stiffness and robustness within the system. Similar to HSS-18,

Specimen HSS-22 was chosen to evaluate the effects of eliminating the fully welded beam

column connection on the overall performance of the system.

The gusset plate thickness and geometry is identical to what was previously tested in HSS-17

(Kotulka 2007). The plate was 3/8 in. thick and uses an 8t elliptical clearance. The 3/8 in. fillet

interface welds were sized to develop the full plastic capacity of the gusset plate. Figure 5.8.1

shows the connection detail for the specimen. Tapered gusset plates have performed well in past

test and have been able to achieve high total drift ranges up to 4.8%. Interface weld damage can

be a major concern because the tapered shape results in a shorter weld length.


120

Specimen HSS-22 was tested August 1, 2007. The specimen failed due to brace fracture at the

plastic hinge while going to peak positive displacement of the 35th cycle. The maximum and

minimum drift ratios achieved were 1.49% and -2.48% for a total drift range of 3.97%. The

frame resisted a maximum and minimum lateral load of 301.8 kip and -132.8 kip for a total range

of 434.6 kip. Figure 5.8.2 shows the actual displacement history for the test. The load verses

drift ratio hysteretic behavior is shown in Figure 5.8.3.



121

The performance of the specimen is summarized in Table 5.8.1 showing peak results for drift

ratio, lateral load, and damage. The controlling failure mode was fracture of the brace and

yielding extending beyond the brace and into the gussets. Severe yielding and interface weld

cracking was exhibited but very little yielding was observed in the framing elements.


5.8.2 Initial Drift Ranges: 0% to 1.25%

Initial yielding (Y1) of the northeast gusset plate occurred at 0.11% drift with the brace in

tension. The resistance in the frame was 101.8 kip when small yielding lines were observed at

the brace end as shown in Figure 5.8.4. Slight upward buckling of the brace could be observed

at -0.12% drift. Yielding (Y1) of the southwest gusset plate was observed at 0.13% drift. Figure

5.8.5 shows small yield lines south of the brace to gusset welds of the southwest gusset plate

connection.


1 6 -0.08 0.07 0.15 -68.0 65.9

7 8 -0.12 0.11 0.22 -99.3 101.8 Y1-NEGE

9 10 -0.16 0.13 0.29 -122.6 132.4 Y1-SWG

11 16 -0.21 0.17 0.38 -134.5 162.9 Y1-SWGE

17 18 -0.22 0.25 0.47 -135.4 185.3

19 20 -0.25 0.35 0.60 -132.6 208.6 B1 Y1-NEGC Y1-NEBF

21 22 -0.31 0.42 0.73 -127.8 226.9 Y1-NEGB

23 24 -0.50 0.52 1.02 -126.0 245.7 B2 Y1-SWGB Y3-NEG Y1-NLBF

25 26 -0.70 0.61 1.31 -125.9 261.5 Y3 B1-NEGB

27 28 -1.11 0.79 1.90 -130.1 287.7 WD-SWGC Y3-SWG

29 30 -1.46 1.02 2.48 -132.8 298.2

WD-SWGB,

NEGC, NEGB

31 32 -1.86 1.21 3.08 -131.4 297.3 BC Y1-SWCF

33 34 -2.48 1.50 3.97 -125.2 301.8

Y5-NEG, SWG;

B1-NEGC,

SWGB, SWGC

- 1.38 - - 171.0 BF

Moderate

Severe

Initial

Drift

Range

Cycle Drift Ratio Load (kips)

HSS-22

Brace Gusset Plates Beams Columns

Performance

122

Figure 5.8.4: Y1 of NE Gusset Plate (0.11%)

Figure 5.8.5: Y1 of SW Gusset Plate (0.13%)

Yielding of the northeast gusset plate increased at 0.17% drift exhibiting long multidirectional

yield lines extending from the brace end to towards the beam and column interface welds. Initial

yielding was also observed at the brace end of the southwest gusset plate. Figure 5.8.6 shows the

increased yielding of the northeast gusset plate at this drift.

123

Figure 5.8.6: Increased Yielding of NE Gusset Plate (0.17%)

At 0.35% drift, initial yielding (Y1) was observed at the column reentrant corner of the northeast

gusset plate. Initial yielding (Y1) was also observed on the northeast beam flange at the gusset

reentrant corner. Figure 5.8.7 shows the yielding at the column reentrant corner. The out-of-

plane displacement of the brace as it buckled in compression exceeded the 2% (B1) of the brace

length at a drift of -0.25%. The slight arching of the buckled brace is shown in Figure 5.8.8.

Figure 5.8.7: Y1 at NE Gusset to Column Reentrant Corner (0.35%)

124

Figure 5.8.8: B1 Buckled Shape of Brace (-0.25%)

Yielding was observed on the northeast and southwest gusset plates along the column interface

welds at 0.42% drift. Figure 5.8.9a and Figure 5.8.9b shows the yield lines stretching from the

beam/column intersection corners out toward the column reentrant corners of the northeast and

southwest gusset plates, respectively. At -0.31% drift, initial yielding (Y1) was observed on the

northeast gusset plate at the beam reentrant corner and yielding increased at the column

reentrant corner. Residual out-of-plane deformation of the brace could be seen after the frame

was returned to the original zero point which would indicate axial lengthening of the brace in

tension.

Yielding of the northeast gusset plate increased with yielding extending over most of the surface

(Y3) at 0.52% drift. The multidirectional yield lines can be seen in Figure 5.8.10 running parallel

a) NE Gusset Plate Yielding b) SW Gusset Plate Yielding

Figure 5.8.9: Yielding along Column Interface Welds (0.42%)

125

to the framing elements and crossing behind the brace end. Yielding also increase on the

southwest gusset place, most notably along the gusset to south beam interface weld. Initial

yielding (Y1) was noted at the north beam flange at the end of the load beam from the

concentrated moment created by the eccentricity of the actuator and the beam. The out-of-plane

displacement of the buckled brace at -0.50% drift reached over 5 in. and is shown in Figure

5.8.11. Initial yielding (Y1) was also observed at the southwest gusset to beam reentrant corner.

Figure 5.8.11: B2 Buckled Shape of Brace (-0.50%)

Figure 5.8.10: Y3 at NE Gusset Plate (0.52%)

126

5.8.3 Moderate Drift Ranges: 1.25% to 2.75%

The calculation for elongation of the brace indicates that the total elongation reached 0.2% at a

drift of 0.61%. The yielding of both gusset plates continued to increase along the interface welds

and in the beam/column intersection corner. Slight edge buckling of the northeast column

along the beam side was also observed. During the opening moment in the frame at -0.70%,

rotation of the northeast shear tab connection could be observed visually.

Yielding of the southwest gusset plate increased to moderate level (Y3) at 0.79% drift. Figure

5.8.12 shows the yield lines covering the majority of the plate and concentrating in the

beam/column intersection corner and along the welds. At -1.11% drift, the buckled shape of the

brace in compression was becoming more linear and forming a hinge at the center as shown in

Figure 5.8.13. Initial cracking was also observed in the plane between the base metal and weld

material at the southwest column reentrant corner. Figure 5.8.14 shows the initiation of cracking

at this location.

Figure 5.8.12: Y3 at SW Gusset Plate (0.79%)

127


Figure 5.8.14: WD at SW Column Reentrant Corner (-1.11%)

Large out-of-plane rotation of the gusset plates was occurring as the brace buckled at -1.46%

drift. As the gusset plate rotates, high stress concentrations occur at the ends of the gusset

interface welds. Initial cracking was observed at both northeast gusset plate reentrant corners

and in the beam reentrant corner of the southwest gusset plate. The crack at the column

reentrant corner of the southwest connection propagated to approximately 3/8 in. as shown in

Figure 5.8.15. All cracking observed is in the heat affect zone (HAZ) of the base metal.

128

Figure 5.8.15: WD SW Gusset to Column Interface Weld Cracking (-1.46%)

5.8.4 Severe Drift Ranges: > 2.75%

Initial yielding of the southwest column flange reentrant corner occurred at 1.21% drift. Overall,

very little yielding has been observed to the framing elements considering the total drift range

has reached 3.08%. The yielding that has been observed is only located at the column flanges

adjacent to gusset plate reentrant corners. As the brace buckled at -1.86% drift, slight plastic

deformation of the brace center was noted in the form of cupping on the underside and bulging

at the vertical sides of the section. Figure 5.8.16a and Figure 5.8.16b show the deformation of

the plastic hinge from above and below.

a) b)

Figure 5.8.16: Damage at SW Gusset Plate Welds (-1.86%)

129

Propagation of the weld cracks at all locations was also observed at -1.86% drift. The crack

length for the southwest gusset to column weld was approximately 7/8 in. and 5/8 in. for the

gusset to beam weld as shown in Figure 5.8.17a and Figure 5.8.17b respectively. The cracks

lengths for the northeast gusset to column and gusset to beam welds were both ¼ in. and shown

in Figure 5.8.18a and Figure 5.8.18b, respectively.

Tearing of the underside of the brace starting at the tap location of the out-of-plane

displacement string potentiometer attachment was observed at the maximum positive drift

achieved of 1.49%. Figure 5.8.19 shows the small tearing at the edges of the tubes section and

the larger hole where the material was tapped to set the 0.11 in. diameter attachment screw.

With the brace in 301.8 kip of tension, necking of the brace could be clearly seen at the hinge as


a) Column Reentrant Corner b) Beam Reentrant Corner

Figure 5.8.18: Damage at NE Gusset Welds (-1.86%)


Figure 5.8.17: Damage at Southwest Gusset Plate Welds (-1.86%)

130

Figure 5.8.20: Necking of Brace at Hinge (1.49%)

The yielding of both the northeast and the southwest gusset plates increased to severe (Y5).

Yielding was extensive and uniform for nearly the entire surface of the plates from the brace to

the interface welds as shown in Figure 5.8.21a and Figure 5.8.21b. Large rotation could be

observed at the northeast beam to column shear tab connection. The inside flange of the north

beam was binding on the inside flange of the east column at the frame reached the peak

maximum drift of 1.48%. The pinching at the beam/column intersection corner of the

northeast gusset plate closed caused local deformation in the form of a bulge in the gusset plate.

Local buckling of all the gusset plate edges (B1) was also observed.

Figure 5.8.19: Tearing of Brace at Plastic Hinge (1.49%)

131

The local deformation at the brace plastic hinge increased to severe (BC) during the final

successful compression cycle at -2.48% drift. Figure 5.8.22 shows the extent of the local

deformation prior to failure the following tension cycle. The crack at the southwest gusset to

column propagated to approximately 1 ¼ in. as shown in Figure 5.8.23. The northeast gusset to

column and gusset to beam weld cracks propagated to approximately 1 ½ in. and 1 ¾ in. Figure

5.8.24 clearly shows the cracks within the HAZ of the base metal at both locations.

Figure 5.8.22: Local Deformation (BC) of Brace (-2.48%)


Figure 5.8.21: Y5 of Gusset Plate (1.48%)

132

Figure 5.8.23: SW Gusset to Column Interface Weld Crack (-2.48%)

Initial yielding of the inside face of the southwest column outer flange opposite the gusset plate

was observed and is shown in Figure 5.8.25. This is the only frame yielding observed beside the

small yield lines observed directly adjacent to the gusset plate reentrant corners at the northeast

beam and southwest column.


Figure 5.8.24: Damage at NE Gusset Welds (-2.48%)

133

The brace fractured (BF) the next tension cycle at 1.38% drift and 171.0 kip resisted by the

frame. Figure 5.8.26 shows the fractured brace which unusually tore along a diagonal line across

the tube section. The residual deformation of the gusset plates after failure shown in Figure

5.8.27a and Figure 5.8.27b was severe. A section of the fractured brace was removed and two

additional cycles were complete to evaluate the stiffness of the frame without the brace.

a) Inside Face of Column Outer Flange b) Outside Face of Column Outer Flange

Figure 5.8.25: Y1 of SW Column (-2.48%)

134


This specimen exhibited weld cracking and severe yielding to the gusset plates while experiencing

little to no yielding of the framing elements. The balance design approach discussed in Chapter

4 is intended to distribute desirable yielding from the brace to the gusset plates and to the

framing elements in order to maximum the ductility of the system. HSS-22 was unable to

capitalize on the potential ductility within the framing elements and concentrated inelastic

deformation and energy dissipation to the brace a gusset plates. Even though the controlling

failure mode of the system was brace fracture at the plastic hinge, the amount of damage at the

welds and severity of the gusset yielding is not encouraging.

Additional instrumentation was added to monitor specific behavior for the shear plate beam-to-

column connection adjacent to the gusset plate. The rotation of the northeast connection

adjacent to the gusset plate was monitored and shown in Figure 5.8.28. It was clear from the

a) NE Gusset Plate Connection a) SW Gusset Plate Connection

Figure 5.8.27: Residual Deformation of Gusset Plate after Brace Fracture (1.38%)

Figure 5.8.26: BF of the Brace Center (1.38%)

135

observations that the beam was rotating with respects to the column shear tab connection at the

higher drift ranges. Figure 5.8.29 shows the rotation of the connection by the markings on the

whitewash and the binding of the inside beam flange against the inside column flange. The

tapered gusset plate adds less rotational stiffness than the rectangular geometry evaluated in HSS-

18. The larger rotational demands of the frame at these connections attributed to the increased

yielding of the gusset plates and the cracking observed at the reentrant corners.

Figure 5.8.28: NE Beam-to-Column Connection Rotation

Figure 5.8.29: NE Beam-to-Column Rotation (1.49%)

136

The stiffness of the fully welded connection also adds torsional resistance to the beam as the

gusset plate rotates out-of-plane to accommodate the displacement of the buckled brace. This

type of rotation could translate to the beams of a composite floor system causing considerable

damage to slab. Instrumentation was added to monitor the torsional rotation of the north beam

12 in. away from the column face. The plot shown in Figure 5.8.30 shows the peak torsional

rotations as the northeast gusset plate rotated. The rotations are small but are increasing. A

stiffer gusset plate configuration than the thin tapered plate used in HSS-22 would increase this

behavior.

Figure 5.8.30: NE Beam Torsion at Connection

5.9 WF-23: Wide-Flange Brace Section


Specimen WF-23 was selected to evaluate an alternate brace section other than the typical

HSS5x5x3/8. A W6x25 wide-flange section was selected for its similar tension capacity but has a

lower expected compressive capacity. The material used to manufacture HSS tubes is carbon

steel under the ASTM designation A500 Gr. B. Experimental tests have shown that the cold

rolled A500 steel tubes do not perform well when subjected to nonlinear cyclic loading. Wide-

flange sections are thought to be a more ductile brace section and have a better tolerance to low

cycle fatigue. Wide-flange shapes are hot-rolled sections manufactured using high strength, low

alloy steel. Also because of the geometry of tubes, approximately 50% of the sectional area is

near the extreme fibers and it has a larger radius of gyration about both axes. Although this

137

geometry is ideal for elastic buckling, the material at the extreme ends experiences the highest

stresses and largest strains in the inelastic buckled state. Typically, initiation of cracking and

tearing of the brace occurs shortly after large strains and local deformation are exhibited at the

brace plastic hinge. Wide-flange sections have less material at the extreme edges when buckling

about the weak axis.

The connection detail was based on discussions with design engineers from Rutherford and

Chekene Structural Engineers in San Francisco, CA and is oriented to buckle out-of-plane about

its weak axis. The connection detail is shown in Figure 5.9.1. It consisted of a slotted extended

brace flanges and a brace web connected with splice plates. The web splice plates are designed

to account for the reduced net section due to the slots in the flanges. The 3/8 in. thick gusset

plate geometry utilizes an 8t elliptical clearance and the 3/8 in. interface fillet welds are sized to

develop the full plastic capacity of the plate. The beam to column connections adjacent to the

gusset plates are fully welded moment connections.

Figure 5.9.1: WF-23 Connection Detail

WF-23 was tested September 25th, 2007 and was subjected to 41 cycles before failing

simultaneously due to fracture of northeast gusset interface welds and complete bolt fracture of

the northwest shear tab connection. The specimen achieved maximum and minimum story

drifts of, 2.35% and -3.21%, for a total drift range of 5.56%. The frame maximum and

minimum lateral resistance was 338.2 kip and -149.5 kip, for a total range of 487.7 kip. Figure

Specimen WF-23

138

5.9.2 shows the actual frame displacement during the test cycles. The resulting hysteretic

behavior is plotted in Figure 5.9.3.

Figure 5.9.2: WF-23 Displacement History

Figure 5.9.3: WF-23 Hysteresis

Whitewash was used to assist in identifying yielding of the brace section for this test. The

nomenclature used to describe damage is similar for the wide flange brace section as for the

framing element described in Sections 5.2 and 5.3. For example, initial yielding o the brace

flange at the northeast end of the brace would be labeled Y1-NEF under the brace section of the

139

performance summary in Table 5.9.1 below. Table 5.9.1 also shows the maximum and minimum

drift ratios, lateral forces resisted by the frame, and performance of components are all peak

values during the test.

Table 5.9.1: WF-23 Peak Results


Initial yielding (Y1) of the northeast gusset plate at the brace end was observed after the first

cycle at 0.07% drift. A few small yield lines visible off the end of the south flange as shown in

Figure 5.9.4. Very slight upward buckling was noticed in the brace at -0.06% drift.



1 6 -0.06 0.07 0.12 -69.5 78.8 Y1-NEGE

7 8 -0.09 0.10 0.18 -99.4 112.1

Y1-NEF,

SWF

9 10 -0.12 0.13 0.25 -120.8 141.0

Y1-NEF,

SWF

11 16 -0.18 0.16 0.35 -128.5 173.6 Y1 Y1-SWGE Y1-SWBW

17 18 -0.27 0.20 0.48 -121.0 193.9 Y1-NEGC

19 20 -0.38 0.24 0.62 -107.0 215.4

B1;

Y3-BH Y1-NLBF

21 22 -0.49 0.29 0.78 -100.0 232.2 B2

23 24 -0.67 0.40 1.07 -103.7 255.5 Y3 Y1-NEBW

25 26 -0.83 0.53 1.36 -109.1 264.9 Y5-BH Y1-NEBF

27 28 -1.16 0.74 1.90 -121.8 291.7

Y1-NEGB, NEGC,

SWGB, SWGC B1-NLBF Y1-NECF

29 30 -1.48 0.98 2.47 -135.5 305.4

Y3-NEG; B1-

NEG, SWG Y1-SWCF Y1-NECW

31 32 -1.80 1.23 3.02 -141.5 316.5 Y3-SWG

33 34 -2.15 1.48 3.63 -145.6 324.9 Y3-SWCF

35 36 -2.49 1.77 4.26 -148.4 330.8

B2-NEG,

SWG Y3-NECF

Y1-SWCW;

B1-NECF

37 38 -2.86 2.05 4.91 -149.5 336.3

WDB-NEGB,

NEGC, SWGC

Y5-NEG,

SWG B1-SWCF

39 40 -3.21 2.35 5.56 -137.3 338.2 SWD-NEGB Y1-SEBW Y3-NEBF Y3-NECW Y5-SWCF

41 41 - 2.32 - - 243.5 WF-NEGB

Performance

HSS-23

Drift

Range

Cycle Drift Ratio Load (kips)Brace Gusset Plates Beams

Moderate

Initial

Columns

Severe

140

At 0.10% drift, initial yielding (Y1) of the brace was observed around each hole at the end of the

slotted flanges at the northeast and southwest brace ends connecting to the gusset plate. The

lateral force resisted by the frame at the initial yielding was 78.8 kip. Figure 5.9.5 shows an

example of the yielding at the north flange of the southwest brace to gusset plate connection.

Initial yielding of the northeast gusset plate was observed between the brace and the north beam

weld. A small amount of yielding was noticed approximately 6 in. away from the gusset edge on

the brace flanges while the brace was in compression at -0.08% drift. This can be seen in Figure

5.9.6a and occurs close to the end of the web slice plate. Initial yielding (Y1) was also noted on

the flange of the southwest brace end 2 in. inside the gusset plate as shown in Figure 5.9.6b.

Figure 5.9.5: Y1 at Brace Flange Reduced Section in Tension (0.10%)

The yielding of the brace, along the north flange, approximately 6 in. off the gusset edge at the

a) b)

Figure 5.9.6: Y1 of Brace Flange in Compression (-0.08%)

141

northeast and southwest ends increased slightly at -0.12% drift. Initial yielding (Y1) of the brace

flange two inches inside the start of the northeast gusset plate, similar to that on the southwest

end, was observed. Figure 5.9.7 shows both of these types of yielding for the north flange at the

northeast brace end. Figure 5.9.8 shows the location of the yielding relative to the web splice

plate of the brace to gusset connection.

Figure 5.9.7: Yielding at NE Brace on Flange (0.12%)

Figure 5.9.8: View of Yielding at NE of Brace (0.12%)

Initial yielding (Y1) of the southwest gusset at the brace end and the southwest beam web

occurred at 0.16% drift. Initial yielding (Y1) of the lower edge of the brace flanges occurred as

the brace buckled at -0.18% drift. Figure 5.9.9 shows the buckled shape of the brace and the

yielding over the length. Increased yielding was observed at the brace ends for the northeast and

southwest gusset plates in compression. Initial yielding (Y1) of the northeast beam flange at the

gusset plate reentrant corner was noted at -0.18% drift.

142

Yielding of the northeast gusset plate column reentrant corner was observed at 0.20% drift. The

yielding at the brace ends of the northeast and southwest gusset plates has increase at this

deformation as shown in Figure 5.9.10.

Figure 5.9.10: Yielding at NE Gusset Plate (0.20%)

At -0.27% drift, weak axis flexural yielding of the brace had increased. Yielding at the top edge

of the brace flanges and significantly more yielding along the bottom edge toward the center of

the brace was observed. Figure 5.9.11a and Figure 5.9.11b show the brace yielding and buckled

shape, respectively. Yielding at the northeast brace end has significantly increased as the gusset

plate rotated out-of-plane. A concentrated yield pattern at the brace end of the gusset plate is


Figure 5.9.9: Y1 on the Lower Edge of the Brace Flange (-0.18%)

143

Figure 5.9.12: Gusset Plate Yielding at Northeast Brace End (-0.27%)

Initial yielding of the north beam flange at the end of the load beam was noted at 0.24% drift.

The out-of-plane displacement of 3.95 in. at the brace center exceeded 2% of the brace length

(B1) at -0.38% drift. Moderate Yielding (Y3) concentrated at the brace center was observed as

the buckled shape began to look more linear. Figure 5.9.13 shows the yielding of the brace

center at -0.38% drift. This yielding is designated in the performance portion of Table 5.9.1 as

Y3-BH, for moderate yielding at the brace hinge.

a) Brace Yielding b) Brace Buckled Shape

Figure 5.9.11: Brace Condition (-0.27%)

144

Figure 5.9.13: Y3 at Brace Center (-0.38%)

Out-of-plane brace buckling exceeded 5 in. at -0.49% drift displacing 5.14 in. (B2). The yielding

at the brace center continued to increase and hinging could be seen in the buckled shape. Figure

5.9.14 shows the buckled shape of the brace in compression.

Figure 5.9.14: B2 Buckled Brace Shape (-0.49%)

Instrumentation monitoring the brace elongation indicated that the brace elongation exceeded

0.2% at 0.40% drift and tensile yielding had clearly occurred. The total brace elongation was 0.35

in. over the brace length, gusset edge to gusset edge, of 132 11/16 in.. Very slight yielding (Y1)

was also observed at the northeast beam web running parallel to the gusset plate interface weld at

0.40% drift.

145


Initial yielding (Y1) of the northeast beam flange approximately 5 in. from the face of the

column and at the beam reentrant corner was observed at -0.83% drift. Yielding of the brace

center has increased to severe (Y5) and is covering both the brace flanges and web. Figure 5.9.15

shows the yielding of the brace center at -0.83% drift.

Figure 5.9.15: Y5 at Brace Center (-0.83%)

At 0.74% drift, both gusset plates increased yielding. Multidirectional yield lines extended from

the brace ends at both the northeast and southwest locations. Initial yielding of the northeast

column flange at the gusset reentrant corner and increased yielding at the northeast beam flange

at the gusset reentrant corner are shown in Figure 5.9.16a and Figure 5.9.16b. Very slight

yielding of the southwest beam inside flange approximately 5 in. from the west column face was

noted.

146

At -1.59% drift, the couple created by the eccentricity of the actuator from the beam caused local

deformation of the north beam flange and yielding across the beam web. Yielding was also

observed at the load beam bolts of the north beam flange. As the out-of-plane rotation of the

gusset plates due to brace buckling increased, initial yielding of both reentrant corners of the

southwest and northeast gusset connection occurred. Increased yielding of the column flange at

the northeast gusset reentrant corner was also observed. Figure 5.9.17 shows the yielding of the

northeast gusset reentrant corners at -1.16% drift.

Yielding increased over the majority of the surface of the northeast gusset plate at 0.98% drift.

Moderate yielding (Y3) was concentrated at the brace end but extended to the reentrant corners

and along the gusset to brace welds as shown in Figure 5.9.18. Local buckling (B1) of all the

a) Northeast Beam Reentrant Corner b) Northeast Column Reentrant

Corner Figure 5.9.17: Y1 at NE Gusset Reentrant Corners (-1.16%)

a) Northeast Column b) Northeast Beam

Figure 5.9.16: Y1 at NE Column and Beam Flanges (0.74%)

147

gusset plate edges was observed at 0.98% drift. Yielding also increase over the southwest gusset

plate and at the northeast column inside flange at the reentrant corner. Slight initial yielding (Y1)

was noted at the northeast column web adjacent to the fully welded beam web connection.


Initial yielding (Y1) at the outside flange of the southwest column occurred at -1.48% frame

drift. Figure 5.9.19 shows the yielding opposite the gusset edge on the outside face of the

southwest column flange. The yielding at the center of the brace was concentrated over an

approximate 18 in. length as shown in Figure 5.9.20. Yielding also increased at all of the gusset

plate reentrant corners.

Figure 5.9.19: Y1 at SW Column (-1.48%)

148

Figure 5.9.20: Yielding at Brace Center (-1.48%)


Yielding of the southwest gusset plate increased to moderate (Y3) at 1.23% drift. Figure 5.9.21

shows yield lines covering the majority of the plate from the brace end extending to the reentrant

corners and along the gusset to brace welds. Yielding also increased at the northeast gusset plate.

Local deformation of the brace center was observed in the form of pinching at the top edges of

the flanges and flaring out at the bottom edges.

Figure 5.9.21: Y3 at SW Gusset Plate (1.23%)

Local edge buckling of the gusset plate edges also increased as the opening and closing moment

at the beam/column connection increased at these higher drifts. Figure 5.9.22 shows edge

buckling of the southwest gusset plate column edge. Significant rotations of the northwest and

southeast shear tabs connections were also observed at 1.23% drift.

149

Figure 5.9.22: Edge Buckling at SW Gusset Plate (1.23%)

Yielding of the framing elements increased at -1.80% drift. The yielding at the outside flange of

the southwest column increased and is shown in Figure 5.8.24. The northeast column flange

adjacent to the gust reentrant corner exhibited increased yielding and yield line running

perpendicular to the length of the column were observed in the web. Yielding of the east

column support was observed at the completion of the cycle at drifts -1.80% and 1.23%.

Figure 5.9.23: SW Column Yielding (-1.80%)

Increased yielding of northeast beam and column inside flanges are shown in Figure 5.9.24a and

Figure 5.9.24b at 1.48% drift, respectively. The yielding at both locations extends over the full

width of the flange at the gusset plate reentrant corners. Initial yielding was observed at the

southwest column inside flange intersecting with the inside flange of the south beam. Figure

5.9.25 shows yielding at the fully welded beam flange.

150

Figure 5.9.25: Y1 at SW Column/Beam Connection (1.48%)

Increase yielding of the northeast column web was also observed at 1.48% drift. Figure 5.9.26

shows the pattern of yielding with lines running perpendicular to the length of the column.

Yielding was also noted at inside face of the southwest beam inside flange as shown in Figure

5.9.27.

a) Northeast Beam b) Northeast Column

Figure 5.9.24: NE Frame Yielding (1.48%)

151

Figure 5.9.26: NE Column Web Yielding (1.48%)

Figure 5.9.27: SW Beam Flange Yielding (1.48%)

The outside flange of the southwest column flange significantly increased to a moderate damage

state (Y3) at -2.15% drift. Figure 5.9.28 shows clear “V” shaped yielding at the inside face of the

outside flange of the southwest column opposite the edge of the gusset plate.

152

Figure 5.9.28: Y3 at SW Column (-2.15%)

All the described damage below occurred at 1.77% drift. Moderate level edge buckling (B2) was

observed at all four gusset plate edges. The local out-of-plane buckling greater than the

thickness of the plate is shown in Figure 5.9.29. Yielding of the southwest column web

increased adjacent to the beam web and gusset plate welds. Figure 5.9.30a and Figure 5.9.30b

shows this yielding and the increased yielding of the northeast column web. In both cases, the

yield lines are running perpendicular to the length of the column. Local deformation of the

northeast column inside flange was observed and is shown in Figure 5.9.31. Also, it was

observed that the top flanges at the center of the brace remained buckled inward while the brace

was in tension.

Figure 5.9.29: B2 Edge Buckling at SW Gusset Plate (1.77%)

153

Figure 5.9.31: NE Column Deformation (1.77%)

Yielding at the northeast column inside flange increase to moderate level (Y3) and extended into

the column web at -2.49%. Figure 5.8.34a and 5.8.34b show the yielding at the northeast column

flange and web located at the edge of the gusset plate. Binding of the south beam inside flange

against the southeast column inside flange was noted due to the rotation of the southeast shear

tab connection.

a) SW Column b) NE Column

Figure 5.9.30: Column Web Yielding (1.77%)

154

Yielding of the northeast and southwest gusset plates has increased to severe (Y5) at 2.05% drift.

Both plates experienced significant yielding over the entire surface and have wide spread yielding

in areas as shown in Figure 5.9.33a and Figure 5.9.33b. Local deformation of the northeast

column inside flange increased and initial deformation (B1) of the southwest column outside

flange was observed as shown in Figure 5.9.34. Buckling of the gusset plate edges also increased

with local deformation of the northeast gusset plate beam edge shown in Figure 5.9.35.


Figure 5.9.33: Y5 of Gusset Plate (2.05%)

a) Column Flange b) Column Web

Figure 5.9.32: Y3 at NE Column (-2.49%)

155

Figure 5.9.34: SW Column Deformation (2.05%)

Figure 5.9.35: Gusset Edge Plate Buckling (2.05%)

Cracking occurred at the interface welds as the gusset plates rotated to accommodate the out-of-

plane displacement of the brace at -2.86% drift. The length of the crack at the northeast gusset

beam reentrant corner was approximately 2 in. and is shown in Figure 5.9.36a. Figure 5.9.36b

shows the crack at the column reentrant corner of the northeast gusset plate reaching

approximately 1 in. in length. A smaller crack of ¼ in. was observed at the southwest gusset

plate column reentrant corner. All of the cracks initiated at the plane between the weld material

and the base metal and propagated in the HAZ of the base material.

156

Residual deformation of the brace center was observed as the brace was in tension at 2.35%

drift. Figure 5.9.37 shows brace center yielded and the pinching at the top edges of the flanges.

Yielding of the framing elements also increased. Both yielding and deformation of the southwest

column flange and web increased. The severe yielding (Y5) shown in Figure 5.9.38a and Figure

5.9.38b cover the full width of the flange and extend through the web. Yielding also increased at

the southwest beam/column moment connection on the flange of the inside column around the

CJP. This is shown in Figure 5.9.39.

Figure 5.9.37: Residual Deformation at Brace Center (2.35%)

a) Beam Reentrant Corner b) Column Reentrant Corner

Figure 5.9.36: Cracking (WD) at NE Gusset Plate Interface Welds (-2.86%)

157

Figure 5.9.39: SW Beam/Column Moment Connection Yielding (2.35%)

Gusset plate edge deformation increased at 2.35% drift at all four locations. Yielding of the

southwest beam inside flange increased to cover a significant area around the gusset plate

reentrant corner. The yielding in the web of the northeast column increased to a moderate level

(Y3) and is shown in Figure 5.9.40.

a) Web and Inside Flange b) Outside Flange

Figure 5.9.38: SW Column Damage (2.35%)

158

Figure 5.9.40: Y3 at NE Column Web (2.35%)

A peak minimum drift of -3.21% was achieved prior to failure the following tension cycle. All

the descriptions below occurred at this drift. Yielding of the northeast column increased and

extended to the outside flange face of the section. The gusset plate interface weld damage

increased as the gusset rotated and the brace buckled. The northeast gusset plate to beam weld

propagated through the base material leaving only approximately 1 in. of weld remaining. Figure

5.9.41a and Figure 5.9.41b show the extent of the weld damage at -2.97% and then at -3.21%,

respectively. The crack at the southwest gusset plate to column weld propagated to

approximately 1 in.

The northeast gusset plate interface welds failed before reaching the peak maximum drift of the

following tension cycle. The drift at failure was 2.32%. All four bolts at the northwest shear tab

connection sheared and severe shear deformation could be seen in the bolts of the southeast

a) SWB at NE Gusset Plate (-2.97%) b) SWB at NE Gusset Plate (-3.21%)

Figure 5.9.41: Severe Damage at NE Gusset to Beam Weld

159

shear tab connection after removal. Figure 5.9.42a and Figure 5.9.42b show the fracture welds at

the northeast gusset plate and the failed northwest shear tab connection, respectively.


WF-23 achieved a large total drift range and yielding was distribution from the brace to the

gusset plate and into the framing elements follows the intended of the Balance Design

Procedure. It is difficult to determine the sequence of the failure between fracture of the

northeast gusset plate interface welds and the northwest shear tab bolt fracture and if one was

possibly the result of the other. The desired failure mode of brace fracture was not achieved.

Figure 5.9.43 shows the condition of the hinge at the brace center at the peak minimum drift of -

3.21%. Yielding is severe and some deformation was noted but the section was not able to fully

develop the concentrated local deformation and large strains associated with the plastic hinge

seen with tube brace sections and other tests using wide-flange brace sections.

a) NE Gusset to Frame Welds Fracture b) NW Shear Tab Bolt Fracture

Figure 5.9.42: System Failure Modes (2.32%)

160

Figure 5.9.43: Brace Center Damage (-3.21%)

The out-of-plane displacement of the brace was greater than any seen in the previous SCBF tests

at UW. Figure 5.9.44 shows the out-of-plane displacement at the center of the brace related to

drift.

Figure 5.9.44: Out-of-Plane Displacement at Brace Center

As the total drift on the frame increased, the rotational demands of the shear tab connections

also increased. Bolt-hole elongation of the beam web from bearing at the northwest connection

was observed after fracture as shown in Figure 5.9.45. The southeast shear tab connection saw

yielding at the beam web adjacent to the connection plate. There was also significant shear

deformation of the bolts. Figure 5.9.46a and Figure 5.9.46b shows the conditions of the

southeast shear tab connection and bolts after system failure.

161

Figure 5.9.45: Bolt-Hole Elongation (Post Test)

5.10 HSS-24: Welded Flange, Bolted Web Frame Connection


Specimen HSS-24 was chosen to evaluate an alternate beam to column connection adjacent to

the gusset plate. Most test specimens in this research program utilized complete joint

penetration (CJP) welds to connect the flanges and web of the beam to the column. These welds

enhance the system performance but the benefit must be weighed against the cost of material

and labor during construction. The HSS-24 connection consisted of a bolted shear plate and

CJP welded flange beam-to-column connection. The connection should be stiffer and stronger

than the simple shear connections evaluated in test HSS-18 and HSS-22.

a) Beam Yielding b) Shear Deformation of Bolt

Figure 5.9.46: Damage at SE Shear Tab Connection (Post Test)

162

The 3/8 in. thick gusset plate geometry uses an 8t elliptical clearance and interface welds sized to

develop the full plastic capacity of the plate. This again as well as the brace is identical to that of

HSS-05 that achieved the drift ranges within the previous UW test programs. The shear tab was

design for the vertical component based of force distribution from the AISC Uniform Force

Method (UFM). The ½ in. thick shear tab is welded to the column with 5/16 in. fillet weld on

each side and bolted to the beam using 4-1 in. diameter A490 bolts tightened to a force of 64 kip

based on AISC Table J3.1 for the Minimum Bolt Pretension. Figure 5.10.1 shows the

connection detail.


HSS-24 was tested in the UW Structures Lab May 7, 2008. The specimen fractured at the brace

center while achieving maximum and minimum drift ratios of 1.94% and -2.50% for a total range

of 4.44%. The maximum and minimum forces resisted by the frame were 339.7 kip and -164.1

kip. The frame displacement history is shown in Figure 5.10.2. The hysteretic behavior of the

specimen is given in Figure 5.10.3.

163

Figure 5.10.2: HSS-24 Displacement History


Peak performance values are given in Table 5.10.1along with a summary of component damage.

Yielding was observed brace and gusset plates and into the framing elements. The damage

observed to the columns was significantly more than what was seen in the beams. The yield

pattern over the gusset plate was also unique with more yielding closer to the column interface

weld than the beam. Some ductile tearing was observed at the end of the gusset interface welds.

The bolt fractures that occurred at the shear tab connections prior to brace fracture were

attributed to the use of a shorter bolt length than in previous tests. The bolt length was

164

incorrectly selected and inadvertently aligned with the threads into the shear plane of the bolted

connection. After the bolts fractured, new bolts were installed in order to complete the test.



Initial yielding (Y1) of the northeast gusset plate was observed at 0.07% drift of the first cycle

and 81.7 kip resisted by the frame with the brace in tension. Small yield lines could be seen

extending from the end of the brace toward the northeast column. Initial yielding (Y1) of the

southwest gusset plate in a similar fashion was observed at 0.11% drift.

At -0.16% frame drift, initial yielding of the southwest beam flange at the gusset plate reentrant

corner was observed. Upward out-of-plane displacement and elastic buckling was observed in

the brace while in compression at a drift of -0.21%. Tensile yielding has also increased at both

the northeast and southwest gusset plates at the brace ends at a drift of 0.23%.

Initial yielding was observed in the form of small whitewash flaking from the column flange of

the northeast gusset plate reentrant corner at this drift. Figure 5.10.4 shows the initial yielding

(Y1) at the northeast column flange at 0.27% drift.


1 6 -0.07 0.07 0.14 -63.7 81.7 Y1-NEGE

7 8 -0.09 0.11 0.21 -87.1 116.1 Y1-SWGE

9 10 -0.12 0.15 0.27 -114.2 145.9

11 16 -0.16 0.20 0.36 -135.1 177.0 Y1-SWBF

17 18 -0.21 0.23 0.44 -144.9 200.2

19 20 -0.30 0.27 0.58 -146.3 222.5 Y1-NECF

21 22 -0.39 0.30 0.69 -144.4 237.7 B1Y1-NEGC,

SWGCY1-SWBW

23 24 -0.63 0.37 0.99 -144.0 268.1 B2 Y1-NLB

25 26 -0.81 0.46 1.27 -145.7 287.0 Y3 Y3-NEG Y1-NEBW

27 28 -1.17 0.68 1.86 -154.8 316.1 Y1-SWCF Y3-NECF

29 30 -1.50 0.98 2.48 -161.6 331.0B1-NEG,

SWG

Y1-NEBF,

SWBFB1-NECF

31 32 -1.83 1.34 3.17 -164.1 336.4WD-NEGB,

NEGC, SWGCY1-SWCF

33 34 -2.14 1.71 3.85 -154.4 337.0 BCB2-NEG,

SWGY3-NECW

35 35 -2.50 1.94 4.44 -148.0 339.7 Y5-NEG Y3-SWCF

36 36 - 2.07 - - 187.2 BF

Columns

Severe

Moderate

Initial

HSS-24

Drift

Range

Cycle Drift Ratio Load (kips)Performance

Brace Gusset Plates Beams

165

Figure 5.10.4: Y1 at NE Column at Gusset Reentrant Corner (0.27%)

All of the following descriptions of yielding occurred at 0.30% drift and a force of 237.7 kip

resisted by the frame with the brace in tension. Increased yielding occurred at the ends of the

brace at the northeast and southwest gusset plates. Yielding of the northeast has extended

beyond the brace end and toward the beam interface weld. Yield lines were also visible along the

brace to gusset plate weld of the southwest gusset plate connection. Initial yielding (Y1) of the

gusset plate was observed at the northeast column reentrant corner and southwest column

reentrant corner.

At -0.39% drift and 144.4 kip of force resisted by the frame with brace in compression, the brace

out-of-plane displacement of 3.56 in. exceeded 2% of the brace total length (B1). The buckled

shape and the brace and the out-of-plane rotation of the northeast gusset plate can be seen in

Figure 5.10.5a and Figure 5.10.5b, respectively. Compressive yielding of the southwest gusset

plate seen as a concentration of yield lines at the brace end occurred as the gusset plate rotated to

accommodate the out-of-plane displacement of the buckling brace. Initial yielding (Y1) of the

southwest beam web was also observed with perpendicular yield lines running along the gusset

interface weld.

166

At 0.37% drift and 268.1 kip frame resistance, yielding increased at the gusset plates and in the

framing elements. Increased yielding along the brace to gusset welds was seen at both the

northeast and southwest gusset plates. Yielding of southwest beam web also increased with yield

lines running perpendicular to and along the gusset interface welds. Yielding increased at the

northeast column flange at the gusset plate reentrant corner. Initial yielding of the northeast

column flange also occurred at the north beam inside flange to column connection. The yielding

resulting from the beam flange forces being transferred to the column as the frame closes is

shown in Figure 5.10.6 at the drift of 0.37%. Initial yielding (Y1) was also observed at the north

beam flange at the end of the load beam due to the concentrated moment from the actuator

eccentricity to the beam work point.

Figure 5.10.6: NE Column Flange Yielding at Beam/Column Connection (0.37%)

a) Buckled Shape b) SW Gusset Plate Rotation

Figure 5.10.5: B1 Level Brace Out-of-Plane Displacement (-0.39%)

167

The buckled brace center out-of-plane displacement reached 5.24 in. and exceeded the depth of

the member (B2) at a drift of -0.63% and -144.0 kip resisted by the frame. The buckled shape of

the brace is shown in Figure 5.10.7. Also at this drift, increased yielding of the northeast gusset

plate resulted in long stretching yield lines from the brace end to the column interface weld.

Yielding also increased on the southwest gusset plate directly at the brace end from the out-of-

plane rotation of the plate as the braced buckled. Residual out-of-plane displacement of the

brace could be seen after the completion of this cycle and the frame was return to the original set

point.

Figure 5.10.7: B2 Level Buckling of Brace (-0.63%)


The potentiometer monitoring the brace elongation reached 0.27 in., more than 0.2% of the

original length (Y3), free edge of gusset plate to free edge of gusset plate, at a drift of 0.46% and

287.0 kip resisted by the frame. Yielding of the northeast column flange at the gusset plate

reentrant also increased at this drift level as shown in Figure 5.10.8 extending upward over the

flange width. Initial yielding (Y1) of the northeast beam web in the form of a two small yield

lines towards the center of the section was observed. Also at this drift, the yielding of the

northeast gusset plate notably increased.

168

Figure 5.10.8: NE Column Flange Yielding (0.46%)

The base of the east column showed compressive yielding over the east flange due to the

overturning forces concentrated at the column based as the frame was pushed out to the drift of

0.46%. This is shown in Figure 5.10.9.

Figure 5.10.9: Yielding at East Column Base (0.46%)

Yielding of the northeast gusset plate significantly increased to the moderate level (Y3) at -0.81%

drift and 145.7 kip of resistance from the frame. The yielding extends over most of the plate

from the brace end, along the brace to gusset weld, and to the interface welds as shown in Figure

5.10.10. Yielding of the southwest beam web increased along the interface weld, with yield lines

running parallel as well as perpendicular to the direction of the weld. It was also noted that

169

additional yielding occurred at the northeast column flange at the gusset plate reentrant corner

with the brace in compression and the frame opening.

Figure 5.10.10: Y3 at NE Gusset Plate (-0.81%)

Yielding of the northeast column at the reentrant corner significantly increased at 0.68% frame

drift, to a moderate damage level (Y3) with evidence of yielding over the width of the flange and

extending through the web. Figure 5.10.11 shows yielding from the inside view of the column

adjacent to the gusset plate reentrant corner. Yielding of the northeast and southwest gusset

plates also increased in the area between the brace to gusset weld and the interface welds.

Yielding of the southwest beam web and the north beam flange at the end of the load beam also

increased at 0.68% drift and 316.1 kip resisted by the frame.

Figure 5.10.11: Y3 at NE Column Flange (0.68%)

170

The shape of the buckled brace in compression began to appear more linear at -1.17% drift.

Figure 5.10.12 shows the buckled shape and the formation of 3 hinges, at the gusset plates and at

the brace center. Again, yielding of both gusset plates visibly increased, although yielding at the

southwest was less than at the northeast gusset plate.


Initial yielding of the southwest column flange at the gusset plate reentrant corner was observed

at -1.17% drift. The yielding of the gusset plate at the southwest column reentrant corner also

increased at this drift. Both conditions can be seen in Figure 5.10.13.

Figure 5.10.13: Increased Gusset Plate Yielding and Y1 at SW Column Flange (-1.17%)

At 0.98% drift, slight deformation of the gusset plate free edges was observed. Yielding

increased across the web of the northeast beam adjacent to the interface weld toward the shear

tab beam to column connection. There was also a slight increase in yielding at the southwest

Gusset Plate

Column

171

beam web. Yielding increased and extended over the northeast and southwest gusset plates.

Slight deformation of the northeast column inside flange was noted along the gusset plate.

Figure 5.10.14 shows the deformation of the column flange at this drift.

Figure 5.10.14: B1 at NE Column Flange (0.98%)

All of the following damage observations occurred during the ensuing compression cycle at -

1.50% drift and frame resistance of -161.6 kip. Yielding increased at the ends of the brace and

along the brace to gusset welds at both the northeast and southwest gusset plates. Initial yielding

(Y1) was observed at the north and south beam flanges at the gusset reentrant corners. This

yielding extended from the edge of the gussets is shown for the northeast and southwest beams

in Figure 5.10.15a and Figure 5.10.15a, respectively. The yielding on the both the NE and SW

gusset plates at the beam and column reentrant corners also increased. Significant increase of

yielding of the northeast column web was observed while long yielding lines extending most of

the length of the adjacent interface weld toward the beam/column connection.

a) NE Beam Flange b) SW Beam Flange

Figure 5.10.15: Y1 of Beam Flanges at Reentrant Corners (0.87%)

172


Yielding of the northeast column in the area around the gusset plate reentrant corner

significantly increased to a moderate damage level (Y3) at 1.34% drift and 336.4 kip of frame

resistance. Figure 5.10.16a and Figure 5.10.16b show the damage over the flange and web of the

northeast column. Yielding also increased at the south flange of the beam to column connection

as shown in Figure 5.10.17.

Figure 5.10.17: Yielding at NE Beam/Column Connection (1.34%)

Again, yielding increased on the gusset plates in the area between the brace and the interface

welds, especially between at the southwest beam interface welds at 1.34% drift. It was also noted

that the yielding of the northeast gusset plate was more severe between the brace and the beam

interface weld rather than brace and the column. The unbalanced yield patterns are likely due to

the inability of the beam to transfer vertical forces from the brace to the column because of the

a) NE Column Flange b) NE Column Web

Figure 5.10.16: Y3 at NE Column (1.34%)

173

reduce stiffness at the bolted shear tab connection. The asymmetric yielding of the northeast

gusset plate is shown in Figure 5.10.18.

Figure 5.10.18: Unbalanced Yielding at NE Gusset Plate (1.34%)

The frame resisted the minimum force of -164.1 kip at a drift of -1.83%. Initial cracking of the

northeast column and northeast beam interface welds were observed at this drift. Initial cracking

was also observed at southwest gusset plate to column weld. Initial yielding (Y1) of the

southwest column flange was seen in the area across from the gusset plate end. Significant

increased of yielding to the northeast column web could be seen and is shown in Figure 5.10.19

with long lines running longitudinally up the column length toward the beam/column

intersection.

Figure 5.10.19: NE Column Web Yielding (-1.83%)

174

Moderate level yielding (Y3) and increased deformation of the inside flange were observed at the

northeast column as the yielding over the section increased at 1.71% drift. Figure 5.10.20a and

Figure 5.10.20b show the condition of both the flange and the web. Gusset plate edge buckling

increased to moderate levels (B2) when the buckling out-of-plane exceeded the thickness of the

3/8 in. gusset plates at this drift. The column edge of southwest gusset plate and beam edge of

the northeast gusset plate are shown in Figure 5.10.21a and Figure 5.10.21b. Yielding also

increased at the outside flange of the north beam at the end of the load beam connection.

The hinge at the brace center exhibited plastic deformation in the form of cupping of the

underside and bulging of the vertical sides at -1.95% drift. Figure 5.10.22a and Figure 5.10.22b

show the brace buckled shape and the local plastic deformations at the brace center due to the

a) SW Gusset Plate b) NE Gusset Plate

Figure 5.10.21: B2 Gusset Plate Edge Deformation (1.71%)

a) b)

Figure 5.10.20: Y3 Web Yielding and Flange Deformation at SW Column (1.71%)

175

large strains at the hinge. Yielding increased at numerous locations including the southwest

gusset plate, southwest column flanges, northeast beam web and northeast column web.

Cracking within the weld metal initiated at the northeast gusset plate to beam weld propagated to

approximately ½ in. Yielding of the southwest column outside flange is shown in Figure 5.10.23

at this drift of -1.95%.

Figure 5.10.23: Yielding of SW Column Flange (-2.14%)

Also at -1.95% drift, bolt fracture occurred at both the northwest and southeast shear tab

connections. It was determined that the ¾ in. diameter A490 bolts used for the shear tab

connections of Specimen HSS-24 were shorter than of those used in previous tests and the

threads were within the shear plane between the shear tab and the beam web. The design of the

shear tab connections for the specimen calls for the threads of the bolts to be outside the shear

a) Buckled Brace Shape b) BC Local Deformation of Hinge

Figure 5.10.22: Brace Shape and BC Deformation (-2.14%)

176

plane. The early fracture of the bolts was attributed an error in fabrication and not an actual

failure mode resulting from testing. The two bolts from the northwest connection and the single

bolt from the southeast connection were replaced in order to complete the test. Initiation of

micro-cracking was observed at the corners of the brace plastic hinge while in tension at 1.71%

frame drift.

The local deformation of the plastic hinge increased with more severe cupping and bulging

visible at -2.14% drift. Yielding increased at the southwest gusset plate between the brace and

the beam interface weld. Weld cracking at the northeast gusset plate to beam weld propagated to

approximately 1.5 in. Figure

The final successful tension cycle prior to brace failure resulted in 339.7 kip resisted and 1.94%

frame drift. Northeast gusset plate yielding significantly increased to severe levels (Y5) with wide

slashing yielding lines covering the entire plate as shown in Figure 5.10.24. Yielding and

deformation also increased at the northeast column with long yield lines continuing to extend up

the length of the column. Figure 5.10.25a and Figure 5.10.25b show the damage to the inside

column face and over the web.


177

Yielding of the southwest column outside flange also increased to a moderate damage level (Y3)

at 1.94%. Yielding lines have spread over the depth of the outside flange as shown in Figure

5.10.26. Gusset plate edge buckling increased and large rotations were visible at the northwest

and southeast shear tab connections. Yielding also increased over the southwest beam web with

yield lines running perpendicular to and along the interface weld.

Figure 5.10.26: Y3 at SW Column (1.94%)

A frame drift of -2.50% was achieved on the final compression cycle before brace failure. The

buckled brace shape is shown in Figure 5.10.27a and the out-of-plane rotation of the northeast

gusset plate is shown in Figure 5.10.27b. The plastic deformation also increased with severe

cupping and bulging at the hinge.

a) b)

Figure 5.10.25: Increased Yielding at NE Column (1.94%)

178

Yielding increased over the southwest gusset plate at -2.50%. Significant increase in yielding

occurred at the outside flanges of the southwest and northeast columns as shown in Figure

5.10.28a and Figure 5.10.28b, respectively. Crack propagation increased to approximately 2 in. at

the northeast gusset plate to beam interface weld. Figure 5.10.29 shows the crack in the weld

material at this location at this drift.

a) SW Column b) NE Column

Figure 5.10.28: Column Yielding (-2.50%)

a) Buckled Brace Shape

Figure 5.10.27: Brace Shape and Gusset Plate Rotation (-2.50%)

b) NE Gusset Rotation

179

Figure 5.10.29: NE Gusset Plate to Beam Weld Crack (-2.50%)

Brace fracture (BF) occurred at the plastic hinge while heading to but before peak. The

maximum frame drift achieved was 2.07% and the frame resistance at failure was 187.2 kip. Two

additional cycles were completed with a section of the brace removed to obtain the stiffness of

the frame. Bolt fracture again occurred at the shear tab connections.


Specimen HSS-24 was capable of achieving large drift ranges, greater than 4%, and was

controlled by the desired failure mechanism of brace fracture at the plastic hinge. The modified

beam to column connection resulted in uneven distribution of yielding over the gusset plates and

between the framing elements. The reduced stiffness of the shear tab connection between the

beam web and column show limited ability to transfer the vertical shear component of from the

brace. This likely resulted in the more severe yielding to the columns. Figure 5.10.30 shows the

yielding over the northeast column web. The unique pattern of yielding over the gusset plates is

shown in Figure 5.10.31a and Figure 5.10.31a. The typical crescent shape of yielding around the

brace end is not clearly visible and there is another significant arch of yielding adjacent to the

gusset to column interface weld.

180

5.11 HSS-25: Heavy Beam, No Net Section Reinforcing


Specimen HSS-25 was chosen to evaluate a frame with a heavier, stiffer beam section and thick,

stiff gusset plate connections. The design of the specimen is identical to that of HSS-11 with the

exception of the removal of the brace net section reinforcing plates. Braced frame research at

UC Berkeley (Yang and Mahin, 2005) supports the potential for brace fracture at the reduced net

section. This failure mode has not controlled in any of the UW tests since the removal of the net

section reinforcing for Specimen HSS-14 (Koltulka, 2007). The heavier beam and stiff gusset

plate connections are expected to put greater demand on the brace due to in-plane bending.

a) SW Gusset Plate b) NE Gusset Plate

Figure 5.10.31: Gusset Plate Damage (Post Test)

Figure 5.10.30: NW Column Web Damage (Post Test)

181

W16x89 beam sections were used rather than the typical W16x45 of previous tests. The beams

were connected to the W12x72 columns with fully welded web and flanges adjacent to the gusset

plate connections. The gusset plates were 7/8 in. thick and utilized at 4*tg elliptical clearance.

The connection detail for this specimen is shown in Figure 5.11.1.


The gusset plate to frame interface welds from HSS-11 called for CJP welds. The gusset plate

was prepared with a double bevel and back-gouging was required to remove slag prior to welding

the opposite side of the plate. A modified method was used for HSS-25. It was decide to only

fully back-gouge the 8 inches from the gusset free edge and partially back-gouge the balance

because of the difficulty to access the full weld length and the time required to perform the work.

Specimen HSS-25 was tested in the UW Structures Lab on June 5, 2008. The specimen

ultimately failed by brace fracture at the brace center hinge location. The maximum and

minimum frame drift ratios achieved were 0.88% and -2.41% for a total range of 3.30%. The

maximum and minimum forces resisted by the frame were 384.6 and -191.9 kip. The resistance

of Specimen HSS-25 was significantly greater than any of the previous test specimen with the

typical W16x45 beam section. The actual displacement history from the test is shown in Figure

5.11.2 and the hysteretic behavior of the specimen is shown in Figure 5.11.3.

182



Particular observation during the test was paid to brace at the ends of the slot connecting to the

gusset plate. While the frame is pushed to positive drifts and the brace is in tension, in-plane

bending occurs and adds additional stresses at the slotted locations of the brace. Initial cracking

originating at the end of the slot did occur but did propagate. The nomenclature for cracking or

tearing of the steel at the brace net section designated in the Table 5.1.1 is NSD, Net Section

Damage. Fracture of the brace is designated as NSF, Net Section Fracture.

183

Table 5.11.1 summarizes component damage at peak values during the test. It is important to

note the level of damage sustained to the columns which is significantly more than what is

typically observed when the frames consists of the W16x45 beam sections and a thinner gusset

plate configuration. Also, very little damage was observed at the heavier W16x89 beams.



Out-of-plane deformation of the brace while in compression was observed at -0.06% frame drift

and -70.9 kip resisted by the frame. Local buckling (B1) of the inside flanges of the northeast

and southwest columns were observed initially at 0.13% drift where the inside flanges of the

beams connect to the column flanges. The force resisted by the frame was 144.0 kip with the

brace in tension. Initial yielding of northeast and southwest column flanges at the gusset plate

reentrant corners occurred at 0.16% frame drift. Significant rotation of the northwest and

southeast shear tab connection was also observed at this drift level.

The following cycle, the maximum drift ratio actually decreased though the induced displacement

of the actuator increased. This could likely be caused by losses with the test set up such as slip of

the connection between the load beam and north beam, between the support beam and the

south beam, or due to rigid body rotation of the frame as the east column uplifts. The yielding at

the northeast and southwest column flanges at the reentrant corners increased slightly at 0.15%


1 6 -0.06 0.06 0.12 -70.9 76.1

7 8 -0.09 0.10 0.19 -104.3 110.8

9 10 -0.12 0.13 0.24 -131.0 144.0

B1-SWCF,

NECF

11 16 -0.20 0.16 0.36 -165.2 169.6

Y1-SWCF,

NECF

17 18 -0.24 0.15 0.38 -182.0 189.3 Y1-SWGE

19 20 -0.30 0.18 0.48 -191.1 215.3 Y1-NEGE

21 22 -0.38 0.22 0.59 -191.9 237.3 NSD

23 24 -0.53 0.25 0.79 -191.1 271.0 B1

25 26 -0.72 0.31 1.03 -190.2 299.0 B2

Y1-SWGB,

NEGB,

NEGC

27 28 -1.05 0.42 1.47 -188.5 334.5 Y1-SWGC

Y1-NECW,

SWCW

29 30 -1.50 0.54 2.03 -187.2 359.2 Y3 Y3-SWG Y3-NECF

31 32 -1.94 0.68 2.62 -189.9 375.3 BC WD-NEGC

33 34 -2.41 0.88 3.30 -178.0 384.6 Y1-SWBF

Y3-NECW,

SWCF; Y5-

NECF

B2-NECF;

B1-NECW

35 35 - 1.11 - - 242.0 - BF - -

BS-NWB,

SEB -

HSS-25

Drift

Range


Brace Gusset Plates Beams

Initial

Columns

Moderate

Severe

184

frame drift. Initial yielding was also observed on the inside face of southwest column flange

approximately 8 in. north of the beam to column connection.

A clear out-of-plane rotation was observed at both the northeast and southwest gusset plates to

accommodate the elastic buckling of the brace at -0.24% drift. The inside flanges of northeast

and southwest columns could be seen deforming as the frame opened at the minimum drift of

the cycle. Very slight, initial yielding (Y1) also occurred at the brace end of the southwest gusset

plate.

Initial yielding (Y1) of the northeast gusset was observed at -0.30% drift at the brace end. The

slight out-of-plane displacement as the brace elastically buckled is shown in Figure 5.11.4.

Figure 5.11.4: Brace Shape (-0.30%)

At a frame drift of 0.22%, slight initial cracking of the brace net section (NSD) was observed at

the southwest end, north side of the brace. The crack initiated along the edge of the brace to

gusset plate weld and the HAZ at this location. Slight yielding of the northeast and southwest

gusset plates increased at ends of the brace. Yielding increased at the brace ends of both the

northeast and southwest gusset plates at -0.38% frame drift. Yielding also increased at the

reentrant corners on the northeast and southwest column flanges.

Instrumentation indicated that the out-of-plane displacement of the buckling brace reached 3.29

in. exceeding 2% of the total length of the brace (B1), 148.75 in., at -0.42% frame drift and -

190.3 kip resisted by the frame. Figure 5.11.5 shows the shape of the buckled brace and the

extent of the out-of-plane displacement. Yielding of the northeast and southwest gusset plates

significantly increased at the brace ends at -0.53% drift. The flaking has begun to spread from

the brace ends but no clear pattern of yielding was visible at this drift. The level of yielding can

185

be seen in Figure 5.11.6a and Figure 5.11.6b for the northeast gusset plate and southwest gusset

plates, respectively.

Figure 5.11.5: B1 Brace Buckled Shape (-0.42%)

Yielding of the northeast column flange increased at 0.31% and can be seen in the area behind

the instrumentation on the inside flange in Figure 5.11.7. Similar yielding was observed at the

southwest column. Yielding increased on the northeast and southwest column flanges at the

reentrant corner. Figure 5.11.8 shows the extent of the yielding at the northeast column flange.


Figure 5.11.6: Gusset Plate Yielding at Brace Ends (-0.53%)

186

Figure 5.11.7: Yielding of NE Column Flange (0.31%)

Figure 5.11.8: Yielding at Reentrant Corner of NE Column Flange (0.31%)

The out-of-plane displacement at the center of the brace increased to 5.16 in., which is greater

than the depth of the brace (B2) at a frame drift of -0.69%. At -0.72% drift, yielding significantly

increased at the northeast and southwest gusset plates. Initial yielding (Y1) was observed

extending beyond the brace ends and into the areas between the brace and the beam interface

welds for the southwest gusset plate and between the brace and both the beam and column

interfaces welds at the northeast gusset plate. Figure 5.11.9a and Figure 5.11.9b show the

yielding over the northeast and southwest gusset plates, respectively. Initial yielding of the

southwest gusset plate was also observed at the beam reentrant corner.

187


At 0.42% drift and 334.5 kip frame resistance, yielding of the northeast and southwest column

flanges increased at the reentrant corners and in the area approximately 8 in. from the interface

weld end. Figure 5.11.10a and Figure 5.11.10b shows the yielding at reentrant corners for the

northeast and southwest columns, respectively. Slight propagation was observed at the north

side of the southwest brace net section. The crack length on the upper side of the slot has

increased to approximately 3/8 in.

Yielding of both the northeast and southwest gusset plates increased as they rotated out-of-plane

to accommodate the brace buckling at -1.05% drift. The gusset plate yielding began to form the

elliptical shaped crescent as shown in Figure 5.11.11 for the southwest connection, including

a) NE Column b) SW Column

Figure 5.11.10: Increased Yielding at Column Flanges (0.42%)


Figure 5.11.9: Increased Yielding of Gusset Plate (0.69%)

188

initial yielding (Y1) between the brace and the column interface weld. Initial yielding (Y1) was

observed at the northeast and southwest column webs at the inside beam flange of the beam to

column moment connection.

Figure 5.11.11: Increased Yielding at SW Gusset Plate (-1.05%)

The instrumentation monitoring the tensile elongation of the brace indicated that the total brace

elongation of 0.31 in. at 0.54% drift over the original length 119.25 in. from free edge of the

gusset to free edge of gusset exceeded 0.2% (Y3). Yielding of the northeast column flange

adjacent to the interface weld has increased to a moderate damage level (Y3). Figure 5.11.12

shows the yielded flange partially obstructed by the instrumentation of the northeast column at

0.54% frame drift. Increased yielding of the northeast and southwest column flanges at the

reentrant corners was observed as well as yielding extending into the web to both locations.

Yielding also increased over the both gusset plates in the elliptical shape.

Figure 5.11.12: Y3 at NE Column Flange (0.54%)

189

The cracking of the southwest brace net section slightly propagated at 0.54% frame drift to

approximately ½ in. on the upper side and 3/8 in. on the lower side of the gusset plate. The

cracks can be seen extending from the slot edges in Figure 5.11.13.

Figure 5.11.13: Crack Propagation at SW Brace Net Section (0.54%)

Significant increased yielding was observed at the northeast and southwest gusset plates at -

1.50% frame drift. The yielding of the southwest gusset plate as it rotated out-of-plane was

described as moderate (Y3) covering most of the plate area, as shown in Figure 5.11.14. Yielding

increased at the northeast column web and inside flange at the location near the gusset plate

edge. This yielding was evidence of a plastic moment, or hinge, occurring at the column as the

frame opens at this minimum drift. Three distinct hinges have begun to form as the buckled

brace shape looked more linear: one at each gusset plate and one at the brace center. The

buckled brace shape and northeast gusset plate rotation are shown in Figure 5.11.15 and Figure

5.11.15b.

190

Figure 5.11.14: Y3 at SW Gusset Plate (-1.50%)

At 0.68% drift, yielding increased to a moderate damage level over the northeast gusset plate as

shown in Figure 5.11.16. Increased yielding was observed over the webs of the northeast and

southwest columns, as well as at the inside flanges and reentrant corners. The cracking of the

southwest brace end increased length to approximately 9/16 in. at the upper side and ½ in. at the

lower side of the gusset plate slot. Figure 5.11.17 shows the extent of the cracking at the net

section of the brace. Initial cracking (WD) of the modified CJP interface weld was observed as

the northeast column reentrant corner.

a) Buckled Shape b) NE Gusset Plate Rotation

Figure 5.11.15: Buckled Brace Shape (-1.50%)

191


Figure 5.11.17: Crack Propagation at Brace SW Net Section (0.68%)

Local deformation of the plastic hinge at the brace center was observed at -1.94% frame drift in

the form of cupping at the underside and bugling on the vertical sides of the section. Yielding of

the northeast column increased over the inside flange and the web. Deformation increased and

initial yielding over the outside flange opposite the gusset plate edge was observed for the

northeast column at this drift. Northeast column damage is shown in Figure 5.11.18a and Figure

5.11.18b. Increased yielding of the southwest column flange and web was also observed.

192


This positive drift level of 0.88% with 384.6 kip resisted by the frame was the final successful

cycles achieved prior to fracture of the brace center at the hinge location. Significant increase in

yielding and deformation of the northeast column was observed in the hinge area near the gusset

plate edge at 0.88% drift. B2 level local buckling of the column flange can be seen in Figure

5.11.19a as well as increased yielding across the web in Figure 5.11.19b. Yielding of the

southwest column web increased similar to the northeast column.

The crack in the weld material of the northeast gusset plate to column interface weld propagated

to approximately 5/8 in. at 0.88% frame drift. The cracks in the brace at the southwest net

section has also increased. Figure 5.11.20 shows the crack lengths extending approximately 5/8

a) Flange Deformation b) Web Yielding

Figure 5.11.19: NE Column Damage (0.88%)

a) Outside Flange Yielding b) Web Yielding

Figure 5.11.18: NE Column Damage (-1.94%)

193

in. on the upper side and 9/16 in. on the lower side of the gusset plate slot. The slot thickness is 1

in. wide leaving 4 in. of material remaining of the slotted side. The cracks total length of 1 3/16

in. is 30% of the material at the cracked face. Finally at this frame drift, slight initial yielding (Y1)

was observed at the southwest beam flange at the reentrant corner.

Figure 5.11.20: Crack Propagation at Brace SW Net Section (0.88%)

The minimum frame drift achieved prior to brace fracture the following cycle was -2.41%.

Figure 5.11.21a shows the out-of-plane deformation of the buckled brace while Figure 5.11.21b

depicts the level of local deformation at the brace plastic hinge.

Moderate yielding (Y3) of the northeast gusset plate was observed with yielding extending over

the majority of the plate surface. Figure 5.11.22 shows the condition of the northeast gusset at -

a) Buckled Brace Shape b) Brace Plastic Hinge

Figure 5.11.21: Brace Condition (-2.41%)

194

2.41% drift. Yielding continued to increase at the northeast and southwest columns at the hinge

areas at -2.41% frame drift. Increased yielding was observed on the outside flange of both the

northeast and southwest columns as shown in Figure 5.11.23a and Figure 5.11.23b, respectively.

Severe yielding (Y5) of the inside flange and increased deformation of both flanges of the

northeast column can be seen in Figure 5.11.24a and Figure 5.11.24b. The northeast column

web also exhibited local buckling (B1) in the form of bulging in the area adjacent to the edge of

the gusset plate.

Figure 5.11.22: Y3 at NE Gusset Plate (-2.41%)

a) NE Column b) SW Column

Figure 5.11.23: Yielding at Column Outside Flanges (-2.41%)

195

Brace fracture occurred while heading to peak frame drift with 242.0 kip of resistance.

Simultaneously, all four bolts at the northwest shear tab connection and one bolt at the southeast

connection shear tab. It is difficult to determine if the failure of the shear tab connections was a

result of the load transfer immediately after the brace fractured or if it was an accurate

representation of what could happen in the field in a severe seismic event. The northwest and

southeast shear tab connections are shown in Figure 5.11.25a and Figure 5.11.25b after failure,

respectively.


The most important aspect of this test was in evaluating the potential for net section failure.

Even though the controlling failure mechanism was brace fracture at the plastic hinge, the test

a) NW Connection b) SE Connection

Figure 5.11.25: Shear Tab Bolt Fracture (Post Test)

a) Y5 at Inside Column b) Column Flange Deformation

Figure 5.11.24: NE Column Damage (-2.41%)

196

did show the possibility for fracture at the net section by exhibiting tearing at relatively moderate

drifts with the brace in tension. The cracks did not propagate to fracture but the effect of the

heavier beam section and the thicker, stiff gusset plate does show greater potential for net

section failure. Specimen HSS-26 looks further into this potential for net section failure with a

specimen with the heavier beam section and a thick, stiff gusset plate.

5.12 HSS-26: Heavy Beam, Near-Fault Drift History


Specimen HSS-26 is identical to specimen HSS-25. The specimen utilizes a heavier beam section

and a thick gusset plate. Figure 5.12.1 shows the connection detail for the specimen. Again, the

brace does not include net section reinforcing at the brace to gusset plate connection. This test

was chosen to evaluate the specimen while subjected to an alternate near-fault imposed drift

history. A study experimentally evaluated the affect of different drift histories and effective net

section area on the potential for net section fracture showed that the requirement for net section

reinforcing is more critical for tension near-fault drift histories (Yang and Mahin, 2005). HSS-26

is the first of the UW SCBF tests that implemented an alternative near-fault drift history.


The design of the tension near-fault drift history was based on the SAC recommended near-fault

drift history for Special Moment Frames outlined in FEMA 355D and the experimental near-

fault drift history used by UC Berkeley in the tests mentioned above. No established near-fault

197

history existed for SCBFs so the amplitudes were chosen based on expected compressive and

tensile yielding of the brace as seen in HSS-25. Figure 5.12.2 shows the designed imposed near-

fault drift history for HSS-26. The intent was to modify the input LVDT displacement during

the test based on the actual frame drift calculated by the frame diagonal displacement to account

for the losses with in the test set up and to more accurately obtain the desired story drift.

Figure 5.12.2: Designed Near-Fault Drift History

The specimen ultimately failed by net section fracture while pushing to the largest positive story

drift with the brace in tension. The fracture occurred at the southwest end of the brace with

initial tearing originating at the north side of the brace at the end of the gusset plate slot. The

maximum and minimum drift ratios achieved were -0.40% and 0.99% for a total range of 1.39%.

The maximum and minimum lateral forces resisted by the frame were -197.0 kip and 392.6 kip.

The total range was 589.6 kip total. This test was started with a small displacement test cycle to

verify instrumentation. The actual displacement history of HSS-26 is show in Figure 5.12.3.

Figure 5.12.4 show the hysteretic behavior the specimen.

HSS26 Near Fault Loading History

-1.0

-0.5

0.0

0.5

1.0

1.5

2.0

2.5

3.0

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16

Cycle

Dri

ft,

%

-0.787%

-1.206"

2.121%

3.251"

2.536%

3.888"

1.705%

2.614"

0.875%

1.341"

0.459%

0.704"

1.290%

1.978"

0.875%

1.341"

-0.984%

-1.508"

-0.044%

-0.068"-0.371%

-0.569"

2.536%

3.888"

-0.787%

-1.206"

Drift, %

Displacement, inches

198



The actual testing of Specimen HSS-26 was completed very quickly because of the early fracture

of the brace. Photos have consistently been a valuable method for documenting the progression

of damage to components at increasing story drifts for previous test. Unfortunately, limited

photos were taken at specific points in the test and it is difficult to determine when exactly

yielding initial occurred because of the nature of this test protocol. Peak values drift and lateral

force and corresponding component damage levels are summarized in Table 5.12.1.

199



The instrumentation monitoring the out-of-plane displacement of the brace center showed that

the displacement of the buckled brace reached 2% (B1) of the total brace length, 148.75 in., at -

0.33% drift and -193.6 kip resisted by the frame. Initial yielding (Y1) of the southwest column

flange and web adjacent to the gusset plate occurred at -0.40% drift. The lateral force resisted by

the frame with the brace in compression was -197.0 kip. Initial yielding (Y1) of the northeast

column flange at the reentrant corner was observed at the same drift ratio. Slight yielding was

also observed at the northeast column flange at the beam to column connection, in the location

of the south flange. The maximum out-of-plane displacement of the brace center at -0.40% was

3.70 in. The buckled brace is shown in Figure 5.12.5.

Figure 5.12.5: Buckled Brace Shape (-0.40%)

Initial yielding (Y1) of the southwest column flange web while the brace was in tension was

observed at 0.15% and 211.3 kip resisted by the frame. The brace elongation instrumentation

indicated that the total elongation of the brace, from end of gusset plate to end of gusset plate, at

0.36% and 327.2 kip frame resistance reached 0.2%, signifying moderate yielding (Y3).


-0.05 0.05 0.10 -72.8 68.6

-0.33 0.15 0.47 -193.6 211.3 B1 Y1-SWCW

-0.33 0.36 0.69 -193.6 327.2 Y3

-0.40 0.74 1.14 -197.0 385.3

Y1-SWNS,

NENS; NSD-

SWNS

Y1-NECF,

SWCF,

SWCW

Y3-NECF,

NECW; B1-

NECF

Moderate 2 2 -0.40 0.99 1.38 -197.0 392.6 NSF Y1-SWGCY3-SWCF;

B1-SWCF

Initial

HSS-26

Drift

Range


Brace Gusset Plates Beams Columns

1 1

200

The test was paused at 0.74% drift with 385.3 kip resisted by the frame to evaluate damage at an

intermediate point before reaching the desired maximum drift ratio. Slight concentration of

yielding (Y1) of brace at the northeast and southwest net sections were observed around the hole

at the end of the slot. It was also noted that a very slight tearing (NSD) of the brace steel at the

southwest net section was observed. Figure 5.12.6a and Figure 5.12.6b show the yielding at the

northeast and southwest net section locations, respectively.

Yielding of the northeast column flange and web significantly increase to a moderate damage

levels (Y3). Figure 5.12.7a and Figure 5.12.7b show the extent of the yielding for each at 0.74%

drift. Slight local buckling could be observed at the northeast column flange in the same area.

Yielding of the southwest column flange also increased.

a) Column Flange b) Column Web

Figure 5.12.7: Y3 at NE Column (0.74%)

a) Northeast b) Southwest

Figure 5.12.6: Yielding at Brace Net Section Locations (0.74%)

201


Fracture of the southwest net section (NSF) occurred at 0.99% drift and 392.6 kip frame

resistance. Yielding around the area that the fracture initiated could be seen clearly. Figure

5.12.8a and Figure 5.12.8b show the fractures net section of the brace. The only gusset plate

yielding (Y1) observed during this test occurred at the southeast gusset plate at the column

reentrant corner observed after the brace fractured.

Damage to the southwest column increased significantly. Moderate yielding (Y3) was observed

over the southwest column flange and yielding increase through the web. Local deformation of

the flange was clearly exhibited adjacent to the edge of the gusset plate. Figure 5.12.9a and

Figure 5.12.9b show the damage to the southwest column. Yielding and location deformation

also increase at the northeast column flange and web. This is shown in Figure 5.12.10a and

Figure 5.12.10b below.

a) b)

Figure 5.12.8: NSF at SW Brace End (0.99%)

202


Specimen HSS-26 proved susceptible to the undesirable failure mechanics of net section fracture

with the elimination of the net section reinforcement under at near-fault drift history. The heavy

beam and thick gusset also play a part in increasing the specimen likeliness to net section

fracture. The majority of strain elongation across the diagonal must occur in the brace rather

than being spread thru the brace, gusset plates and framing elements.

This test showed the difficulties in implementing an appropriate near-fault drift history to a full

scale SCBF test specimen. The SAC recommended near-fault drift history is intended for a

much more flexible system, Special Moment Frames, and accurately scaling that first maximum

drift amplitude needs further review. The point at which the fracture occurred was in the same

a) Flange Deformation a) Flange and Web Yielding

Figure 5.12.10: NE Column Damage (0.99%)

a) Flange Deformation b) Y3 Yielding at Flange and Web

Figure 5.12.9: SW Column Damage (0.99%)

203

range as maximum positive drift value achieved before fracture of the brace at the plastic hinge

in HSS-25 with the normal increasing amplitude cyclic loading protocol, 0.88% compared to

0.99%. The maximum frame resistance with the brace in tension was also similar, 384.6 kip for

HSS-25 and 392.6 kip for HSS-26.

204

Chapter 6: Data Analysis

6.1 Introduction

The data analyses used to evaluate the global and local responses of the nine specimens within this

test program are provided in this chapter. Descriptions of the analysis methods and the results are

provided including comparisons to illustrate a specific behavior or response. Plots of the analysis

results that are not included in this chapter are compiled in Appendix C.

Analysis of the system response was evaluated in Section 6.2 including evaluation of the lateral load

verses story drift hysteretic response and nonlinear system stiffness. The response over the brace

diagonal, including the brace and gusset plates was analyzed and discussed in Section 6.3. Analysis

of the frame response was presented in Section 6.4. The distribution of system resistance between

the brace and the framing elements was evaluated in Section 6.5, which also assessed of the analysis

methods for determining forces in the brace and the frame. The energy dissipated by each

specimen was determined in Section 6.6. Lastly, Section 6.7 compares the damage levels observed

during the test with regards to performance level criteria within Performance Based Design

Methods.

Extensive calculations and data interpretation of the test data was necessary for creating the

comparison plots. The methods and equations used for calculating the variables for each

comparison are outlined in Appendix D. A combination of software including Matlab Version

R2007a and Microsoft Office Excel were used to analyze the data and to create comparison plots.

In cases where instrumentation was damaged or electronic error during recording occurred, the

data was determined erroneous and unusable for the comparison plots. Corrections where also

made for any adjustments or unintended movement of instrumentation during testing such as re-

positioning of instruments or bumping string potentiometer wires.

6.2 System Response

This section evaluates the system response of the nine specimens within this test program. The

two factors that best signify the performance of the system are ultimate drift capacity and lateral

load resistance. Comparisons are presented to evaluate the ductility and load resistance of the

specimens to evaluate the effect of the test parameters on the system performance. Sections are

205

provided to evaluate and discuss lateral force verses drift hysteresis and the elastic and inelastic

stiffness based on backbone curves from the enveloped hysteretic response.

The ultimate drift capacity of the specimen is the clearest indicator of system ductility and the

ability of the lateral bracing system to withstand extreme seismic demands. The dynamic response

to seismic loading inherently induces large story drifts during a severe seismic event. The capability

of a specimen to maintain lateral load resistance while achieving large drift levels shows superior

performance.

The strengths of the specimens are evaluated to determine how each test parameter affects the

lateral resistance of the system. The single diagonal configuration of these test specimens exhibits

an unsymmetrical response and has different inelastic resistance with the brace in tension and in

compression. For an actual SCBF system, brace directions are offset to provide equal resistance in

both directions. The lateral resistance is normalized by the lateral load associated with brace tensile

yielding, , neglecting resistance of the frame. The load is calculated using the nominal brace yield

force, , and the brace angle, , as shown in Equation (1.2.1).

(6.2.1)

The elastic stiffness is inherently important for the design of an SCBF when considering moderate,

more frequently experienced earthquakes, and performance based design criteria for immediate

occupancy. Once the brace buckles in compression and yields in tension, the nonlinear behavior

affects energy dissipation and the ultimate resistance of the system. The inelastic post buckling and

tensile yielding performance are critical to the response of the system for more severe and less

frequent seismic events. The ability of the system to effectively dissipate energy and maintain

resistance at larger drift ranges improves the seismic performance and increases the performance

based design levels of life safety and collapse prevention for a maximum credible design

earthquake.

A global performance summary for the nine test specimens is presented in Table 6.2.1. The

ultimate drift capacity and total range, as well as the ultimate positive and negative resistances and

total drift range are summarized. Description of system failure mechanisms and the relevant test

parameters for each specimen are also included.

206

Table 6.2.1: Global Performance Summary

When discussing ultimate drift capacity, total drift range is normally used because it is

representative of the complete response of the system in both directions. WF-23 with the W6x25

wide-flange brace section reached the highest total drift range of 5.56% and drift ratios in both

directions of the nine specimens in this test program before failure occurred. The ultimate failure

was fracture of the northeast gusset plate interface welds. Of the specimens with HSS5x5x3/8

brace sections, HSS-24 with the bolted shear plate welded flange beam-to-column connection and

thin rectangular gusset plates reached the highest total drift range, 4.44%. HSS-18, HSS-21 and

HSS-20 utilized thin rectangular gussets but different beam-to-column connections and reached

4.19%, 4.14%, and 3.97% total drift ranges, respectively. HSS-22 utilized a bolted shear plate

beam-to-column connection and thin tapered gusset plates, and achieved 3.98% total drift range.

HSS-25 incorporated a heavier beam section and a thick rectangular gusset without net section

reinforcement and only achieved 3.30% with brace fracture as the ultimate failure. HSS-26 was an

identical design as HSS-25 but subjected to a tension dominated near-fault loading protocol which

resulted in a total drift range of 1.38% and fracture at the brace net section. Lastly, HSS-19 only

achieved 1.31% total drift due to an undesired failure mechanism controlling and fracture occurring

at the brace extension plate early in the test.

Expectedly, specimens HSS-25 and HSS-26 exhibited the greatest resistance because of the

increased frame resistance from the larger beam sections for the positive direction and stiffer gusset

Range Min Max

HSS-18 4.19 -2.59 1.60 -0.76 1.59Brace

Fracture

Bolted Shear Plate

Connection

HSS-19 1.31 -1.01 0.31 -0.51 1.12Brace

Fracture

Bolted WT Brace

Connection

HSS-20 3.97 -2.28 1.69 -0.93 1.59Brace

Fracture

18 Bolted Beam End

Plate Conn.

HSS-21 4.14 -2.55 1.59 -0.81 1.73Brace

Fracture

14 Bolted Beam End

Plate Conn.

HSS-22 3.98 -2.48 1.50 -0.66 1.50Brace

Fracture

Bolted Shear PL. and

Tapered GP

WF-23 5.56 -3.21 2.35 -0.58 1.30

Interface

Weld

Fracture

WF brace section

HSS-24 4.44 -2.50 1.94 -0.80 1.69Brace

Fracture

Bolted Web Weld

Flange Beam/Col Conn.

HSS-25 3.30 -2.41 0.88 -0.95 1.91Brace

Fracture

Heavy Beam and

Gusset Plate

HSS-26 1.38 -0.40 0.99 -0.98 1.95

Brace Net

Section

Fracture

Heavy Bm and GP with

Near-Fault Loading

ResistanceDrift Ratio, % Failure

MechanismDiscription

P/Py

207

plate connections increase the buckling capacity in the negative direction. HSS-22 showed the

lowest resistance for specimens with the HSS tubular braces, which is also expected considering the

simple shear tab beam-to-column connection and the thin tapered gusset plates providing more

flexible end supports for the brace in compression.

6.2.1 Force vs. Drift Response

The lateral force verse drift ratio hysteresis behavior is commonly used as a basis for evaluating

overall performance. The ultimate drift capacity, ultimate resistance and initiation of yielding can be

extracted from the plots. Hysteretic plots also indicate the level of ductility within the system and

the energy dissipation. Large, wide loops suggest greater energy dissipation, damping, and

distribution of yielding, while tight loops or pinching indicate reduced dissipation of energy.

The hysteretic plots are provided in Table 6.2.2 and are arranged with a detail of each gusset plate

design. The load cell in the actuator was used to record the lateral load applied the specimens and

the drift ratio is taken as the horizontal frame drift divided by the story height in percent. The

horizontal frame drift is calculated from the diagonal elongation from the southwest to the

northeast work points, divided by the cosine of the original brace angle, 45o. This method was

chosen to determine the frame drift rather than instrumentation measuring the relative horizontal

displacement at the northeast corner from a fixed point on the strong floor, because it is

independent of horizontal frame slip at the base connection and rigid body rotation of the frame,

which was significant at larger drifts and required multiple corrections during analysis.

Table 6.2.2: Hysteresis Comparison

HS

S-1

8

Bolted Shear Plate Beam-to-

Column Connection

3/8” Thick Rectangular Gusset

Plate with 7*tg Elliptical

Clearance

208

HS

S-1

9

WT Bolted Brace to Gusset

Plate Connection

½” Rectangular Gusset Plate

with 7*tg Elliptical Clearance

CJP Welded Beam-to-Column

Connection

HS

S-2

0

Bolted Beam End Plate

Connection to Column

18 Bolt Configuration

3/8” Rectangular Gusset Plate


HS

S-2

1

Bolted Beam End Plate

Connection to Column

14 Bolt Configuration



HS

S-2

2

Bolted Shear Plate Beam-to-

Column Connection

3/8” Thick Tapered Gusset

Plates with 7*tg Elliptical

Clearance

209

WF

-23

W6x25 Wide-Flange Brace

Section



CJP Welded Beam-to-Column

Connection

HS

S-2

4

Bolted Shear Plate CJP Welded

Flange Beam-to-Column

Connection



HS

S-2

5

W16x89 Beam Section CJP

Welded to Column



No Net Section Reinforcing

Typical Loading Protocol

HS

S-2

6

W16x89 Beam Section CJP

Welded to Column



No Net Section Reinforcing

Near-Fault Loading Protocol

6.2.2 System Stiffness

This section evaluates the initial and nonlinear stiffness for the nine test specimens. Envelope

curves, or backbone curves, were created using the values from lateral load and drift ratio for each

210

specimen. Figure 6.2.1 shows an example of a backbone curve for a SCBF with a single diagonal

brace, and it illustrates the key values are used to determine the initial elastic stiffness of the system,

, and the post-yield tangent stiffness after tensile yielding or brace buckling occurs, and ,

respectively.

Figure 6.2.1: Backbone Curve Description

The elastic stiffness and the inelastic tangent stiffness in tension and compression were calculated

and summarized in Table 6.2.3 for each specimen. Resistances at yield and buckling, and ,

and the drift ratios at which these occurred are also included in the summary.

KPt

KE

KPc

RYt

RYc

ΔYc ΔYt

211

Table 6.2.3: Elastic and Tangent Stiffness Summary

Regression analysis was used to determine the elastic stiffness from initial small cycles of the testing

protocol by fitting a linear equation over the enveloped load verses drift response. The slope of the

line is equivalent to the stiffness of the system. The same method was used to determine the post-

yield and post buckling tangent stiffness based on the load verse drift response beyond yield. A

leveling or decrease in compressive resistance and then the obvious transition to the nonlinear

response marks the occurrence of buckling. The buckling resistance and the drift ratio at buckling

are taken as the maximum resistance in the transition region and the associated drift from the

backbone curve. The transition between the elastic and nonlinear responses is less clearly defined

with the brace in tension, since no definitive point of yielding is observed as when the brace

buckles in compression. The intersection of the elastic stiffness and the post-yield tangent stiffness

was taken as the drift at yield for comparison, as shown in Figure 6.2.1. The lateral resistance in the

system at this drift is considered the yield resistance. The backbone curves for the nine test

specimens are shown in Table 6.2.4. Elastic stiffness, post-yield and post-buckling tangent

stiffness, and the resistance and drift ratio at brace buckling and yielding are also included for each

specimen.

Post

Yield

Stiffness,

KPt (k/in)

Post

Buckling

Stiffness,

KPc (k/in)

HSS-18 614.5 24.8 0.7 221.0 0.315 -151.9 -0.235

HSS-19 739.0 N/A -1.6 224.3 0.276 -92.5 -0.092

HSS-20 664.2 11.0 2.6 186.4 0.350 -181.2 -0.134

HSS-21 604.5 19.8 1.4 262.1 0.362 -155.1 -0.338

HSS-22 581.4 10.3 -1.0 206.5 0.351 -134.5 -0.224

WF-23 740.2 19.4 9.9 225.0 0.274 -128.5 -0.129

HSS-24 682.8 10.9 3.8 252.4 0.328 -144.9 -0.205

HSS-25 808.4 72.3 -6.2 291.8 0.294 -190.2 -0.292

HSS-26 776.3 42.0 N/A 295.7 0.296 -168.1 -0.167

Δyc (%)Specimen

Elastic

Stiffness,

KE (k/in)

Tangent Stiffness

Ryt (kips) Δyt (%) Ryc (kips)

212

Table 6.2.4: Backbone Curve Comparison H

SS

-18

HS

S-1

9

HS

S-2

0

HS

S-2

1

HS

S-2

2

WF

-23

HS

S-2

4

HS

S-2

5

HS

S-2

6

213

The backbone curves give an interesting global comparison for considering how gusset plate

geometry and beam-to-column connection type affects the elastic and post-yield and buckling

tangent stiffness of the system. Specimens HSS-25 and HSS-26 showed the largest elastic stiffness,

808.4 k/in and 776.3 k/in respectively. Both specimens had heavier beam sections and the larger

geometries with the 7/8” thick gusset plates. At positive drifts, these specimens achieved the

largest resistance at initial yield. At negative drifts, HSS-25 showed the greatest compressive

buckling resistance, -190.2 kips at -0.29% drift, because of the increased rigidity of the gusset plate

connections and subsequent affect on the brace slenderness ratio. The post-buckling tangent

stiffness of -6.2 k/in for HSS-25 was the lowest of all the specimens.

HSS-22 had the lowest elastic stiffness, 581.4 k/in, because of the bolted shear plate beam-to-

column connection and the thin tapered gusset plates. This also resulted in a lower buckling

resistance because of the increased flexibility at the brace ends, -134.5 kips at -0.22% drift. The

post-yield tangent stiffness, 2.0 k/in, was lower than specimens with rectangular gusset plates and

the post-buckling tangent stiffness, -1.0 k/in, showed a loss of resistance over the plastic region.

WF-23 showed a larger elastic stiffness by comparison, 740.2 k/in, because of the increased brace

area of the W6x25. Buckling occurred at a resistance of -128.5 kip and -0.13% drift. The specimen

continued to increase resistance post yield in the positive direction and showed a post-yield

stiffness of 16.2 k/in while achieving the largest positive drift levels. After an immediately loss in

resistance after buckling, WF-23 exhibited a positive stiffness of 9.9 k/in at increasing negative

drifts.

HSS-18, HSS-20, HSS-21 and HSS-24 all had identical gusset plate designs but varied beam-to-

column connections. Similar elastic stiffness was seen for all four specimens ranging between 604.5

k/in for HSS-21 and 682.8 k/in for HSS-24. HSS-20, with the 18 bolt beam endplate connection,

had a larger stiffness than HSS-21 with the 14 bolt configuration, of 664.5 k/in. HSS-18 showed

an elastic resistance of 614.5 k/in with the bolted shear plate beam-to-column connection. Post-

yield tangent stiffness ranged between 3.3 k/in for HSS-20 and 15.3 k/in for HSS-20. Similar post-

buckling tangent stiffness was shown for the four specimens between 0.7 k/in for HSS-18 and 3.8

k/in for HSS-24.

The backbone curves for HSS-18, HSS-20, HSS-21 and HSS-22 show small changes in the stiffness

prior to a clear transition to the inelastic response in the positive direction. It is difficult to

determine what behavior attributes to this small change in stiffness. It also makes the

214

determination of initial tensile yielding someone subjective. The values tabulated above are for

comparison purposes and do not reflect the actual drift level or lateral resistance associated with

tensile yielding of the brace. Brace yielding is discussed in greater detail in a later section.

Comparing the elastic stiffness for HSS-25 and HSS-22, a specimen with heavier beam fully welded

and thick rectangular gusset is 28.1% stiffer than a specimen with the bolted shear plate beam with

thin tapered gusset plates. Specimens HSS-18 and HSS-22 both shared the same beam-to-column

detail but the rectangular geometry exhibited a 5.7% increase in elastic stiffness.

6.3 Brace and Gusset Plate Response

The behavior of the diagonal brace is the primary inelastic response component in an SCBF

system. The ability for the brace to freely buckle, elongate in tension, and cycle between inelastic

post-buckling and tensile yield deformation will determine the cyclic response of the frame. When

the brace fractures, the lateral resistance of the system is diminished, and this is viewed as the

controlling failure mode of the system. The behavior of the brace beyond the elastic stages under

cyclic loading is critical to the overall performance of the SCBF.

This section compares the performance of components over the full brace diagonal including the

gusset plates. The inelastic brace behavior is summarized below in Figure 6.3.1. Each of the

specimens followed this model with the exception of HSS-19 which buckled and eventually

fractured the splice plate to the connection. HSS-26 was subjected to an alternate tension

dominated near-fault loading protocol which did not induce negative drift required to fully develop

buckling in the brace.

215

Figure 6.3.1: Idealized Inelastic Brace Behavior Under Cyclic Loading (Kotulka 2007)

Stage 1 shows the buckled shape of the brace as it deflects due to initial imperfections and reaches

the critical buckling load. Curvature increases as load and deflection increase eventually forming a

plastic hinge at the brace center, as shown in Stage 2. Stage 3 shows how the shape of the buckled

brace has changed and strain is more concentrated at the hinge region after significant yielding

occurs. As the loading returns to zero, residual strain at the plastic hinge and residual out-of-plane

displacement of the brace are evident. As the brace is pulled in tension, the hinge behavior is

reversed increasing strain accumulation at the plastic hinge, shown as Stage 4. At Stage 5, the brace

is yielding in tension. Tensile strain is not evenly distributed over the length but more likely

concentrated at the plastic hinge region. Finally, fracture of brace section typically occurs at the

plastic hinge while in tension.

The concentration of strain at the center plastic hinge appears to be the factor which ultimately

leads to cracking and then fracture of the brace. One of the objectives within this research is to

delay the onset of these large strains by developing gusset plate connection designs that reduce the

local strain concentration and extend of the life cycle of the brace.

6.3.1 Brace Response

The brace forces were determined using strain gauge data recorded during the tests. Figure 6.3.2

shows the placement of strain gauges for the typical HSS5x5x3/8 and W6x25 brace cross sections.

The exact locations of the strain gauges along the length of the brace are shown in Appendix B,

and are approximately at the southwest quarter-point of the brace length. This section first

216

evaluates the buckling capacity of the braces the experimental buckling brace force to the nominal

buckling force based on AISC specifications. Secondly, the complete inelastic response is

determined using analysis methods that account for the nonlinear behavior of the brace material to

evaluate brace response beyond yielding.

Figure 6.3.2: Brace Strain Gauge Locations

6.3.2 Brace Buckling Capacity

The loading protocol used in testing the specimen begins with small induced drift levels which

allow the specimens response to remain elastic. The brace strain gauge records can be converted to

brace force prior to yielding. After yielding, a more sophisticated analysis method is required to

convert strain to stress.

(6.3.1)

The critical buckling load determined experimentally is compared to the nominal buckling capacity

of the brace as calculated per AISC Design Specifications. The nominal buckling capacity is

calculated using a K factor of 1.0 and the actual brace length. The values are summarized in Table

6.3.1. In addition, the effective length coefficient, K, based on the experimental buckling force, ,

and the resulting effective slenderness ratio, , using the actual brace length, l, was evaluated for

each specimen and provided in the table.

The AISC Seismic Provisions limits the effective slenderness ratio to which is 100.43 for

A500 B/C ( ) HSS steel and 96.33 for A992 wide-flange sections ( ). This

4

3

2

11 2

4 3

= strain guage

W6x25HSS5x5x3/8

217

gives a comparison of the gusset plate connection flexibility. A more flexible connection results in

values closer the effective length factor for pinned-pinned end supports, 1.0, while more

rotationally stiff gusset connections produce values closer to the theoretical fixed end condition,

0.50.

Table 6.3.1: Experimental vs. AISC Brace Forces

Observations from the table are as follows:

HSS-19 brace buckled at the extension plate prior to reaching the critical buckling force of

the brace. All of the other specimens buckled over the length of the brace at or below the

calculated nominal buckling capacity which experimentally supports the AISC equations

for calculating nominal buckling capacities of compression members.

The thin tapered gusset plates for HSS-22 resulted in a lower experimental buckling force

and calculated effective length factor, , equal to 1.0.

As expected, the increase stiffness from the 7/8” thick gusset plates for HSS-25 and HSS-

26 increased the experimental critical buckling force and resulted in a equal to 0.62. Here

the rigidity of the gusset plates increases the critical buckling load and using a equal to 1.0

and the actual brace length underestimates the nominal buckling capacity by 30%.

All of the actual slenderness ratios for specimens with HSS tubular brace sections are far

below the limit established within the AISC Seismic Provisions. However, WF-23 with the

wide-flange brace section did exceed the limit but this was anticipated in the design of the

specimen and deemed acceptable in order to maximize the brace section size within limits

of the UW test setup and actuator capacity.

Experimental AISC

Pcr Pn=FcrAg

HSS-18 188.6 176.2 1.07 0.96 81.3

HSS-19 115.9 179.9 0.64 N/A N/A

HSS-20 199.9 176.2 1.13 0.86 72.4

HSS-21 182.7 176.2 1.04 0.96 81.1

HSS-22 176.3 176.7 1.00 1.00 84.3

WF-23 167.0 165.4 1.01 0.99 103.8

HSS-24 206.7 176.2 1.17 0.82 68.8

HSS-25 241.4 185.7 1.30 0.62 49.3

HSS-26 241.1 185.7 1.30 0.62 49.5

Specimen

Experimental

Effective Length

Factor, K

Experimental

Slenderness

Ratio (Kl/r)

Exp/AISC

Pcr/Pn

218

6.3.2.1 Nonlinear Brace Response

Nonlinear brace responses were determined by evaluating peak values and brace hysteresis behavior

over the full test. The hysteretic response of the brace uses actual calculated brace forces and

displacement over the diagonal work point to work point. Strain gauges were placed around the

brace section at the southwest quarter point to record the strains during each test. These strains

were converted to stress using a plasticity model created using MatLab. The elastic-plastic stress-

strain relationship illustrated in Figure 6.3.3 defines the nonlinear behavior of the A500 B/C steel.

The model addresses large cyclic strains by calculating the plastic strain accumulated prior to

reversal for each cycle. The model calculates stress equal to in the elastic range until yielding

occurred, where it was taken as the input yield stress until the reversal point. Unloading and

reloading stiffness follows the modulus of elasticity, , equal to 29000ksi. The material properties

of the A500 B/C tubular braces sections and the A992 W6x25 brace section used in WF-23 were

not attained from coupon test. The yield stress in the model was taken as RyFy from the AISC

Seismic Provision, 55ksi and 64.4ksi, for the respectively. Figure 6.3.4 shows a theoretical strain

record under cyclic loading and the resulting stress-strain response from the plasticity model.

Figure 6.3.3: Elastic-Plastic Stress/Strain Behavior

Fy

E=29000ksi

219

Figure 6.3.4: Plasticity Model under Repeat Cyclic Loading

Brace hysteresis plots for each specimen are shown in Table 6.3.2. The plots were created using

the brace forces from the strain gauge data verses the total displacement over the frame diagonal.

Ideally, brace elongation would be a more telling variable for the brace behavior but the

instrumentation capturing the elongation was affected by the out-of-plane rotation at the gusset

plates at negative drifts. A simplified approach was used to correct for this effect but the resulting

values were deemed unreliable for this comparison. The peak brace force values normalized by the

brace yield force, , and the maximum elongation of the frame diagonal as a

percent of the total length, work-point to work-point, are summarized in Table 6.3.3. Brace

elongation over the actual brace lengths is also included for positive drifts only.

Table 6.3.2: Brace Hysteresis Comparison

HS

S-1

8

HS

S-1

9

Fy

2.0εy

6.0εy

10.0εy

8.0εy

4.0εy

-1.0εy

1.0εy

2.0εy

-Fy

220

HS

S-2

0

HS

S-2

1

HS

S-2

2

W

F-2

3

HS

S-2

4

HS

S-2

5

HS

S-2

6

221

Table 6.3.3: Brace Resistance Summary

For all of the specimens with HSS brace sections, the braces reached their maximum resistance in

tension and then maintained strength or degraded slightly as positive diagonal deformation

increased. The force in WF-23 continues to increase with positive deformation and reached the

maximum resistance immediately prior to failure of the system. In compression, the largest brace

force was recorded immediately prior to buckling and resistance decreased as negative deformation

increased because of P*Δ effects. The wide-flange brace section in WF-23 quickly loses resistance

after buckling while the specimens with HSS brace sections appear to degrade at a slower rate. The

drop in compressive resistance appears more severe for HSS-22 than HSS-25 which would suggest

that the fixity of the gusset plate connections contributes to this behavior.

6.3.3 OOP Response of Brace and Gusset Plates

This section compares the behavior of the brace and gusset plates in compression and their effect

on the ultimate drift capacity of the system. Out-of-plane displacement of the brace is evaluated in

Section 6.3.3.1. The brace is forced geometrically to buckle out-of-plane as the frame is pushed to

negative drifts. The level of out-of-plane displacement should be empirically similar at a given total

drift range regardless of gusset plate size and thickness. The shape and the curvature of the

buckled brace are greatly affected by the gusset plate rotational stiffness. Increased curvature of the

brace leads to strain accumulation at the brace center and fracture at smaller frame deflections

(Kotulka 2007).

Gusset plate rotations are discussed and compared in Section 6.3.3.2. A more flexible gusset plate

will allow the brace to buckle similarly to a pinned-end member. Stiffer gusset plate connections

result in the double curvature typical of a fixed-end member, as shown in Figure 6.3.5. The

curvature at the brace center is less severe at a given out-of-plane displacement for a brace with

more flexible connections. Comparisons of buckled brace shape are compared in Section 6.3.3.3.

Pmax Tmax Compression Tension Compression Tension

HSS-18 -188.6 398.0 -1.29 0.81 N/A 0.78

HSS-19 -117.1 238.2 -0.50 0.15 N/A 0.17

HSS-20 -198.2 398.0 -1.14 0.85 N/A 0.94

HSS-21 -175.0 398.0 -1.28 0.80 N/A 0.90

HSS-22 -174.0 389.7 -1.31 0.79 N/A 0.84

WF-23 -167.0 351.5 -1.71 1.26 N/A 1.65

HSS-24 -175.0 391.6 -1.26 0.97 N/A 1.04

HSS-25 -237.4 398.0 -1.21 0.45 N/A 0.69

HSS-26 -218.3 398.0 -0.22 0.49 N/A 0.70

Brace Elongation,%Specimen

Experimental Brace

Force, kipsFrame Diagonal Elongation,%

222

Figure 6.3.5: Shape and Curvature of Buckled Brace with Different End Conditions

Damage to the interface weld connecting the gusset plates to the framing elements typically initiates

while the brace is in compression. Section 6.3.3.4 discusses weld damage and compares cracking as

observed during the tests.

6.3.3.1 Brace Out-of-Plane Displacement

The maximum out-of-plane displacements at the brace center are determined and discussed in this

section. The instruments used to measure out-of-plane displacement of the brace were fixed at one

end to the strong floor, and the instruments developed an angle of inclination with the frame

deflection. Corrections were made to account for the changing angles as the frame deflected.

These corrections are summarized in Appendix D.

Displacement at the brace midpoint should be relatively close comparatively based on geometry of

all the specimen frames. This is based on OOP displacement of the brace being directly related to

the story drift. The relationship between horizontal drift range and brace out-of-plane

displacement is illustrated in Figure 6.3.6 using Equations (6.3.2) through (6.3.5). The angle in-

plane angle of the brace is given as .

(6.3.2)

(6.3.3)

(6.3.4)

(6.3.5)

223

Figure 6.3.6: Horizontal Drift Range to Brace OOP Displacement Relationship

Figure 6.3.7 plots the out-of-plane displacement of the brace center verses the total drift range for

three of the specimens; WF-23 with the wide-flange brace section, HSS-24 utilizing the thin

rectangular gusset plate and bolted shear plate welded flange beam-to-column connection, and

HSS-22 with the thin tapered gusset plate and shear plate beam-to-column connection. The brace

displacement is normalized by the actual brace length, end to end.

Figure 6.3.7: Brace Out-of-Plane vs. Total Drift Range Comparison

The trends in the behavior are relatively close for all of the HSS specimens but the displacement of

the wide-flange brace section for WF-23 was greater than both HSS-22 and HSS-24 at the smaller

224

drift ranges, less than 3%. The specimen with thin tapered gusset plates, HSS-22, showed larger

out-of-plane displacements at the larger drift ranges.

6.3.3.2 Gusset Plate Rotations

The rotation of the gusset plates as the brace buckled out-of-plane in compression presented in this

section. The gusset plate designs for all of the specimens in this test program utilized an elliptical

clearance to permit rotations and the development of hinges at the brace ends. Instrumentation

recorded the linear displacements at two points near the free edge of the gusset plate, and ,

and were used to determine gusset plate rotations. The displacements are divided by the distance

from the adjacent beam or column flange to the instrumentation point of contact, and , and

averaged to give rotation in radians, . This method for is illustrated in Figure 6.3.8.

Figure 6.3.8: Gusset Plate Rotation

The gusset plate thickness and geometry attributed to the level of rotation achieved during the tests.

Figure 6.3.9 shows the rotations in radians at the northeast gusset plate for three of the specimens,

HSS-22, HSS-24, and HSS-25, over total drift range. The thin tapered gusset plates from HSS-22

were more flexible and achieved larger out-of-plane displacement and rotation than rectangular

gusset plates at a given drift range. Also, the 3/8” rectangular gusset plates in HSS-24 achieved

larger rotations than the 7/8” thick rectangular gusset plates of HSS-25.

225

Figure 6.3.9: Example of NE Gusset Plate Rotations

Table 6.3.4 summarizes the northeast gusset plate rotations and compares values at two drift ranges

and immediately prior to system failure. The first selected drift range for comparison is at 1.25%.

In all cases visible rotation of the gusset plates was observed and out-of-plane displacement of the

brace center exceeded the B2 damage level. The second drift range is at 2.75%. For most of the

specimen, this is immediately prior to observing local deformation in the plastic hinge at the brace

center. The maximum-recorded rotations with the corresponding total drift ranges are included for

each specimen. An average rotation of the NE gusset plate for the seven specimens is calculated

and also included for comparison.

Table 6.3.4: NE Gusset Plate Rotation Comparison Summary

HSS-22 with the thin tapered gusset plates exhibited the largest rotations of all the specimens. At

1.25% drift range, the rotation reached 0.1094 radians, 22% more than the average rotation. The

rotation reached 0.1546 radians at 2.75% drift range, 19% larger than the average. HSS-25 utilized

a 7/8” thick gusset plate, which resulted in smaller rotations than the other specimen at both

compared drift ranges, 0.0785 radians at 1.25% and 0.1181 radians at 2.75%. This is 12.5% and 9%

Rotation Drift Range

HSS-18 0.0868 0.1212 0.1435 4.19 Unwelded Beam/Col Conn. 0.375 Rectangle

HSS-20 0.0838 0.1254 0.1484 3.97 18 Bolted Beam End Plate Conn. 0.375 Rectangle

HSS-21 0.0876 0.1262 0.1494 4.14 14 Bolted Beam End Plate Conn. 0.375 Rectangle

HSS-22 0.1094 0.1546 0.1971 3.98 Unwelded Beam/Col Conn. 0.375 Tapered

WF-23 0.0934 0.1378 0.1907 5.56 WF brace section 0.375 Rectangle

HSS-24 0.0882 0.1257 0.1535 4.44 Bolted Web CJP Flange Conn. 0.375 Rectangle

HSS-25 0.0785 0.1181 0.1272 3.30 Heavy Beam and Gusset Plate 0.875 Rectangle

Average 0.0897 0.1299 0.1585 4.23 - - -

SpecimenRotation

at 2.75%Discription

Gusset Plate

Thickness, in.

Gusset

Shape

Rotation

at 1.25%

Maximum Recorded

226

below the average rotations from the seven tests in this comparison, respectively. HSS-18, HSS-20,

HSS-21, and HSS-24 all had identical gusset plate designs and the behaviors exhibited by all four

specimens were very similar.

6.3.3.3 Brace Buckled Shape and Curvature

Displacements over the brace length were used to create the buckled brace shape for comparison.

Potentiometers recorded the out-of-plane displacements at four points along the northwest half of

the brace which were mirrored to the southeast portion of the brace assuming symmetry. A picture

of the buckled brace shape is created that can be used to evaluate the curvature over the brace

length. The values for displacement and the locations of the instrumentation along the brace are

shown as a percent of the original brace length. Figure 6.3.10 illustrates the approach for plotting

the buckled brace shapes in this section. Straight lines are used connecting the known

displacements along the brace but theory of the actual curvature can be made based on the change

in slopes between lines.

Figure 6.3.10: Instrumentation for Determining Buckled Brace Shape

Brace shape was compared at the different total drift ranges rather than only negative story drift

because the out-of-plane displacement and resulting shape are dependent on a combination of both

the tensile yielding and residual elongation and the buckling behavior of the brace in compression.

The same descriptions for initial (0% to 1.25%) and moderate (1.25% to 2.75%) drift ranges used in

Chapter 5 when evaluating test performance are used here to show the progression of brace shape.

The comparison for the six specimens in Figure 6.3.11 was taken for initial drift ranges from 1.27%

to 1.44% and for moderate drift ranges from 2.61% to 3.04% for Figure 6.3.12. The B1 and B2

damage levels for brace buckling were typically reached within the initial drift ranges. In the

Plotted Buckled Shape

Brace End

Edge of G

usset

Brace 14 Point

Brace Midspan

Original Brace Position

NorthwestDirection

Loriginal

%= poteniometer/Loriginal

Estimated Curvature

227

moderate ranges, the brace began showing less parabolic shape and tended to form a more

triangular shape as the plastic hinge formed at the brace midpoint.

The brace shapes for HSS-19 was not included in the comparisons because the brace buckled

downward and immediately formed a hinge at the extension plates connecting the brace to the

WTs. The brace shapes for HSS-21 were left out due to erroneous potentiometer data at the brace

center and at the gusset plates. HSS-26 was also left out because the test protocol focused on the

tension dominated near-fault loading history and fracture of the brace occurred prior to significant

brace buckling.

Figure 6.3.11: Brace Shape Comparison at Initial Total Drift Range (Approx. 1.25%)

The shape of WF-23 with the W6x25 brace section shows the largest displacements of the six

specimens compared at this initial total drift range. All of the other specimens display similar

maximum out-of-plane displacement at midpoint but there are some variations in the shape of the

brace. It can be seen that the gusset plate rotations for HSS-25 near the lowest of those compare

but the displacement at the ¼ points in near to highest. This is an example of double curvature

over the brace length due to the increase rigidity of the heavier gusset plates. The shape of the

other specimens are somewhat clustered and display a parabolic shape over the length more typical

of pinned-pinned connections. The buckled brace shapes are shown in Figure 6.3.12 for moderate

drift ranges, approximately 2.75%.

228

Figure 6.3.12: Brace Shape Comparison at Moderate Total Drift Ranges (Approx. 2.75%)

The progression of the buckled brace into a more triangular shape at the moderate total drift ranges

can be seen. All of the specimens are showing increased pinching near the brace midpoint as the

plastic hinge formed. This increased curvature indicates increased local strains in this region and

strain accumulation occurring at a smaller drift ranges could limit the life of the brace under cyclic

loading. HSS-25 again shows smaller rotations at the gusset plates more closely resembling a

compressive member with fixed-fixed end supports. HSS-20 achieved a larger midpoint

displacement but relatively lower ¼ point displacements which resulted in great curvature over the

plastic hinge region. HSS-25 and HSS-20 both achieved the smallest total drift range of the

specimens compared, 3.30% for HSS-25 and 3.97% for HSS-20. The gusset plate rotations for

HSS-22 were the largest of those compared and the midpoint out-of-plane displacement was the

largest of the specimens with the tubular HSS brace at this drift range. This suggests at this drift

range that the inelastic behavior of the brace and gussets plates has progressed to higher level for

HSS-22, which only achieved 3.98% total drift range. HSS-18, HSS-24 and WF-23 exhibited more

of a parabolic curve over the length but also show higher curvature at the brace midpoint and also

resulted in higher total drift ranges. Local deformation, cupping and bulging, was observed shortly

after 2.75% drift range in all specimens with the exception of WF-23, which fractured at the

northeast gusset to frame interface welds before local deformation could occur.

The brace shapes shown in Figure 6.3.13 are from the last successful full cycle prior to brace (or

system) failure. Increased curvature over the brace lengths with severe pinching at the brace center

was visible for all test specimens, with the exception of WF-23. The buckled brace of Specimen

WF-23 maintained a half-sine shape throughout the test and never exhibited the local deformation

229

of a fully formed plastic hinge at the brace center. All five of the other specimens with the HSS

brace sections showed extreme local deformation at the plastic hinge which is reflected in the

curvature over the brace shape.

Figure 6.3.13: Buckled Brace Shape Immediately Prior to Failure

The buckled shape of the brace is important to the overall life of the system, the best evidence of

this being the brace behavior of specimen WF-23. The W6x25 wide-flange brace section was able

to achieve larger out-of-plane displacements and delayed the formation of the plastic hinge at the

brace center. A logical reason for this variation is the significant difference in local deformation as

brace buckling occurs. WF-23 showed a gradually curving buckled shape throughout the entire

test, and resulted in the largest total drift range.

HSS-24 and HSS-18 achieved larger total drift ranges compared to the other HSS brace section

specimens in this comparison. HSS-18 and HSS-24 both displayed a more favorable buckled brace

shape at 2.75% and the severity of curvature before fracture was less than those of HSS-22, HSS-20

and HSS-25. Buckled braces that maintain a half-sine wave shape, similar to that of a pinned-

pinned member, delayed the formation of the plastic hinge at the midpoint and improved the

ultimate total drift range achieved by the system. Conversely, HSS-22 showed the largest rotation

at the gusset connections, most resembling pinned-pinned end supports, but resulted in early

concentration of damage at the brace midpoint and smaller total drift range. This suggests that the

brace behavior and the brace life are also influenced by the rigidity of the beam-to-column

connection and not just the gusset plate behavior.

230

However, it is clear that the gusset plate behavior in compression plays a crucial role to the overall

performance of the system and the ultimate total drift range. Stiffer gusset plates like those used in

HSS-25 induce double curvature into the buckled brace shape at smaller drift levels and ultimately

limit the life of the brace. Larger curvature demands of the brace result in greater inelastic strain at

the brace center. Once the plastic hinge is formed, the HSS brace sections quickly develop severe

local buckling in compression and micro-cracks in tension, eventually propagating through the

section. By delaying the localization of damage that is related to the local strain and curvature of

the brace and that immediately precedes buckling, the overall life of the system will be increased.

Table 6.3.5 arranges the test by their total drift ranges and compares observations related to the

buckled brace shape.

Table 6.3.5: Drift Range and Brace Shape Comparison

6.3.3.4 Weld Damage

The behavior and damage to the beam to column interface welds are discussed in this section. The

interface welds are required to transfer load from the gusset plates to the beams and columns, and

provide the necessary resistance and deformability to allow the brace and gusset plate to rotation

out-of-plane. Current AISC design procedures size the welds for the maximum expected

combination of tension and shear from the expected brace force and provide the 2t linear clearance

at the end of the brace to allow for out-of-plane rotation of the brace. Experimental research has

shown that additional stress and deformation demands occur on the gusset plates and interface

welds at the reentrant corners because of gusset rotation and the opening and closing of the frame

Range Min Max

WF-23 5.56 -3.21 2.35Least Deformation of

Hinge at FailureWF brace section 0.375 Rectangle

HSS-24 4.44 -2.50 1.94Pinching at Hinge Prior to

Failure

Modified Beam/Col

Conn.0.375 Rectangle

HSS-18 4.19 -2.59 1.60Pinching at Hinge Prior to

Failure

Bolted Shear Plate

Beam/Col Conn.0.375 Rectangle

HSS-21 4.14 -2.55 1.59 N/A14 Bolted Beam End

Plate Conn.0.375 Rectangle

HSS-22 3.98 -2.48 1.50

Largest Gusset Rotations

and Pinching at Hinge at

Moderate Drift Range

Bolted Shear Plate

Beam/Col Conn.0.375 Tapered

HSS-20 3.97 -2.28 1.69Pinching at Hinge at

Moderate Drift Range

18 Bolted Beam End

Plate Conn.0.375 Rectangle

HSS-25 3.30 -2.41 0.88Smallest Gusset Rotations

and Double Curvature

Heavy Beam and

Gusset Plate0.875 Rectangle

HSS-26 1.38 -0.40 0.99 N/A

Heavy Bm and GP

with Near-Fault

Loading

0.875 Rectangle

HSS-19 1.31 -1.01 0.31 N/ABolted WT Brace

Connection0.500 Rectangle

Gusset

ShapeBuckled Brace ShapeSpecimen

Drift Ratio, %Discription

Gusset

Plate

231

under cyclic loading. At large drift ranges, cracking of the weld material or in the heat affected

zone (HAZ) of the gusset plate has been observed (Johnson 2005). If these cracks grow to

sufficient length that they reach their critical fracture stability limit, weld fracture may occur. A

schematic illustrating the additional demands on the gusset plates is provided in Figure 6.3.14.

Figure 6.3.14: Additional Demands on Gusset Plate Interface Welds

Complete fracture of the gusset to frame interface welds is an undesirable failure mechanism

SCBFs but some ductile tearing can be tolerated if it is controlled (Kotulka 2007). Ductile tearing

at the reentrant corners essentially reduces the gusset plate rotational stiffness and permits greater

brace end rotations. As the stiffness decreases, the effective length factor increases along with

the slenderness ratio of the brace. This potentially increases the ultimate drift capacity of the brace

by reducing the strain demand at the plastic hinge at the brace center.

During design of the nine specimens, the interface welds were sized to develop the full capacity of

the gusset plates or CJP welds were utilized. Interface weld cracking still occurred for HSS-22, WF-

23, HSS-24 and HSS-25. Fracture of the northeast gusset interface welds for WF-23 ultimately

controlled as the system failure mechanism. This occurred prior to fully developing the plastic

1-

Brace

in C

ompr

essio

n

Opening Momentat Connection

Tension AcrossGusset Plate

Column

Brace

SECTION 1-

Out-of-Plane Rotationof Gusset Plate

232

hinge at the wide-flange brace section midpoint. The level of interface weld tearing is summarized

in Table 6.3.6. The length of weld tearing is given as a percentage of the total weld length for both

the gusset to beam and gusset column edges. The scale used to show the crack length is

proportioned because the majority of damage was less than 10% of the total length. The negative

drift ratio at which cracking initiated or propagated is also provided.

Table 6.3.6: Weld Damage Propagation Summary

In HSS-22 cracking initiated in the gusset plates at moderate drift levels of -1.11% (1.90% total drift

range) and propagated to approximately 5.7% of the total interface weld length at the northeast

gusset and 5.4% at the southwest at -2.46% (3.97% total). The bolted shear plate beam-to-column

connection and the thin tapered gusset plates provide less stiffness at the frame connection which

increased the level of rotation at peak drift levels. HSS-22 noted severe yielding and deformation

of the gusset plates and nearly zero yielding of the framing elements.

The performance of WF-23 exceeded all other specimens for ultimate drift capacity and the local

displacement of the brace center. Initial cracking of the reentrant corners did not occur until the

negative drift level of -2.86% (4.91% total drift range) and was immediately observed as 7.2% of

total weld length for the northeast gusset connection. Weld damage then propagated to 49% of the

entire length at -3.21% drift (5.56% total drift range) and fractured during the subsequent tension

cycle. Additional available ductility is recognized for A992 wide-flange brace sections compared to

A500 B/C HSS sections. However, this increased deformation capacity also increases the ductility

demand of the gusset plate connection, including the interface welds, to accommodate the large

100

NE -1.46

SW -1.11

NE -2.86

SW -2.86

NE -1.83

SW -1.83

NE -1.94

SW

*B indicates cracking in base material (HAZ)

*WF

*Numbers indicate Negative Drift Ratio in % at Propagation

Specimen

WF-23

0

Length of Weld Tearing, % of Total LengthInitial

Crack 10 50

HSS-25

HSS-24

**(B) *-2.86

*-2.46

*-1.86

*-1.95

HSS-22

5

*-3.21

*-2.46

*-2.41

*-2.50

*-1.86

233

out-of-plane rotations of the gussets and increased drift levels in the frame needed to develop and

eventually fracture the brace at the plastic hinge.

HSS-24 and HSS-25 both showed only moderate levels of weld damage during the tests. For HSS-

24, cracking initiated at -1.83% drift (3.17% total) and propagated to a maximum 4.3% of the total

weld length at the northeast gusset plate prior to brace fracture and -2.50% drift (4.44% total).

HSS-25 initiated cracking at the northeast gusset plate at -1.94% drift (2.62% total). At the

maximum negative drift level of -2.41% (3.30% total drift range), cracking propagated to only 1.2%

of the total weld length.

The behavior of HSS-22, HSS-24 and HSS-25 show that ductile tearing of the interface welds is not

detrimental to the seismic performance of the system. Both achieved large negative drift levels and

the level of damage observed did not jeopardize the integrity of the welds to transfer forces

between the gusset plate and the framing elements. The damage to the interface weld observed in

WF-23 clearly exceeded what could be considered acceptable or beneficial. Specimen HSS-22, on

the other hand, maintained maximum resistance in the system and in the brace in tension and the

controlling failure mechanism occurred at the brace plastic hinge.

6.3.4 Brace Behavior in Tension

This section compares both brace and gusset plate elongation at total drift ranges. As the frame is

pushed to positive drifts, the distance across the diagonal must elongate to accommodate the

changed geometry. AISC Seismic Provisions for Capacity Based Design of a SCBF size the

diagonal brace section to yield in tension prior to any other component. The design of the

connections is based on the maximum expected force delivered from the braced and includes

strength reduction factors, φ, to limit the yielding to only the brace. As discussed in Chapter 4, the

Balance Design Approach intends to utilize the additional potential ductility in the system beyond

the brace to extend the life of the system at more severe drift ranges. Yielding of the gusset plates

in tension can reduce the strain demand over the brace length and potentially increase the life of

the system under cyclic loading.

6.3.4.1 Brace Elongation

This section compares the in-plane elongation of the brace. Instrumentation on the test specimen

measured the change in length of the brace, ΔL. The elongation is plotted as a strain value, in/in,

calculated as ΔL/Lo, with Lo being the brace length between gusset plate free ends not including

the overlapping splice length within the gusset plate. The strain calculated here is an

approximation of the actual brace strain. After buckling, the brace does not fully straighten at zero

234

displacement but hinges at the brace center. Yielding is concentrated at the hinge location rather

than over the full brace length.

Brace elongation is compared to the overall total drift ranges in Figure 6.3.15 for four of the

specimen for comparison. Specimen HSS-22 with thin tapered gusset plates, HSS-24 with 3/8 in

thick rectangular gusset plates, HSS-25 with 7/8 in. thick rectangular gusset plates, and WF-23 with

3/8 in thick gusset plates and the W6x25 brace section. Brace Elongation plots for all nine

specimens are included in Appendix C.

Figure 6.3.15: Brace Elongation vs. Total Drift Range

The W6x25 brace section in WF-23 was ASTM A992 steel and elongated more at a given drift

range than the HSS brace sections. The brace in HSS-25 having thicker gusset plate connections

also exhibited more elongation than the other specimens at similar drifts. The heavier 7/8” gussets

appear to increase the elongation in the brace. Specimen HSS-24 achieved slightly greater

elongation with the 3/8 in. rectangular gusset plates than HSS-22 having the tapered geometry.

Brace axial stress verses brace strain is plotted in Figure 6.3.16. Brace stress was calculated as

using the brace axial forces, , from Section 6.3.1.

235

Figure 6.3.16: Brace Axial Stress vs. Brace Strain

The experimental stress/strain relationship from the brace in tension closely resembles that of the

brace material coupon test. As expected for A500 B/C steel, the HSS braces does not strain

harden after yielding, where as the wide-flange section exhibited a clear yield plateau, and strain

hardening occurred further increasing the brace stress. This behavior could be one benefit of using

wide-flange brace sections. No coupon values were obtained for the A992 Grade 50 W6x25 brace

section but the yield stress is much lower than what would be expected.

6.3.4.2 Gusset Plate Elongation

Gusset plate elongation is taken as the elongation occurring from the beam to column work points

to the free edges of the gusset plates. It is calculated by subtracting the recorded brace elongation

data from the diagonal work point displacement data. The result is then divided by the original

length across the diagonal minus the brace length, free end from free end. The Balance Design

Procedure promotes tensile yielding of the gusset plates to increase the ductility of the system by

utilizing a β factor of 1.0 for yielding across the effective Whitmore width. The capacity of the

gusset plates must be sufficient to develop yielding of the brace section and as strain hardening

occurs in the brace, the gusset plates should yield, increasing the total elongation over the diagonal.

Gusset plate elongation verses total drift range is present in Figure 6.3.17.

236

Figure 6.3.17: Gusset Plate Elongation Comparison

Specimen HSS-22 exhibited the most gusset plate elongation until a total drift range of 2.5%.

Specimen HSS-18 achieved that most gusset plate elongation prior to failure. Both specimens

utilized the bolted shear plate beam-to-column connections adjacent to the gusset connections.

Significant yielding was observed over the entire area for the tapered gusset plates and the weld

damage observed at higher drift ranges was more than seen in any of the other test with HSS brace

sections. The heavier gusset plate used in HSS-25 showed no observed tensile yielding during the

test, which is consistent with the plot above. The gusset plate elongation of WF-23 was less than

the other specimens with thin rectangular gusset plates.

The gusset plate elongation can be compared to the brace elongation as illustrated in Figure 6.3.18.

The Balance Design Procedure has been used to increase the total ductility of the system by

extending yielding beyond the brace and into the gusset plate connections. Tensile yielding of the

gusset plates is encouraged by implementing a β factor of 1.0 in the design. However, the extent of

yielding should not limit the force developed in the brace and as a result brace yielding. A line

showing brace elongation equal to gusset plate elongation is provided for comparison. Specimens

that fall below the line show yielding of the gusset plate greater than in the brace such as HSS-18

and HSS-22, both of which utilized the bolted shear plate beam-to-column connections. HSS-20,

WF-23 and HSS-26 exhibit a response dominated by brace yielding and minimal tensile yielding in

the gusset plates. HSS-20 and HSS-24 show yielding balanced between both the brace and the

gusset plates.

237

Figure 6.3.18: Brace Elongation verses Gusset Plate Elongation

6.4 Frame Response

The in-plane rigidity of the beam-to-column connections is not easily determined and is a

combination of the actual beam-to-column connection fixity and the in-plane stiffness of the gusset

plate. A frame designed with a great elastic stiffness relative to brace stiffness will increase the

demands on the framing elements. As the system increases drift range and yielding and buckling of

the brace occur, the response of the frame plays a more crucial role to the response of the system.

The following sections evaluate the performance of the frames by comparing column moments and

shears for each specimen in Sections 6.4.1 and Section 6.4.2. The lateral frame resistance is

discussed in Section 6.4.3 including hysteretic plots of the frame response. Lastly, rotations of the

northwest and southeast shear tab connections are evaluated in Section 6.4.4

6.4.1 Column Moments

This section evaluates the column moments directly adjacent to the free edge of the gusset plate

where yielding and deformation typically were observed during the tests. Two gauges were

positioned at the outside center extreme fiber of ach flange at locations where the column was

expected to remain elastic. Plane sections remain plane implies profile is linear over the depth and

stresses are calculated by elastic theory. Moments are calculated as described in Figure 6.4.1 with

modulus of elasticity, , for structural steel equal to 29000ksi.. By having the moments at two

locations over the length of the column, the moment at the edge of gusset plate can be determined

238

through equilibrium. The method for calculating the moment at the edge of the gusset plate is also

illustrated in Figure 6.4.1.

Figure 6.4.1: Column Moment Calculation

The maximum moment from extrapolation is limited to the plastic moment of the sections

calculated as using the actual yield stresses obtained from material coupon tests

summarized in Table 6.4.1. Material properties for HSS-18 were not available and in some cases,

the results from the coupon tests were unavailable. In those cases, the yield stress was taken

as . The column moments at the northeast and southwest gusset plates are plots verses the

story drift in Figure 6.4.2 and Figure 6.4.3, respectively, for specimens HSS-22, HSS-24 and HSS-25

to illustrate different moment responses with different gusset plate design and beam-to-column

connection details. Moment is normalized by nominal or plastic moment, of the W12x72

column section. Figures illustrating column moments are available for all nine specimens in this

test series in Appendix C.

W12x72

a

b

a

b

a

b

E M=I*( a- b)d

Mgusset < Mp

Mnorth

Msouth

d

a

a

b

b

Lv

Lg

(Mnorth-Msouth)Lv

Mgusset = *Lg - Msouth

239

Table 6.4.1: Framing Element Material Properties

Figure 6.4.2: Column Moment at Edge of NE Gusset Plate

North South East West

HSS-18 N/A N/A N/A N/A

HSS-19 62.73 63.57 58.94 57.2

HSS-20 60.54 59.22 N/A N/A

HSS-21 62.35 60.1 56.75 56.65

HSS-22 60.53 60.53 47.74 59.13

WF-23 60.47 61.95 59.7 53.85

HSS-24 62.05 62.72 54.69 N/A

HSS-25 56.79 56.79 N/A N/A

HSS-26 N/A N/A 55.14 58.46

Specimen

Actual Yield Stress, Fy (ksi)

Beams Columns

240

Figure 6.4.3: Column Moment at Edge of SW Gusset Plate

The increased stiffness from the CJP welded beam-to-column connection, the heavier W16x89

beam sections, and larger gusset plates resulted in increased moments at the edge of gusset plate for

HSS-25. Specimen HSS-22, which utilized the bolted shear plate beam-to-column connection and

3/8 in. thick tapered gusset plates, exhibited the lower negative moment in the northeast column

with the brace in compression but comparable moments at the southwest column. In tension,

specimen HSS-22 exhibited little moment resistance at the edge of the tapered gusset plates for

both columns. Specimen HSS-24 utilized the bolted web CJP welded flange beam-to-column

connection and 3/8 in. rectangular gusset plates and achieved smaller moments than HSS-25 but

greater than HSS-22.

6.4.2 Column Shears

Column shears were calculated using the column moments from the strain gauge data from each

specimen and tabulated in Appendix C for all nine specimens. Moments were calculated at the two

locations over the length of the columns. Having no external forces applied over this length, the

shear is uniform between the two locations. The shear forces in each column can then be

calculated directly as illustrated in Figure 6.4.4.

241

Figure 6.4.4: Column Shear Calculations

The total shear resisted by both columns is equal to the lateral force resisted by the frame.

Maximum values of frame resistances were determined at peak drift ratios for the nine specimens.

The enveloped curves are plotted together in Table 6.4.2 for comparison. The frame resistance is

normalized by , the nominal lateral resistance of the theoretical frame assuming pinned

connections at the NW and SE shear tab connections and fixed connections at the NE and SW

connections adjacent to the gusset plates. The nominal resistance was determined as the force that

develops of the column at the fixed connections. was calculated in Equation (6.4.1)

and used a height of the frame, , equal to 144 in.

(6.4.1)

a

b

a

b

a

b

E M=I*( a- b)d

Mnorth

Msouth

Vcolumn

(Mnorth-Msouth)Lv

Vcolumn =

a

a

b

b

Vframe =Veast col.+Vwest col.

d

Lv

242

Table 6.4.2: Enveloped Frame Resistance vs. Drift Ratio H

SS

-18

HS

S-1

9

HS

S-2

0

HS

S-2

1

HS

S-2

2

WF

-23

HS

S-2

4

HS

S-2

5

HS

S-2

6

243

The initial slope of the plotted line is the elastic stiffness of the frame which was calculated using

the first small cycles from the enveloped resistances verses horizontal drift in inches for each

specimen in Table 6.4.2. In general, the plots show a nonlinear response for frame resistance that

gradually arches from the elastic to the inelastic region at larger drifts. The variation in response

was more substantial at positive drifts than at negative drifts because the in-plane stiffness of the

gusset plate influenced the in-plane stiffness of the beam-to-column connection greater in the

positive direction than the negative. The ratio of frame stiffness over total stiffness has been

calculated to evaluate the relative contribution from the frame during the elastic response of the

total system. Frame elastic stiffness, frame stiffness ratio, maximum resistances and corresponding

drift ratios are summarized in Table 6.4.3.

Table 6.4.3: Frame Stiffness and Peak Resistance Summary

The specimen with the heavier beam sections and stiffer gusset plate connections, HSS-25 and

HSS-26, exhibited the largest frame stiffness as expected but not the largest positive or negative

frame resistances. The bolted end plate beam to column connection also resulted in larger initial

stiffness of the frame, as seen with HSS-20 and HSS-21. The thin tapered gusset plates and bolted

shear plate beam-to-column connections of HSS-22 resulted in the lowest initial stiffness, while

specimens with thin rectangular gusset plates exhibited comparable initial stiffness, regardless of

beam-to-column connection detail.

In the positive direction, WF-23 achieved the largest frame resistance, 122.0 kip. In the negative

direction, the 18 bolt beam end plate connection specimen, HSS-20, achieved the largest negative

resistance, -138.0 kip, followed by WF-23 with 126.2 kip. Not including HSS-19 and HSS-26, the

KE,frame

KE,frame/

KE,total

Comp Tension Range Comp Tension Range

HSS-18 87.8 0.143 -109.8 81.5 191.3 -2.59 1.60 4.19

HSS-19 89.1 0.121 -68.5 49.0 117.5 -1.01 0.31 1.31

HSS-20 98.9 0.149 -138.0 95.3 233.3 -2.28 1.69 3.97

HSS-21 99.1 0.164 -120.0 109.7 229.7 -2.55 1.59 4.14

HSS-22 81.3 0.140 -107.5 62.4 170.0 -2.48 1.50 3.98

WF-23 89.1 0.120 -126.2 122.0 248.2 -3.21 2.35 5.56

HSS-24 87.0 0.127 -122.6 88.9 211.5 -2.50 1.94 4.44

HSS-25 118.8 0.147 -127.9 108.4 236.2 -2.41 0.88 3.30

HSS-26 143.4 0.185 -76.6 100.2 176.8 -0.40 0.99 1.38

Frame Resistance, kips

Specimen

Frame Stiffness,

k/inFrame Drift Ratio, %

244

thin tapered gusset plates and bolted shear plate connections in HSS-22 achieved the smallest

positive and negative resistances, -107.5 kip and 62.4 kip.

6.4.3 Frame Hysteretic Response

Hysteresis plots of the total column shear, or frame resistance, verses frame drift ratio are

compared in Table 6.4.4. The widening of the hysteretic loops at large drift ranges indicates the

nonlinear response of the frame. Wide, full loops represent more inelastic deformation of the

framing elements and increased energy dissipated by the frame.

Table 6.4.4: Frame Hysteresis Comparison

HS

S-1

8

HS

S-1

9

HS

S-2

0

HS

S-2

1

HS

S-2

2

WF

-23

245

HS

S-2

4

HS

S-2

5

HS

S-2

6

6.4.4 Shear Tab Connection Rotations

The shear tab connection rotations are compared in this section. Bolt elongation of shear

connections is considered a yield mechanism that can attribute to the ultimate drift capacity of the

system. Bolt fracture at the shear tabs is also recognized as a potential failure mode. The desired

behavior of the shear tabs is to be able to rotate to accommodate large drift ranges while

maintaining shear resistance. Potentiometers were connected to the inside of the beam flanges at

the shear tab connection to record horizontal displacement between beam and the column flange

outside face. The displacements were used to calculate rotation of the shear tab connection in

radians in Equation (6.4.2) and illustrated in Figure 6.4.5. Rotations are plotted verses the story

drift for the northwest shear tab in Figure 6.4.6 and for the southeast in Figure 6.4.7.

(6.4.2)

246

Figure 6.4.5: Shear Tab Rotation Calculation

Figure 6.4.6: Northwest Shear Tab Rotation Comparison

Colu

mn

Beam

Lro

t

b=Potentiometers

sheartab (radians) =Lrot

( a- b)

247

Figure 6.4.7: Southeast Shear Tab Rotation Comparison

The shear tab rotations exhibit a linear response to story drift. The behavior is closely banded for

all specimens at the southeast connection. There appears to be more discrepancy at the northwest

connection which could likely be attributed to the influence of the load beam connected at the

northwest corner of the specimen.

6.5 Distribution of System Resistance

This section compared the resistance of the system by dividing the contribution of story shear into

brace and frame resistances. In order to compare the calculated frame and brace resistance, a

measure is necessary to determine the accuracy of the analysis methods used to determine

resistance. Section 6.5.1 discusses the methods used to determine resistance and Section 6.5.2

assesses the distribution between brace resistance and frame resistance as a percentage of the total

system resistance over the full nonlinear response.

6.5.1 Evaluation of Analysis Methods for Resistance

The brace and frame resistances were calculated from component strain gauge data based on the

analysis methods describe in the previous sections. This section evaluates the accuracy of the

analysis methods for calculating brace and frame resistance by checking for equilibrium with the

applied lateral load acting on the specimen. The horizontal component of the brace force plus the

column shears should equal the applied lateral load as shown in Equation (6.5.1).

248

(6.5.1)

Table 6.5.1 through Table 6.5.9 summarizes the results from checking equilibrium between the

applied load verses internal forces calculated through strain gauge data for each specimen. In each

table the plot on the left compares the enveloped brace axial forces obtained directly through the

brace strain gauge data as described in Section 6.3.2.1 and the brace force based on the measured

column shears minus the calculated shear in the columns as described in 6.4.2. The expected

horizontal force resisted by the brace only is determined by subtracting column shears from the

applied lateral load. The brace shear is divided by the cosine of the brace angle to obtain the

expected brace force and shown in Equation (6.5.2).

(6.5.2)

The right plot in the tables is a comparison of the system hysteretic response calculated using

applied actuator load, , and the total calculated horizontal resistance, , which is

determined by adding the frame shear resistance from Section 6.4.3 with the horizontal

components of the brace resistance as described in Section 6.3.2.1. Percent error is calculated at

three positive and negative drift levels; at tensile yield and buckling, at 1.0% drift, and at the

ultimate positive and negative drift levels. Percent error is calculated using Equation (6.5.3).

(6.5.3)

249

Table 6.5.1: HSS-18 Equilibrium Evaluation


250



251


Table 6.5.6: WF-23 Equilibrium Evaluation

Bra

ce F

orc

e C

om

pa

ris

on

Sy

ste

m C

om

pa

ris

on

% Error in

Compression

5.5

-2.7

-8.3

WF-23 Equilibrium Evaluation

% Error Between Applied and Calculated Total Resistances

Location of System Plot % Error in Tension

Drift% at Brace Yield/Buckling 2.6

Ultimate Drift% 9.6

+/-1.0% 7.3

252



253


The applied and calculated resistances were within a maximum 25% at brace yield and at ultimate

drifts for all specimens in this test series. Typically, the largest differences between the calculated

resistance and the applied force occurred at maximum compressive forces immediately after brace

buckling. The calculated values for brace and frame resistance are used in the following section to

evaluate the distribution of lateral force resisted by each.

6.5.2 Distribution of Resistance

The nonlinear response an SCBF system is a combination of axial response of the brace and the

flexural response of the frame. At small drifts, the majority of resistance can be attributed to the

brace but as drift increases, the demands on the frame also increase and frame action plays are

more significant role to the total resistance of the system. This section evaluates the distribution of

resistance between the brace and the frame. SCBFs under severe seismic loading are expected to

respond nonlinearly to accommodate the large drift demands. The nonlinear response of the

system is a combination of the brace and frame responses with the contribution of each varying as

drifts increase. This section looks to evaluate the relationship between the brace and frame

resistances over the full drift range and whether the distribution attributes to increased drift

capacity and delayed brace fracture. The percentages of the total story shear resisted by the brace

and by the frame are calculated for each specimen in Equation (6.5.4) and (6.5.5), respectively.

254

(6.5.4)

(6.5.5)

Table 6.5.10 shows the distribution of resistance over drift ratio for each specimen. These plots are

an effective illustration of how the system of brace and frame work together to resist the applied

load. The previous section showed that the behavior of frame resistance at negative drifts was

similar for all of the specimens. At positive drifts, the beam to column connection type and the

gusset plate stiffness have a greater effect on the demands placed on the frame. In these plots the

rate of transfer between brace resistance to frame resistance as drifts increase reflects the efficiency

of the system. A steep drop off at negative drifts shows less than ideal performance of the brace

and places more demand on the frame at earlier drift levels. A flat slope at positive drifts levels

indicates that resistance in the frame is not being developed and demands are isolated to the brace.

The specimens that showed the highest performance for ultimate drift capacity and resistance

exhibited a balance between brace and frame resistance.

Table 6.5.10: Distribution of System Resistance

HS

S-1

8

HS

S-1

9

HS

S-2

0

HS

S-2

1

255

HS

S-2

2

WF

-23

HS

S-2

4

HS

S-2

5

HS

S-2

6

HSS-24 performed well for comparison purposes in that it reached comparable negative drift levels

and the largest positive drift levels and total drift range of the other specimens with HSS braces

within this test program. In the plot above, the transition between brace resistance contribution to

frame resistance is fairly gradual and the lines cross at approximately -1.25% drift. HSS-22 on the

other hand reached comparable negative drift levels but was near the bottom when comparing

positive drift levels and total drift range. The distribution plot for HSS-22 shows a steep decline in

brace resistance and relied more on the frame resistance at an earlier negative drift level,

approximately -0.55%. The thin tapered gusset plate and bolted shear plate beam-to-column

connection appear to result in a less balanced response, placing more demand on the brace and

gusset plates in tension and more on the framing elements in compression at larger drift ranges.

HSS-24 exhibited a more balanced response and showed more brace contribution at larger positive

and negative drift levels prior to brace fracture.

256

The plots of HSS-18, HSS-20 and HSS-21 show the majority of resistance is attributed to the brace

until transition occurred at smaller negative drift levels, approximately -0.90%, -0.55%, and -0.70%

for HSS-18, HSS-20 and HSS-21, respectively. The brace in HSS-25 showed the slowest rate of

degradation which is reflected in the distribution plot. The majority of resistance transferred from

the brace to the frame at approximately -1.6% drift. At positive drifts, the stiffness of the heavier

beam section and the large gusset plates increased the influence of the frame contribution to the

total resistance immediately after brace yielding.

The distribution plot for WF-23 shows that the brace resistance quickly dropped off and the

majority of story shear was resisted by the frame at early negative drift levels, approximately -

0.65%. At negative drifts, the opening moment of the beam-to-column connection increase the

demands at the gusset plate reentrant corners and interface welds. This could be reason why WF-

23 experienced severe damage at the gusset plate interface welds and eventually fractured.

Significant weld damage was also observed for HSS-22 which has similar behavior.

The distributions of resistance are summarized in Table 6.5.11. The percentages of resistance are

tabulated at drifts levels corresponding to buckling of the brace in compression and yielding in

tension and at ultimate drift levels. It can be noted that the balance between resistances for HSS-

24, which is thought to have shown better performance, was better compared to HSS-22. At

ultimate drifts, the distribution of resistance for HSS-24 was 29.3% brace and 70.7% frame in the

negative direction, and 74.6% brace and 25.4% frame for the positive direction. For HSS-22, the

resistance in the negative direction was calculated as zero for the brace relying completely on the

frame. In the positive direction, 81.0% of the total shear was resisted by the brace and 19.0% by

the frame. This seems to support that a more balance response at negative drift levels after

buckling of the brace has occurred can improve the performance of the system. Also, HSS-25

shows larger contribution of frame resistance at the ultimate negative drift level and resulted in

limited total drift capacity and early brace fracture. This suggests that a stiffer frame resulting in

higher frame resistance at ultimate drift levels can also work adversely against the overall

performance of the system. Of the nine specimens in this research, HSS-24 exhibits a balanced

response between brace and frame resistances that resulted in the largest ultimate drift capacity.

257

Table 6.5.11: Resistance Distribution Summary

6.6 Energy Dissipation

The nonlinear response of SCBF systems dissipate energy through brace buckling and yielding of

the brace, gusset plates and framing elements. This section evaluates each specimen by calculating

the energy dissipation of the total system as well as energy dissipated by the brace and by the frame.

Energy dissipation of a cyclic system is calculated using Equation (6.6.1) with equal to applied

force and Δ equal to displacement in the direction of the applied force. The variable is the initial

increment and is the total number increments of a cycle.

(6.6.1)

For each specimen in this test program, the total energy dissipation is calculated using load and

horizontal frame displacement data from the instrumentations. Energy dissipated per cycle is the

area enclosed within each completed hysteretic loops and calculated using Equation (6.6.2). An

illustration of dissipated energy is shown in Figure 6.6.1.

(6.6.2)

Brace Frame Brace Frame Brace Frame Brace Frame

HSS-18 82.5 17.5 83.6 16.4 16.1 83.9 76.9 23.1

HSS-19 81.9 18.2 80.2 19.8 10.1 89.9 76.0 24.0

HSS-20 79.5 20.5 86.7 13.4 18.3 81.7 73.4 26.6

HSS-21 79.2 20.8 83.6 16.4 13.2 86.8 71.6 28.5

HSS-22 85.4 14.6 87.8 12.2 0.0 100.0 81.0 19.0

WF-23 87.2 12.9 82.7 17.3 5.7 94.3 67.1 32.9

HSS-24 84.9 15.2 86.8 13.2 29.3 70.7 74.6 25.4

HSS-25 87.3 12.7 85.0 15.0 34.0 66.0 71.8 28.2

HSS-26 78.1 21.9 83.2 16.8 62.8 37.2 70.7 29.3

% Resistance at Brace Yield/Buckling

-Drift% +Drift%

% Resistance at Ultimate

-Drift% +Drift%Specimen

258

Figure 6.6.1: Energy Dissipation (Kotulka 2007)

Dissipation of energy is an interesting global comparison but should be noted that the total energy

dissipated is dependent on the number of cycles completed during the test. Although all the tests

with the exception of HSS-26 were subjected to the same induced loading protocol, the actual

displacement response per specimen varied due to losses in the test setup. A specific specimen that

experienced significant losses during the testing procedure between the actual and induced

displacement could have completed more cycles in order to reach a given drift. This would

increase the total dissipation of energy at a given drift range. In order to be directly comparable, all

specimens would have had to been subjected to identical displacement histories. Regardless, the

displacement histories from this test program were similar and the energy dissipation comparisons

are still valuable and reflect the total inelastic response of the system. Total energy dissipation

verses total range is compared in Figure 6.6.2.

259

Figure 6.6.2: Total Energy Dissipation Comparison

The rate of total energy dissipation is relatively similar for all of the specimens with the exception

of HSS-25. Specimens that achieved larger drift ranges essentially dissipated greater total energy. A

more creditable evaluation in the how each test parameter affects energy dissipation is to

distinguish between energy dissipated through the brace diagonal and energy dissipated by the

framing elements. Figure 6.6.3 shows a comparison of energy dissipated over the brace diagonal

which includes the contributions from the brace and the gusset plates combined. The actual brace

force determined from strain gauge data and displacements over the frame diagonal, work point to

work point, were used to calculate energy dissipation.

260

Figure 6.6.3: Brace Energy Dissipation Comparison

The rate of energy dissipation for the HSS-25 brace diagonal is higher than the other specimens

because the increased stiffness of the gusset plate connections develops higher resistance in the

brace. Also, the brace in HSS-25 showed the least degradation of resistance after buckling and the

largest reloading stiffness from the hysteretic plots, both of which increase energy dissipation. The

wide-flange brace section of WF-23 dissipated energy over the diagonal at a lower rate for total

drift ranges greater than approximately 2.75%. This is consistent with the specimen achieving a

lower maximum tensile brace force and the steep degradation of compressive resistance post

buckling. HSS-18, HSS-20, HSS-21, HSS-22 and HSS-24 all show very similar energy dissipation

behavior for the brace diagonal. At drift ranges of approximately 4%, close to the maximum drift

range for all five, it is possible to distinguish between the values with HSS-22 dissipating the least

energy and HSS-24 and HSS-21 dissipating the most. Comparison of the energy dissipation by the

framing elements are provided in Figure 6.6.4

261

Figure 6.6.4: Frame Energy Dissipation Comparison

Specimen HSS-22 dissipated the most energy through the frame at a given drift range. The HSS-22

frame hysteresis from Table 6.4.4 shows the bolted shear plate connection exhibited the lowest

resistance but also the response is highly inelastic in both the positive and negative direction. This

created wide hysteretic loops and dissipated more energy than the other specimens for a given drift

range greater than 2.0%. Interestingly, specimen HSS-22 exhibited very little observable frame

damage during the test but still dissipated energy effectively. Observations of HSS-24, on the other

hand, note significant column yielding and deformation at larger drift ranges but have lower total

energy dissipation from the frame at a given drift range. This suggests that observed inelastic

deformation of the framing elements does not necessarily equate to the ability of the frame to

dissipate energy.

Figure 6.6.4 shows that energy dissipation for all the specimens other than HSS-22 are fairly banded

up to approximately 3.0% total drift range when the energy dissipated by WF-23 and HSS-20

increased for a given drift range. WF-23 dissipated the most energy through the frame followed by

HSS-22. The least energy dissipated through the frames was by HSS-19 and HSS-25. A summary

of the maximum energy dissipated by the system, brace diagonal and by the frame are displayed in

Table 6.6.1.

262

Table 6.6.1: Percent Total Energy Dissipation Summary

HSS-18, HSS-20, HSS-21, and HSS-24 dissipated between 4825 kip-in and 5237 kip-in of energy

and reached total drift ranges from 3.97% to 4.44%. These specimens all shared identical gusset

plate designs but varied the beam-to-column connection type. The welded beam flanges and

bolted web plate connection for HSS-24 dissipated the most total energy while the 14-bolt beam

end-plate connection for HSS-21 dissipating the least. The distribution of energy dissipated by

these specimens between the brace diagonal and the frame were also similar and ranged between

76% and 81% of the total energy dissipated.

HSS-22 with the tapered gusset plate and bolted shear plate beam-to-column connection dissipated

less energy, 4589 kip-in, than the specimens with rectangular gusset plates. HSS-25 also dissipated

less energy, 4181 kip-in, and utilized the thick rectangular gusset plate and heavier beam section.

Both showed lower amounts of energy dissipated by the braces, 3341 kip-in and 3397 kip-in, for

HSS-22 and HSS-25 respectively. HSS-22 showed more energy dissipated by the frame than the

stiffer HSS-25 but neither were able to dissipate energy throughout the entire system as the

specimen with thin rectangular gusset plates.

WF-23 reached the largest total drift range and showed in Figure 6.6.4 that the rate of energy

dissipation from the frame increased at larger drift ranges, whereas energy dissipated by the brace in

Figure 6.6.3 increased evenly at drift ranges beyond what was achieved by any other specimen.

WF-23 dissipated 8300 kip-in of energy total and exhibited the largest contribution through energy

dissipated by the frame than the other specimens, 36%.

Table 6.6.2 shows energy dissipated by the system, brace diagonal and by the frame for specimens

HSS-18 thru HSS-25. Energy dissipation through the brace was calculated using the brace forces

determined in Section 6.3.2.1. The frame resistances determined in Section 6.4.3 were used o

calculated the frame energy dissipation.

kip-in % kip-in %

HSS-18 5000 3848 77 997 20 4.19

HSS-19 620 483 78 55 9 1.31

HSS-20 5019 3824 76 1044 21 3.97

HSS-21 4825 3891 81 976 20 4.14

HSS-22 4589 3341 73 1248 27 3.98

WF-23 8300 5259 63 2963 36 5.56

HSS-24 5237 4161 79 972 19 4.44

HSS-25 4181 3397 81 630 15 3.30

Specimen

Total Energy

Dissipated,

kip-in

Brace Diagonal Energy

DissipationTotal Drift

Range, %

Frame Diagonal

Energy Dissipation

263

Table 6.6.2: Energy Dissipation of Brace vs. Frame

HS

S-1

8

HS

S-1

9

HS

S-2

0

H

SS

-21

HS

S-2

2

WF

-23

HS

S-2

4

HS

S-2

5

The breakdown of energy dissipation can be taken further and the energy dissipated by the brace

diagonal can be divided into energy dissipated by the actual brace section and by the gusset plates.

The same energy calculation is used as with the diagonal with the exception that the brace uses the

displacement over the actual brace length and the gusset plates use the difference between the

diagonal and the brace displacements. Unfortunately, the corrections for the instrumentation that

264

recorded the displacement over the actual brace length accounting for gusset plate rotations does

not provide a level of accuracy needed for the energy calculation. However, reliable values were

obtained for WF-23 and the distribution of total energy dissipation is shown for the brace, the

gusset plates and the frame. The percentages for each for WF-23 are as follows: 52.4% Brace,

10.9% Gusset Plates, and 35.7% Frame.

Figure 6.6.5: WF-23 Energy Dissipation by Component

6.7 Performance Level Comparisons

This section compares the performance levels for each component within the system related to

immediate occupancy, life safety and collapse prevention. The nonlinear responses of SCBFs

subjected to seismic loading are expected to produce damage in the form yielding and deformation.

More favorable behavior would be to achieve larger drift ranges while minimizing the level of

damage in the components in order to achieve a higher performance level based on Performance

Based Design criteria. Visual observations during the test and instrumentation were used to

determine at what drift levels specific levels of damage occurred. Damage state designations were

established in Chapter 5 based on severity and potential for inducing system failure. The total drift

range was also separated into three performance levels in Chapter 5, initial, moderate and severe, is

referred to when describing when damage occurred.

The performance of each specimen is evaluated in a tabulated format that uses shading to denote

progression of damage as total drift range increased. Component performance levels are compared

separately for the brace, gusset plates and the framing elements. Designations for component

265

damage levels are used which were described in Chapter 5, Section 5.2. The damage is separated

into tensile and compressive system behaviors depending on the direction of drift during the test.

Table 6.7.1 below compares brace performances.

Table 6.7.1: Brace Performance Comparison

The performance of the brace for HSS-18, HSS-20, HSS-21 HSS-22, and HSS-24 are all similar

with severe damage not occurring until severe drift ranges. The onset and progression of brace

buckling in compression (B1 and B2) and yielding in tension (Y3) occurred near the same drift

levels for all four specimens. However, HSS-24 developed the local deformation associated with

the plastic hinge (BC) at a higher drift range and was able to achieve larger total drift ranges than

the other three.

WF-23 saw severe yielding of the brace in the region near the mid span at moderate drift ranges but

this damage did not appear to affect the integrity of the brace. The wide-flange brace section was

the only specimen to be painted with white wash over the length of the brace and the level of

yielding observed was significant. However, even well into severe drift levels, no local deformation

associated with the formation of the plastic hinge was observed. Fracture did not occur in the

brace but the northeast gusset plate interface welds fractured.

266

Specimen HSS-25 and HSS-26 had no net section reinforcement. The brace performance for HSS-

25 shows that net section damage (NSD) occurred at an initial drift range in tension and can be

seen as the darker shading associated with the severe damage state. Initial buckling and the early

progression occurred at similar drifts as with the specimens mentioned above but the local

deformation at the plastic hinge in compression and eventual brace fracture tension both occurred

at smaller drift ranges.

When comparing the performance of the HSS-26 brace, it should be considered that the specimen

was subjected to the alternative tension dominated near fault loading protocol. The brace was able

to achieve initial buckling (B1) but the majority of brace response occurred in tension. It is seen

that net section damage occurred at initial drift levels proceeded by net section fracture and an early

moderate drift range.

Specimen HSS-19 is an example of poor connection performance because the connection did not

permit development of the brace. Buckling occurred in the brace extension plates connecting the

HSS5x5x3/8 to the two WT4x17.5 sections. Moderate and then severe damage states were

observed at initial drift ranges and fracture at an early moderate drift range. Severe yielding and

buckling in the concentrated area of the extension plate drastically shortened the life of the brace.

Gusset plate performances are compared in Table 6.7.2

267

Table 6.7.2: Gusset Plate Performance Comparison

HSS-18, HSS-20, and HSS-21 exhibited similar gusset plate performances. Minor yielding in

compression and tension were observed at initial drift ranges. Moderate yielding occurred earlier

for HSS-21, followed by HSS-20 and HSS-18 near the transition between initial and moderate drift

ranges. At severe drift ranges, all three exhibited severe yielding over their entire area and

deformation along the free edges extending from the reentrant corners. The gusset plate

performance of HSS-24 was also comparable. The progression of damage followed those of HSS-

18, HSS-20 and HSS-21 for initial and moderate drift ranges, but in the severe drift range, weld

damage was observed and severe yielding was not noted until immediately prior to fracture failure.

HSS-22 saw moderate damage at initial drift ranges and weld damage at moderate drift ranges.

Severe yield was also not observed until immediately prior to failure.

The gusset plate performance for WF-23 varied from those with HSS brace sections. Tensile

yielding also occurred early in initial drift range but compressive yielding as the gusset plate rotated

out-of-plane was not observed until the moderate drift range. This is interesting considering the

brace behavior and the level of buckling at initial drift ranges shown in Table 6.7.1. The gusset

plate exhibited significantly less damage at initial and moderate drift ranges than the specimens with

268

HSS brace sections. Well into the severe drift range, severe yielding was noted and interface weld

damage occurred. The damage propagated becoming the controlling failure mechanism.

HSS-25 and HSS-26 utilized the heavier 7/8” gusset plates. HSS-25 showed no damage in tension

throughout the test but did exhibit yielding at the plate rotated out of plate. Minor yielding was

observed in the initial drift range followed by moderate yielding in the moderate range. Weld

damage was also observed in the moderate drift range but only slightly propagated. The gusset

plates for HSS-26 stay exhibited only limited yielding at one reentrant corne after failure.

Only minor damage was observed in the gusset plates for HSS-19 because of the reduced

maximum brace forces and location of hinging at the extension plate connecting the brace to the

southwest gusset plate. Minor yielding did occur in the initial drift range as the gusset plates rotated

out-of-plane but no damage was observed with the brace in tension prior to failure.

Table 6.7.3: Framing Element Performance Comparison

The framing element performance varies considerably for HSS-18, HSS-20, HSS-21 and HSS-24.

Gusset plate and brace designs were identical for these four tests but the crucial test parameter was

in the beam-to-column connection detail. The un-welded connection for HSS-18 was the most

flexible connection and minimized damage. Minor yielding occurred in the beams and columns

269

during the moderate drift range. Moderate yielding was noted during the severe drift range

immediately prior to failure.

HSS-20 and HSS-21 both utilized the bolted beam endplate connection detail with an 18 bolt

configuration for HSS-20 and 14 bolts for HSS-21. Bolt exhibited minor yielding of the beams and

columns with the brace in tension at the initial drift range but HSS-21 also saw bolt fracture. Minor

yielding with the brace in compression for occurred at moderate drift ranges for both. At severe

drift ranges, HSS-21 showed moderate yielding of the columns earlier than HSS-20 also saw

moderate yielding of the beam prior to brace fracture. For HSS-24 with the modified beam-to-

column connection, the flanges were fully welded but the web was bolts with a shear tab. The

performance of the framing elements show the majority of damage concentrated to the columns.

Minor yielding was observed in both at initial drift ranges but the progression of yielding increased

in the columns only. Moderate yielding in tension and local deformation was observed at the

moderate drift range and moderate yielding with the brace in compression was observed in the

severe range. Only minor yielding was observed in the beams through the end of the test.

HSS-22 was likely the most flexible frame of the nine specimens tested having the un-welded beam-

to-column detail and thin tapered gusset plates. Very little damage of the framing elements was

observed during the test. Minor yielding occurred in the beams at the initial drift range and not in

the columns until the severe drift range, both with the brace in compression.

The frame performance for WF-23 shows significant damage delayed until severe drift. Only

minor damage was observed through initial and moderate drift ranges. Moderate yielding of the

columns with the brace in compression and tension were not observed until the severe drift range.

Finally, severe yielding of the columns and moderate yielding of the beams were observed prior to

interface weld failure at the ultimate drift range.

The heavy beam sections and larger gusset plates used in HSS-25 and HSS-26 inherently increased

the frame stiffness. For both, damage was concentrated to the column rather than the beams.

HSS-25 saw minor yielding with the brace in tension at initial drift levels and moderate yielding at

moderate drift ranges. Also at moderate drift ranges, minor yielding was observed with the brace in

compression. Minor beam yielding was finally observed at severe drift levels as well as severe

yielding and local deformation of the columns. Bolt shearing of the northwest shear tab

connection occurred simultaneously with brace fracture at system failure. For HSS-26, all of the

yielding occurred in the columns within the initial drift range. No damage of the framing element

was observed for HSS-19.

270

Chapter 7: Comparison of Design Parameters

7.1 Introduction

The nine tests within this program are an extension of the previous research by Shawn Johnson

(2004), David Herman (2007), and Brandon Kotulka (2007). This chapter compares the results

from these nine tests with selected tests from the previous 17 in order to evaluate the effect each

design parameters has on overall performance. The design parameters evaluated in this chapter are

as follows

Beam-to-column connection method with thin rectangular gusset plates (HSS-18, HSS-20,

HSS-21, and HSS-24)

Beam-to-column connection method with thin tapered gusset plates (HSS-22)

Wide-flange verses HSS tube brace sections (WF-23)

Bolted connections for SCBFs (HSS-19, HSS-20, and HSS-21)

Net Section Reinforcement Requirement (HSS-25 and HSS-26)

Five specimens from the previous test programs were used for comparison that are either reference

specimen to simulate the current AISC design procedures, or specimen with similar design to those

in this series. A brief description of each is provided below:

HSS-01 and HSS-12

Both designed to simulate the current AISC design specifications and procedures for

rectangular gusset plates

Provide base-line comparisons of specimen performance utilizing current standards

HSS-05

Designed during the development of the Balance Design Procedure and utilized thin

rectangular gusset plates with elliptical clearance at brace ends

CJP welded web and flange beam-to-column connection

Identical gusset plate design as HSS-18, HSS-20, HSS-21, and HSS-24 from this test series

Achieved large ultimate drift levels and is used as an upper bound for comparing

performance for specimens with thin rectangular gusset plates

271

HSS-17

Designed following the BDP and utilized thin tapered gusset plates with elliptical clearance

requirement

CJP welded web and flange beam-to-column connection

Identical gusset plate design as HSS-22

Achieved large ultimate drift levels and is used as an upper bound for comparing

performance for specimens with tapered gusset plates

HSS-11

Designed to evaluate potential for net section fracture of the brace by utilizing larger beam

sections and thicker gusset plates to increase likelihood of occurrence

Identical specimen design as HSS-25 and HSS-26 from this test series, except included net

section reinforcement at brace to gusset plate connection designed per AISC Equation

D3.1 and Table D3.1

Summaries of these specimens including the design information are shown in Table 7.1.1. Peak

performance results for each are summarized in Table 7.1.2. An overview of the design and

experimental results are presented for each of the comparison test specimens within Sections 7.1.1

through 7.1.5.

Table 7.1.1: Design Summary of Reference Specimens for Comparison

272

Table 7.1.2: Peak Performance Summary for Reference Specimens

Comparisons are made between specimens with similar gusset plate designs and geometry to isolate

the relevant design parameter. Section 7.2 considers the variations of beam-to-column connections

for thin rectangular gusset plates by comparing the performances of HSS-18, HSS-20, HSS-21, and

HSS-24 with HSS-05 and the two AISC reference specimens. Section 7.3 compares the

performance of HSS-22 with HSS-17 and the two AISC reference specimens to evaluate the effect

of beam-to-columns connections for thin tapered gusset plates. Specimen WF-23 evaluated the

performance of a wide-flange brace section with thin rectangular gusset plates designed using the

Balanced Design Procedure. The performance of this specimen is compared with HSS-05 and the

two AISC reference specimens in Section 7.4.

Bolted connections as a design parameter are evaluated in section 7.5. The performance of HSS-

19, HSS-20, HSS-21 and HSS-24 are compared to HSS-05 and the AISC reference specimens to

evaluate the benefits and draw-backs of each type of bolted connection for performance and

constructability. Specimens HSS-25 and HSS-26 were the continuation of the investigation for net

section fracture of the brace evaluated previously by Kotulka (2007). The potential for net section

fracture and the necessity of net section reinforcement are discussed in Section 7.6.

7.1.1 Specimen HSS-01 Overview

Specimen HSS-01 was the first specimen tested within the SCBF research at UW and was designed

following AISC specifications and procedures and was intended to be used as a reference specimen

for subsequent test (Johnson 2005). The gusset plate design utilized strength reduction factors and

resulted in ½” thick rectangular gusset plates. The 2t linear clearance was implemented from the

end of the brace and resulted in 34” x 30” dimensions adjacent to the beam and column. The

interface welds was designed for the vertical and horizontal components of the expected brace

force based on the AISC Uniform Force Method, resulting in ¼” fillets welds. Net section

Range Min Max

HSS-01 2.65 -1.64 1.01 -0.95 1.64

Interface

Weld

Fracture

AISC Reference

Specimen

HSS-05 4.96 -3.09 1.87 -0.80 1.76Brace

Fracture

Balanced Designed

Thin Rectangular GP

HSS-11 3.96 -2.34 1.63 -1.00 2.04Brace

Fracture

Thick GP & Heavy

Beam

HSS-12 3.49 -2.10 1.39 -0.90 1.86Brace

Fracture

AISC Reference

Specimen with CJP

Interface Welds

HSS-17 4.94 -2.79 2.15 -0.79 1.77Brace

Fracture

Balanced Designed

Thin Tapered GP

P/Py

ResistanceSpecimen

Drift Ratio, % Failure


273

reinforcing was included for the reduced slotted area of the brace at the gusset plate connection. A

fully welded moment connection was used to connect the beam and column adjacent to the gusset

plates.

The specimen achieved a total drift range of 2.74%, -1.64% in the negative direction and 1.11% in

the positive. The system resisted -191.7 kip with the brace in compression and 330.6 kip with the

brace in tension, for a range of 522.3 kip total. The specimen ultimately failed due to complete

weld fracture of the southwest gusset plate interface welds. The hysteretic response and the gusset

plate connection detail are shown in Table 7.1.3.

Table 7.1.3: Specimen HSS-01 Hysteresis and Gusset Plate Detail

HS

S-0

1

Inelastic behavior primarily occurred in the brace as tensile yielding and buckling. Minor yielding

was also observed in the columns adjacent to the gusset connections. Cracking within the weld

material was observed at both gusset plate connections at early moderate total drift ranges, 1.78%.

The cracks propagated within the weld material and eventually separated the southwest gusset plate

from the framing elements in tension. Cracking of the southwest gusset plate to beam interface

weld is shown in Figure 7.1.1 at 2.45% total drift range. Figure 7.1.2 shows the southwest gusset

plate fully separated from the frame. The gusset plate appears undamaged after the completion of

the test.

274

Figure 7.1.1: Interface Weld Cracking at SW Gusset Plate (1.78% Total Drift Range)

Figure 7.1.2: Weld Fracture at SW Gusset Plate (2.65% Total Drift Range)

Specimen HSS-01 demonstrated the issues with the current AISC design specifications and

procedures. The ½” plate thickness and 2t linear clearance can result in gusset plate designs of

large proportions with increased rotational stiffness. The welds were sized for forces determined

from the Uniform Force Method and as a result showed that they were susceptible to cracking to

accommodate gusset plate rotation as the brace buckled.


HSS-12 was tested to establish a baseline specimen reflecting AISC design specifications and

procedures with increased interface weld capacity (Kotulka 2007). The specimen utilized the

275

identical ½” gusset plate design and geometry using AISC strength reduction factors and the 2t

linear clearance. The interface weld design varied from HSS-01 by incorporating CJP welds to

connect the gusset plate to the framing elements and discourage the potential for weld fracture.

This specimen also utilized CJP welded moment connections to connect framing elements and

included net section reinforcing at the brace to gusset plate connection.

Fracture of the brace at the plastic hinge was the controlling failure mechanism. The specimen

reached positive and negative drift levels of 1.39% and -2.10% for a total drift range of 3.49%.

Positive and negative lateral resistances of 373.3 kip and -180.2 kip were achieved totaling 553.5

kip. Cracking did not occur along the CJP interface welds connecting the gusset plates to framing

elements. The hysteretic response of HSS-12 and the gusset plate detail are shown in Table 7.1.4.


HS

S-1

2

The specimen achieved the desired brace behavior by yielding in tension and buckling in

compression. The brace behavior progressed to inelastic buckling eventually developing the local

deformations of prior hinging at the brace center and fracturing in tension. The gusset plates again

exhibited limited rotational capacity and minimal yielding within the 2t clearance region. Also, no

tensile yielding was observed during the tests. Moderate yielding was observed in the beams and

columns but the majority of inelastic damage occurred within the brace. The performance of this

specimen, as well as HSS-01, are considered the baseline for comparison purposes providing a

lower bound expected limit for ultimate drift capacity.


Specimen HSS-05 was designed during the development of the Balanced Design Procedure with a

thinner gusset plate exceeding the AISC design procedure to encourage yielding in tension. The

resulting 3/8” thick gusset plates incorporated a 8t elliptical clearance to reduce the geometry to

276

25” x 21” at the beam and column. The interface welds were designed for the plastic capacity of

the gusset plates using a β factor of 0.80. The beam-to-column connections were fully welded

moment connections and net section reinforcing was also included at the brace to gusset plate

connection.

The total drift range, 4.96%, achieved by HSS-05 has been considered an upper mark when

assessing performance of subsequent tests. The specimen reached negative and positive drift levels

of -3.09% and 1.87% and resisted -161.0 kip with the brace in compression and 353.5 kip in

tension for a total range of 514.5 kip. The system ultimately failed due to brace fracture at the

plastic hinge but not until after severe tearing of the gusset plate welds had occurred. Table 7.1.5

shows the lateral load verses drift ratio hysteresis and the gusset plate connection detail.


HS

S-0

5

The specimen exhibited inelastic deformation of the brace and gusset plates in tension and

compression, as well moderate yielding and local buckling of the framing elements at larger drift

levels. The gusset plate connections experienced moderate yielding (Y2) in tension and developed

elliptical yield patterns extending from the reentrant corners around the end of the brace as out-of-

plane buckling occurred. The northeast gusset plate is shown in Figure 7.1.3.

277

Figure 7.1.3: HSS-05 NE Gusset Yielding (4.96% Total Drift Range)

The progression of yielding of the gusset plates as they rotated out-of-plane was limited by the

onset of ductile tearing of the interface welds. Cracking began at the free edges within the weld

material and propagated inward. This allowed the plate to deform as the brace buckled without

inducing flexural strain through the elliptical region as typically seen in both previous and

subsequent tests with similar gusset plate designs. Severe cracking extended completely through

the northeast gusset to column weld and total crack length at the northeast connection reached

54% to the total interface weld lengths, beam at column combine. Total cracking of the southwest

interface welds exceeded 20%. Cracking of the northeast gusset to column weld is shown in Figure

7.1.4. Regardless, at maximum drift levels, the local deformation and strain accumulation at the

brace plastic hinge produced tearing of the tubular section resulting in brace fracture prior to

separation of the gusset plate from the framing elements.

Figure 7.1.4: Cracking of NE Gusset to Column Weld (4.96% Total Drift Range)

278

HSS-05 achieved a larger total drift range than any other specimen test within this program and

exhibited yielding beyond the brace and into the gusset plates and framing elements, as desired by

the Balanced Design Procedure. For these reason, HSS-05 is used in comparisons of performance

parameters as the upper bound but obtainable goals for the performance of the specimens within

this test series.

7.1.4 HSS-17 Specimen Overview

The 3/8” tapered gusset plates for HSS-17 were designed following the Balanced Design Approach

and used an 8t elliptical clearance at the end of the brace for rotation. The interface welds were

design for the plastic capacity of the gusset plate using a β factor of 0.65, resulting in 3/8” fillet

welds on each side. The framing elements were connected with CJP welded moment connection

adjacent to the gusset plates. The brace to gusset plate connection did not included net section

reinforcing plates.

The specimen performed well in that the maximum total drift range achieved was close to that of

HSS-05. In the positive direction, 2.15% drift was achieved, and -2.79% in the negative direction,

for a total drift range of 4.94%. The maximum positive and negative resistances were 355.0 kip and

159.0 kip, for a total range of 514.0 kips. The controlling failure mechanism of the system was

brace fracture at the plastic hinge in tension. The lateral load verses drift ratio hysteresis and the

gusset plate connection details are shown in Table 7.1.6.


HS

S-1

7

Inelastic behavior was observed in the brace in the form of buckling and tensile yielding, in the

beams and columns and, most notably, in the gusset plates. The 3/8” tapered gusset plates had

lower in-plane and out-of-plane stiffness than rectangular plates of the same thickness which

resulted in severe yielding and deformation observed during the test. Severe cracking at both the

279

NE and SW column interface welds within the based material (HAZ) was also observed at severe

drift levels. Figure 7.1.5 shows the level of gusset plate yielding and the extent of weld damage

observed at the maximum negative drift level prior to brace fracture the following tension cycle.

Figure 7.1.5: Severe Gusset Plate Yielding and Severe Weld Damage (-2.79%)

Moderate yielding of the beams and columns in the area adjacent to the gusset reentrant corner was

observed at severe drifts levels but the majority of visible damage occurred within the brace and

gusset plates. Cracking of the base material at the NE column interface weld propagated to 15” at -

2.79% drift, which is 43% percent of the total weld length connecting the gusset plate to the frame.

The damage at the brace plastic hinge lead to fracture before interface weld cracking continued but

this suggests an inherent issue with thin, tapered gusset plates. The smaller geometry, increased

deformation, and decreased lengths of interface weld increases the potential for weld damage

(Kotulka 2007). However, HSS-22 showed that large ultimate drifts are achievable with tapered

gusset plates and the performance will be held as the upper bound for comparison with HSS-22.


The design of specimen HSS-11 was based on the parametric nonlinear study evaluating varies

beam sizes and gusset plate thickness on the performance of the brace (Herman 2007). HSS-11

utilized a heavier beam section, W16x89, than what had been tested in previous specimen with a

7/8” thick gusset plate because the increased frame and gusset plate stiffness resulted in larger

inelastic demand on the brace and early brace fracture. This specimen is included in the

280

comparison to HSS-25 and HSS-26 because the design is identical with the exception of the

inclusion of net section reinforcing at the brace to gusset connection.

The specimen ultimately failure due to fracture at the brace midpoint hinge and achieved a total

drift range of 3.96%. The positive and negative resistances observed were 410.9 kip and -200.3 kip,

for a total of 611.2 kip. The lateral load verses drift ratio hysteresis and the gusset plate connection

detail are shown in Table 7.1.7


HS

S-1

1

Inelastic behavior for HSS-11 occurred in the brace as bucking in compression and yielding in

tension. The 8t elliptical clearance at the brace ends provide sufficient area for the gusset plates to

rotation out-of-plane but minimal yielding was observed in tension. The column section also

experience significant yielding and deformation in the region adjacent to the gusset plate reentrant

corners.

7.2 Beam-to-Column Connections for Thin Rectangular Gusset Plates

This section compares the performances of specimens with 3/8” thin rectangular gusset plates

(TRGP) within this test series (HSS-18, HSS-20, HSS-21, and HSS-24) with those of the AISC

reference specimens, HSS-01 and HSS12, and with HSS-05. Specimen HSS-05 utilized identical

gusset plate thickness and geometry as the four TRGP specimens from this series, while the two

AISC reference specimens are included to give a comparison to specimen with gusset plates

designed by the current procedure. The objective of this section is to evaluate the effect the beam-

to-column connection detail has as the primary design parameters with regards to system

performance.

281

Evaluation of the system responses is compared in Section 7.2.1. The response of the brace

diagonal, including the brace and gusset plate responses in compression and tension, is evaluated in

Section 7.2.2. Comparison of the frame response is discussed in 7.2.3 followed by comparison of

energy dissipation in Section 7.2.4. Data interpretation and analysis methods for generating plots

are outlined in Chapter 6 and reference for clarity.

7.2.1 TRGP System Response Comparison

The peak performance values for TRGP specimens are compared in Table 7.2.1. The table

includes ultimate drift levels and total drift range as a measure of system ductility. The maximum

positive and negative resistances of the system are normalized by , the theoretical lateral force

associated with tensile brace yielding, shown in Equation (1.2.1).

(7.2.1)

Maximum resistance of the system typically occurs after the brace buckles in compression and

yields in tension, and is largely attributed to the contribution of frame resistance and the ability of

the brace to maintain compressive resistance post-buckling. For these reasons, maximum

resistance is limited as a parameter for assessing performance of the system. However, it does

provide a general value for comparing system resistance at a glance. Sections 7.2.2 and 7.2.3

investigate brace and frame resistances over the full nonlinear response of the system, respectively.

Table 7.2.1: Peak Performance Summary Comparison for TRGPs

Range Min Max

HSS-01 2.65 -1.64 1.01 -0.95 1.64

Interface

Weld

Fracture

Simulate AISC Design

with Fillet Welds

HSS-05 4.96 -3.09 1.87 -0.80 1.76Brace

Fracture

CJP Welded Bm/Col

Thin Rectangular GP

HSS-12 3.49 -2.10 1.39 -0.90 1.86Brace

Fracture


with CJP Interface

Welds

HSS-18 4.19 -2.59 1.60 -0.76 1.59Brace

Fracture

Bolted Shear Plate

Connection

HSS-20 3.97 -2.28 1.69 -0.93 1.59Brace

Fracture

18 Bolted Beam End

Plate Conn.

HSS-21 4.14 -2.55 1.59 -0.81 1.73Brace

Fracture

14 Bolted Beam End

Plate Conn.

HSS-24 4.44 -2.50 1.94 -0.80 1.69Brace

Fracture

Bolted Web Weld

Flange Beam/Col Conn.

P/Py

ResistanceSpecimen

Drift Ratio, % Failure


282

The four specimens from the current test series reached total drift ranges less than HSS-05 but

greater than both the AISC reference specimens, HSS-01 and HSS-12. Specimen HSS-24, with the

welded flanges and bolted web plate beam-to-column connection, was most similar to HSS-05 and

achieved the largest total drift range, 4.44%, of the four within this test series. A slight loss in

maximum resistance was observed in the positive direction with the modified beam-to-column

connection.

Both the specimens with bolted beam end plate connections saw significant reductions in total drift

range than what was achieved with the CJP welded connection in HSS-05. The 14-bolt

configuration of HSS-21 achieved 4.14% total drift range followed by HSS-20, which achieved

3.97%. The maximum resistance of HSS-20 was notably larger with the brace in compression and

smaller while in tension than HSS-05. Specimen HSS-21, however, showed comparable maximum

resistances in both directions. Specimen HSS-18 utilized the bolted shear plate beam-to-column

connection and resulted in a lower ultimate drift capacity, 3.96% total drift range. The maximum

positive and negative resistances were both less than specimen HSS-05.

The enveloped lateral force verse drift ratio response in the positive directions is compared in

Figure 7.2.1. The lateral force is normalized by , the nominal yield force of the brace. The

specimens utilizing the bolted shear plate connection and the 18 bolt beam end plate connection

exhibited the least resistance after yielding occurred. The 14 bolt beam end plate connection and

the bolted web CJP welded flange connection showed greater resistance post yielding, but were still

below the maximum positive resistance achieved by the CJP welded beam-to-column connection in

HSS-05. All four connections allowed the specimens to retain or continue increasing resistance

prior to failure, where as HSS-05 lost positive resistance at larger drift levels.

283

Figure 7.2.1: TRGP Load vs. Drift (Positive Envelop)

The enveloped load verses drift ratio response is shown in the negative direction in Figure 7.2.2.

As expected, the larger geometry and increased thickness, ½ in., of the AISC reference specimens

resulted in greater resistances at brace buckling, with the exception of the 18 bolt beam end plate

connection of HSS-20 which exhibited a larger elastic stiffness and resistance at brace buckling than

specimens with similar gusset plate designs. The specimens with the bolted shear plate connection

(HSS-18), the 14 bolt beam end plate connection (HSS-21), and the bolted web CJP welded flange

connection (HSS-24) exhibited similar system resistance at buckling. The bolted shear plate

connection saw the lowest resistance post buckling for a given negative drift level.

284

Figure 7.2.2: TRGP Load vs. Drift (Negative Envelop)

7.2.2 TRGP Brace and Gusset Plate Comparison

This section compares the responses of the brace and gusset plates in compression and tension.

Section 7.2.2.1 evaluates the brace forces by comparing the buckling capacity of the brace and the

full nonlinear brace response. The physical behaviors of the brace diagonal in compression are

compared in Section 7.2.2.2 including out-of-plane brace displacement, gusset plate rotation in

buckling, and residual out-of –plane displacement of the brace and gusset plates. Brace and gusset

plate elongation in tension are compared in Section 7.2.2.3.

7.2.2.1 TRGP Brace Force Comparison

Comparisons of the brace buckling capacities are shown in Table 7.2.2. The critical buckling force

in the brace, , was determined using strain gauge data as described in Chapter 6. Also included

in the table are the design nominal buckling capacities as calculated in the AISC Design

Specifications, , the effective length factor, , based on the experimental critical buckling load,

, and the experimental slenderness ration, .

285

Table 7.2.2: TRGP Experimental Buckling Capacity Comparison

The critical buckling loads determined experimentally for the five specimens with similar brace

lengths and gusset plate designs varied between 174.8 kip and 199.9 kip. This resulted in effective

length factors between 0.86 and 1.01, which is supports the use of an effective length factor of 1.0

in the design. The bolted shear plate connection has less in-plane stiffness than the CJP welded

connection or the bolted web CJP welded flange connection but exhibited a larger buckling

capacity than both. However, the 18 bolt beam end plate connection has greater in-plane stiffness

than the bolted shear tab connection and resulted in a larger buckling capacity, 199.9 kip. This

comparison suggests that the beam-to-column connection method and in-plane stiffness do not

affect the buckling capacity of the brace.

The full nonlinear brace responses are shown in Table 7.2.3 for the TRGP specimens and the two

AISC reference specimens. Brace force verses diagonal elongation, work point to work point, are

plotted. Brace force is normalized by the nominal tensile yield force of the brace, ,

calculated in Equation (7.2.2). The brace forces for the test specimens in this series were calculated

as described in Chapter 6. The full strain gauge record for the AISC reference specimen with fillet

interface welds and the TRGP specimen with CJP welded beam-to-column connections were not

available and the brace force was extracts by subtracting the column shears from the applied lateral

load to determine the horizontal load resisted by the brace.

(7.2.2)

Experimental AISC

Pcr Pn=FcrAg

AISC w/ Fillet 209.5 200.7 1.04 0.94 67.4

AISC w/ CJP 195.7 200.7 0.98 1.04 74.5CJP Welded

Beam/Col Conn174.8 176.2 0.99 1.01 85.0

Shear Plate

Beam/Col Conn188.6 176.2 1.07 0.96 81.3

18 Bolt Beam End

Plate199.9 176.2 1.13 0.86 72.4

14 Bolt Beam End

Plate182.7 176.2 1.04 0.96 81.1

Bolted Web CJP

Welded Flange 175.0 176.2 0.99 1.01 84.9

Design ParameterExp/AISC

Pcr/Pn

Experimental

Effective Length

Factor, K

Experimental

Slenderness

Ratio (Kl/r)

286

Table 7.2.3: TRGP Brace Responses A

ISC

w/

Fil

let

AIS

C w

/C

JP

CJP

Weld

ed

Beam

/C

ol

Co

nn

.

Sh

ear

Pla

te

Beam

/C

ol

Co

nn

.

18 B

olt

ed

Beam

En

d

Pla

te

14 B

olt

Beam

En

d

Pla

te

Bo

lted

Web

CJP

Fla

ng

e C

on

n.

The degradation of brace compressive capacity is compared for the TRGP specimens by plotting

the brace compressive force verses total drift range in Figure 7.2.3. To be consistent, the brace

force in this comparison is determined by subtracting the column shears from the applied lateral

load to determine the horizontal force component in the brace for all five specimens. The AISC

reference specimens are not included in this comparison to focus on the influence the beam-to-

column connection design parameter has on the compressive capacity of the brace.

287

Figure 7.2.3: TRGP Brace Compression Degradation Comparison

The bolted shear plate, bolted web CJP welded flange, and the CJP welded beam-to-column

connections resulted in similar brace compressive capacity degradation. The bolted beam end plate

connection both degraded more significantly at smaller drift ranges, less than 1.5% total, but still

achieve similar minimum capacities. The ratios of capacity remaining immediately prior to

fracture, , over the critical buckling load, , are shown in Table 7.2.4. The specimens with

the 18 bolt and 14 bolt beam end plate connections retained the largest compressive capacity

compared to their critical buckling load, 0.28 and 0.24, respectively. The CJP welded beam-to-

column connection retained the least, 0.20, but all were between 0.20 and 0.28 and the behavior

was similar in all cases.

Table 7.2.4: TRGP Brace Compressive Capacity Degradation Ratio

CJP Welded Beam/Col

Conn164.2 32.3 0.20

Shear Plate Beam/Col

Conn182.3 43.6 0.24

18 Bolt Beam End

Plate207.7 58.9 0.28

14 Bolt Beam End

Plate168.3 43.7 0.26

Bolted Web CJP

Welded Flange 173.7 41.3 0.24

Pmin/*PcrPcr*

* critical buckling load calculated by subtracting column shears

from applied lateral load for horizontal componet resisted by

brace

Design Parameter Pmin

288

7.2.2.2 TRGP Brace Displacement and Gusset Plate Rotation Comparisons

Brace displacement and gusset plate rotations are compared in this section to evaluate the behavior

of the brace in compression and determine how the beam-to-column connection detail influences

the brace response. The out-of-plane behaviors of the brace center and gusset connections reflect

the shape and curvature of the buckled brace. Larger curvatures at smaller drift ranges increase the

strain at the brace plastic hinge, potentially limiting the life of brace.

The displacement at the brace midpoint for TRGP specimens are compared in Figure 7.2.4

including the AISC reference specimen HSS-12. The values for displacement have been plotted as

a percentage of the total brace length. The midpoint displacement values for specimen with the

CJP welded beam-to-column connection (HSS-05) are limited because of loss of instrumentation to

a total drift range of 4.30%, along reached 4.96%. Similarly, the 18 and 14 bolt beam end plate

connection specimens (HSS-20 and HSS-21) were plotted up to a total drift range of 3.30% and

1.96% but reached 3.96% and 4.14% total drift range, respectively. Data for the bolted shear plate

connection (HSS-18) was also limited to 3.41% but achieved 4.19% total.

Figure 7.2.4: TRGP Brace Out-of-Plane Displacement Comparison

The center displacements of TRGP specimens from this test series and the CJP welded beam-to-

column connection were similar for a given drift level, with the exception of the 18 bolt beam end

plate connection (HSS-20), which were notably larger. The AISC reference specimen with CJP

interface welds (HSS-12) exhibited somewhat different behavior by increasing more for a given

increment of total drift range beyond 1.25% and resulted in larger displacement at the maximum

drift range of 3.49%.

289

Gusset plate rotations over total drift range for the northeast connection are shown in Figure 7.2.5.

The southwest gusset plate rotations are not shown and assumed to be similar. Stiffer gusset plates

connections limit rotation at the brace ends resulting in more severe curvature of the brace than

more flexible connections.

Figure 7.2.5: TRGP NE Gusset Plate Rotation Comparison

The gusset plate rotations for the TRGP specimens in this test series were nearly identical for a

given drift range with the bolted shear plate connection (HSS-18) achieved slightly less rotation at

failure. The specimen with CJP welded beam-to-column connection (HSS-05) exhibited smaller

rotations at a given drift range less than 3.25% but increased at larger drift ranges achieving larger

maximum rotations than any other specimen. As shown in Figure 7.2.4, the 18 bolt beam end plate

connection (HSS-21) exhibited the largest midpoint displacements but only achieved comparable

gusset rotations at a given drift range. This would have resulted in a more pinched brace shape and

increased strain accumulation at the plastic hinge, limiting the life of the brace and fracturing at

lower total drift ranges. The same can be said for the AISC reference specimen with CJP interface

welds (HSS-12). Larger center out-of-plane displacements and smaller gusset plate rotations

resulted in increased curvature and more strain at the plastic hinge location which is consistent with

the observed early brace fracture and limited ultimate drift capacity.

The residual out-of-plane displacements at the brace center and the gusset rotations are shown in

Figure 7.2.6 and Figure 7.2.7, respectively. These illustrate the level of plastic deformations that

occurred over the brace diagonal for a given drift level. Specimens that experienced larger plastic

290

deformation required additional force to straighten the brace in tension and increase the level of

plastic strain accumulated at the brace plastic hinge.

Figure 7.2.6: TRGP Residual Brace OOP Displacement Comparison

Figure 7.2.7: TRGP Residual NE Gusset Plate Rotation Comparison

The specimen with the 18 bolt beam end plate connection (HSS-20) and the AISC reference

specimen (HSS-12) both showed the largest residual displacement of the brace and the smallest

residual gusset plate rotations. This combination resulted in increase curvature in the braced shape

and larger plastic strain at the brace center. The AISC reference specimen achieved significantly

291

smaller total drift range than any of the TRGP specimens and the 18 bolt beam end plate

connection specimen was the lowest of the five TRGP specimens. The specimen having the bolted

web and CJP welded flanges (HSS-24) showed the opposite behavior having smaller brace

displacement and larger gusset plate rotations. This specimen achieved the largest total drift range

of the four TRGP specimens in this test series.

7.2.2.3 TRGP Brace and Gusset Plate Elongation Comparisons

This section compares the behavior of the brace and gusset plates in tension. Brace elongation for

the five TRGP specimens are plotted over total drift range in Figure 7.2.8. Brace elongation is

normalized by the original brace length from edge of gusset to edge of gusset and given as a unit-

less measure, or in./in. The data for the CJP welded beam-to-column connection specimen (HSS-

05) is incomplete in that it was not available beyond 4.30% total drift range. The elongation of the

braces are well banded for all of the TRGP specimens with the exception of the bolted shear plate

connection (HSS-18), which exhibited less elongation at a given drift range.

Figure 7.2.8: TRGP Brace Elongation Comparison

Elongation occurring in the gusset plates was determined by subtracting the brace elongation from

the diagonal displacement, work point to work point, and normalized by the length from edge of

gusset plate to the beam/column work point. Additional description of the analysis method is

available in Chapter 6. The gusset plate elongations over total drift range are compared for the

TRGP specimen in Figure 7.2.9. It can be seen that the gusset plate elongation for the bolted shear

plate connection specimen (HSS-18) were greater for a given drift range greater than 2.0%. Also,

292

gusset plate elongation for the 14 bolt beam end plate beam-to-column connection (HSS-21) was

lower than the other TRGP specimens for a given drift level.

Figure 7.2.9: TRGP Gusset Plate Elongation Comparison

Figure 7.2.10 compares the elongation of both the brace and gusset plates. The line labeled “One-

to-One” is provided to designate a balance between brace and gusset plate elongation. The BDP

encourages gusset plate yielding beyond the brace to increase the total ductility of the system and

this comparison illustrates the balance between brace and gusset plate elongation.

Figure 7.2.10: TRGP Brace vs. Gusset Plate Elongation

293

The specimen with CJP welded beam-to-column connections (HSS-05) and the specimen with

bolted shear plate and CJP welded flanges (HSS-24) reached the largest total drift ranges, 4.96%

and 4.44% respectively, and both show elongation balanced between the brace and the gusset

plates, and slightly more brace elongation at larger drift ranges. The specimen having bolted web

CJP welded flange connections (HSS-18) exhibited a response more dominated by gusset plate

elongation and achieved a lower maximum drift range, 4.19%. The 14 bolt beam end plate

connection (HSS-21) showed the opposite having more elongation occurring in the brace

compared to the gusset plates and also achieved a smaller drift range, 4.14%.

7.2.3 TRGP Frame Response Comparison

The frame responses are compared using the calculated columns shears determined in Chapter 6 to

evaluate the resistance of the frame with varying connection details. Column strain gauge data

from AISC reference specimen with CJP interface welds (HSS-12) contained errors and was not

included in this comparison. Figure 7.2.11 and Figure 7.2.12 display frame resistances verses story

drift ratio in the positive and negative directions, respectively. The frame resistance, is normalized

by , the theoretical shear force to achieve the plastic moment of the column, , at the beam-

to-column intersection.

Figure 7.2.11: TRGP Frame Resistance vs. Drift Comparison (+ Direction)

294

Figure 7.2.12: TRGP Frame Resistance vs. Drift Comparison (- Direction)

Observations of the frame resistance comparison are below:

The AISC reference specimen with fillet interface welds exhibited higher resistances in

both directions because of the CJP welded beam-to-column connection and larger, stiffer

gusset plate geometry.

Both specimens with beam end plate connections (HSS-20 and HSS-21) showed higher

resistances comparatively. The 1” thick end plate extended the full length of the beam

and gusset plate shortening the unbraced length of the column and increased the stiffness

of the frame.

The specimen with bolted shear plate beam-to-column connections (HSS-18) was

intuitively less stiff than the other TRGP specimens and exhibited the lowest resistance in

the negative direction and lower resistance in the positive direction comparatively.

The bolted web CJP welded flange connection (HSS-24) resulted in similar frame

resistance as the CJP welded beam-to-column connection of HSS-05 in the negative

direction but lower resistance in the positive.

Comparisons of the horizontal shear resisted by the brace and the frame are shown in Table 7.2.5

as a percentage of the total lateral shear. Values at brace yielding and at the ultimate drift levels are

presented in Table 7.2.6. As discussed in Chapter 6, larger total drift ranges were obtained when

295

the distribution of system resistance was approximately 75-80% brace and 20-25% frame at the

ultimate negative drift levels. Typically when brace contribution declined rapidly at smaller negative

drift levels, brace fracture occurred earlier and smaller total drift ranges were achieved.

Table 7.2.5: TRGP Distribution of Resistance

AIS

C w

/F

ille

t

CJP

Weld

ed

Beam

/C

ol

Co

nn

.

Sh

ear

Pla

te

Beam

/C

ol

Co

nn

.

18 B

olt

ed

Beam

En

d

Pla

te

14 B

olt

ed

Beam

En

d

Pla

te

Bo

lted

Web

CJP

Fla

ng

e

296

Table 7.2.6: Summary of TRGP Resistance Distribution

7.2.4 TRGP Energy Dissipation

The energies dissipated by the TRGP specimens are compared in this section. Table 7.2.7 displays

the total energy dissipated as well as energy dissipated by the brace diagonal and by the frame

individually. The total energy is calculated using the applied lateral load and the horizontal

displacement of the frame as described in Section 6.6, which also outlines the equations used to

determine energy dissipated by the brace diagonal and by the frame. Energy dissipated by the brace

diagonal considers the inelastic responses of both the brace and the gusset plates and is calculated

using the brace force from Section 6.3.1.2 and the axial deformation over the brace diagonal, work

point to work point. Frame energy dissipation is calculated using the total column shears as

determined in Section 6.4.2 and the frame horizontal displacement. The total energy dissipated by

the system, brace diagonal and by the frame are summarized in Table 7.2.8 for each TRGP

specimen and the AISC reference specimens.

Table 7.2.7: TRGP Energy Dissipation

AIS

C w

/F

ille

t

AIS

C w

/C

JP


AISC w/ CJP 78.0 22.1 80.3 19.7 36.0 64.0 73.2 26.8CJP Welded

Beam/Col

Conn 80.3 19.7 83.8 16.2 19.7 80.3 68.2 31.8Shear Plate

Beam/Col

Conn 82.5 17.5 83.6 16.4 16.1 83.9 76.9 23.1

18 Bolt Beam

End Plate 79.5 20.5 86.7 13.4 18.3 81.7 73.4 26.6

14 Bolt Beam

End Plate 79.2 20.8 83.6 16.4 13.2 86.8 71.6 28.5Bolted Web

CJP Welded

Flange 84.9 15.2 86.8 13.2 29.3 70.7 74.6 25.4

Specimen

% Resistance at Brace Yield/Buckling % Resistance at Ultimate

-Drift% +Drift% -Drift% +Drift%

297

CJP

Weld

ed

Beam

/C

ol

Co

nn

.

Sh

ear

Pla

te

Beam

/C

ol

Co

nn

.

18 B

olt

ed

Beam

En

d

Pla

te

14 B

olt

Beam

En

d

Pla

te

Bo

lted

Web

CJP

Fla

ng

e C

on

n.

Table 7.2.8: Summary of Energy Dissipation for TRGP

All of the TRGP specimens within the current test series (HSS-18, HSS-20, HSS-21, and HSS-24)

showed similar total energy dissipation and distribution between energy dissipated by the brace

diagonal and the frame. These specimens dissipated between 4825 kip-in and 5237 kip-in of

kip-in % kip-in %

AISC w/ Fillet 3334 3181 95 153 5 2.65

AISC w/ CJP 3219 N/A N/A N/A N/A 3.49CJP Welded

Beam/Col Conn6941 5451 79 1490 21 4.96

Shear Plate

Beam/Col Conn5000 3848 77 997 20 4.19

18 Bolt Beam End

Plate5019 3824 76 1044 21 3.97

14 Bolt Beam End

Plate4825 3891 81 976 20 4.14

Bolted Web CJP

Welded Flange 5237 4161 79 972 19 4.44

Design Parameter

Total Energy

Dissipated,

kip-in



Range, %

Frame Diagonal

Energy Dissipation

298

energy, with 76%-81% occurring over the brace diagonal. The specimen with CJP welded beam-

to-column connections (HSS-05) dissipated significantly more total energy, 6941 kip-in, with similar

distribution between brace diagonal and frame as the current specimens.

All of the TRGP specimens showed significant improvement in energy dissipation compared to the

two AISC reference specimens (HSS-01 and HSS-12). The AISC designed reference specimen with

fillet interface welds (HSS-01) dissipated 3334 kip-in of total energy with 95% over the brace

diagonal, whereas the TRGP specimen with CJP welded beam-to-column connections (HSS-05)

showed a 108% improvement in energy dissipation with the more compact, thinner gusset plate

using the Balance Design Procedure resulting in a more ductile response of the system. The TRGP

specimen with bolted web CJP welded flange connections (HSS-24) resulted in 24.5% decrease in

total energy dissipation compared the CJP welded web and flange connection, 23.7% decrease from

the brace diagonal and most notably 34.8 % less energy dissipated through the frame. The bolted

shear plate beam-to-column connection (HSS-18) saw 28.0% less total energy dissipated by the

system than the CJP welded connection. The brace diagonal and frame dissipated 29.4% and

33.1% less energy, respectively.

The 18 bolt and 14 bolt beam end plate connections (HSS-20 and HSS-21) resulted in 50.5% and

44.7% improvements in energy dissipated compared to the AISC reference specimen with fillet

interface welds (HSS-01), but could not achieve similar levels as the TRGP specimen with CJP

welded beam-to-column connections. The 18 bolt configuration saw a 27.7% loss in total energy

dissipation and the 14 bolt configuration lost 30.5%.

7.2.5 TRGP Summary

This comparison has shown that modifications from the CJP welded beam-to-column connection

utilized in HSS-05 resulted in reductions in ultimate drift capacity and energy dissipation, but not

necessarily in total resistance. All TRGP specimen is this test series achieve larger total drift ranges

and more desirable brace behavior that either AISC reference specimens. The specimen with the

bolted web CJP welded flange connection (HSS-24) showed the closest similarity in response to full

CJP welded beam-to-column connection (HSS-05) by achieving the larger total drift range and

dissipating more energy than the others but the decrease in these important performance

parameters was significant by only modifying the beam web connection to the column. The bolted

shear plate beam-to-column connection (HSS-18) resulted in more significant reduction in ultimate

drift capacity and total resistance of the system, which can be attributed to the decrease in frame

stiffness and early brace fracture.

299

The stiffness of the beam-to-column section also appears to influence the level of residual

deformation of the brace and gusset plate out-of-plane. Under cyclic loading, larger strain demands

are required to straighten the brace in tension once plastic deformation has occurred, increasing the

total plastic strain accumulated at the brace center. This could potentially lead to early brace

fracture at smaller drift ranges as illustrated for the bolted shear plate connection (HSS-18), which

saw larger residual out-of-plane displacements than the bolted web CJP flange connection (HSS-24)

and failed at a lower drift range, 4.19% compared to 4.44%.

Both bolted end plate connection (HSS-20 and HSS-21) exhibited decreases in ultimate drift

capacity and energy dissipation compared to the CJP welded beam-to-column connection (HSS-

05), but exhibited increased stiffness and maximum resistance. The response of the frame with the

bolted connection, resistance verses drift ratio, was also comparable to that of HSS-05 and achieved

larger resistances at a given drift level.

7.3 Beam-to-Column Connections for Thin Tapered Gusset Plates

Thin tapered gusset plates (TTGP) have shown the ability to achieve large ultimate drift capacities

and also reduce the amount of material and welding time during construction. The geometry of the

tapered plates allows for free rotations as the brace buckles out-of-plane but also reduces the

buckling capacity of the brace by increasing the effective slenderness ratio. Specimen HSS-17 and

HSS-22 utilized the identical gusset plate designs, but HSS-17 consisted of CJP welded web and

flanges, rather than bolted shear plate connection in HSS-22, adjacent to the gusset plates. The

specimen with CJP welded beam-to-column connections (HSS-17) achieved an ultimate drift range

of 4.94% and is considered the upper bound performance for specimen with tapered gusset plates.

This section evaluates the performance of HSS-22 having the bolted shear plate beam-to-column

connection by comparing it to HSS-17. The two AISC reference specimens (HSS-01 and HSS-12)

are also included in the some of the comparisons to illustrate improved behavior of the system by

utilizing the Balance Design Procedure for TTGP specimens.

Comparison of the system response is discussed first in Section 7.3.1 followed by a comparison of

the response of the brace and gusset plates in Section 7.3.2. The response of the framing elements

are compared in Section 7.3.3 followed by a comparison of energy dissipation for TTGP specimen

in Section 7.3.4.

300

7.3.1 TTGP System Response Comparison

The system responses were evaluated by comparing the lateral load verses drift response for the

two TTGP specimens and the two AISC reference specimens. Table 7.3.1 summarizes the peak

values of drift and maximum lateral load resisted, and includes a description of the controlling

failure mechanism and design parameter for each specimen. The resistance is normalized by the

theoretical lateral force associated with brace yielding calculated in Equation (1.2.1).

Table 7.3.1: Peak Performance Summary for TTGP

Both TTGP specimens achieved larger drift ratios than either AISC reference specimen. The

TTGP specimen with CJP welded beam-to-column connections (HSS-17) achieved a total drift

range of 4.94%, which is 41.5% greater than the AISC reference specimen with CJP interface welds

(HSS-12). The total drift range of the bolted shear plate connection specimen (HSS-22), 3.98%,

increased by only 14.0% compared to HSS-12. Both TTGP specimen exhibited smaller maximum

negative resistance than the AISC reference specimens and the specimen with the bolted shear

plate connections (HSS-22) also exhibited lower maximum positive resistance than any test in this

comparison.

The enveloped values for the lateral load verses drift ratio response were used to compare the

global response of the system. Figure 7.3.1 compares the global responses of the two TTGP

specimens and the two AISC reference specimens in the positive direction. The responses in the

negative direction are shown in Figure 7.3.2.

Range Min Max

HSS-01 2.65 -1.64 1.01 -0.95 1.64

Interface

Weld

Fracture


with Fillet Welds

HSS-12 3.49 -2.10 1.39 -0.90 1.86Brace

Fracture


with CJP Interface

Welds

HSS-17 4.94 -2.79 2.15 -0.79 1.77Brace

Fracture

CJP Welded Bm/Col

Thin Tapered GP

HSS-22 3.98 -2.48 1.50 -0.66 1.50Brace

Fracture

Bolted Shear Plate

Conn Thin Tapered GP

Resistance

P/Py

Failure

MechanismDiscriptionSpecimen

Drift Ratio, %

301

Figure 7.3.1: TTGP Load vs. Drift (Positive Envelope)

The TTGP specimens both exhibit initial tensile yielding at relatively small positive drift levels,

approximately 0.5%, but continued to increase resistance as drift increased. The AISC reference

specimens, especially the specimen with fillet interface welds (HSS-01), retained stiffness until a

more clearly defined point of yielding occurred at approximately 1.30%. The TTGP specimen with

CJP welded beam-to-column connections.

Figure 7.3.2: TTGP Load vs. Drift (Negative Envelope)

302

In the negative direction, the point of brace buckling can be seen followed by the nonlinear

response of the system with the brace in compression. Both TTGP specimens exhibited similar

behavior post buckling, maintaining or slightly decreasing resistance at larger drift levels. The CJP

welded beam-to-column connection (HSS-17) exhibited both a greater resistance at buckling and

maintained a larger resistance until failure than the bolted shear plate connection (HSS-22). The

AISC reference specimens exhibited greater buckling capacity and compressive resistance over the

nonlinear response than either TTGP specimen, which is expected considering the larger gusset

plates and stiffer brace end supports.

7.3.2 TTGP Brace and Gusset Plate Comparison

The response of the brace and gusset plates for the two TTGP specimens (HSS-17 and HSS-22)

are compared in this section to evaluate the effect the beam-to-column connection between has on

brace and gusset plate behaviors. The two AISC reference specimens are including in some of the

comparisons to illustrate how the behavior of the thin tapered gusset plates differ and are more

beneficial than the gusset plates design to simulate current AISC procedures.

The brace forces and nonlinear brace response are compared in Section 7.3.2.1. Comparison of the

brace out-of-plane displacement and gusset plate rotations are included in Section 7.3.2.2. Brace

and gusset plate elongation in tension are compared in Section 7.3.2.3.

7.3.2.1 TTGP Brace Force Comparison

Comparisons of the brace force and nonlinear brace responses are evaluated in this sub-section.

First, the critical buckling capacity of the brace, , was determined from strain gauge data directly

on the brace. Values obtained from the tests were compared to the design nominal buckling

capacity, , in Table 7.3.2. Also included in this table is the ratio of the experimental critical

buckling load over the nominal design value and the experimental effective length factor, , and

slenderness ratio, .

303

Table 7.3.2: TTGP Experimental Buckling Capacity Comparison

The buckling capacity of the brace for the TTGP specimen with CJP welded beam-to-column

connections (HSS-17) was significantly larger than the nominal design value. The resulting

effective length factor, , equaled 0.81, which is surprising considering the increased flexibility of

the brace end connection because of the thin gusset plates and tapered geometry compared to the

effective length factors observed with the AISC reference specimens. The bolted shear plate

specimen exhibited an experimental buckling capacity equal to the design value and equal to 1.0.

The nonlinear bracer response is plotted for the two TTGP specimens in Table 7.3.3. The

hysteretic plots were created using the brace strain gauge data and the elongation over the brace

diagonal, work-point to work-point, as described in Chapter 6, Section X 6.3.2.1.

Table 7.3.3: TTGP Brace Response

CJP

Weld

ed

Beam

/C

ol

Co

nn

.

Sh

ear

Pla

te

Beam

/C

ol

Co

nn

.

A comparison of the nonlinear response of the brace in compression was made to evaluate the

degradation for brace capacity post buckling. Figure 7.3.3 shows brace compression over total drift

range and illustrates the level of force degradation exhibited by the two TTGP specimens with

different beam-to-column connection types. The compressive force is normalized by the design

nominal buckling capacity. Again, the nonlinear brace force used for this comparison was

calculated by subtracting the columns shear described in Chapter 3, Section 6.4.2 from the applied

lateral load to the system. Table 7.3.4 is provided summarizing the compressive response, including

Experimental AISC

Pcr Pn=FcrAg

AISC w/ Fillet 209.5 200.7 1.04 0.94 67.4AISC w/ CJP 195.7 200.7 0.98 1.04 74.5CJP Welded

Beam/Col Conn 208.8 176.7 1.18 0.81 67.7

Shear Plate

Beam/Col Conn 176.3 176.7 1.00 1.00 84.3


Pcr/Pn

Experimental

Effective Length

Factor, K

Experimental

Slenderness

Ratio (Kl/r)

304

the ratio of minimum compressive capacity prior to failure, , over the critical buckling

capacity.

Figure 7.3.3: TTGP Compressive Degradation Comparison

The ratio of capacity for the TTGP specimen with the bolted shear plate beam-to-column

connection (HSS-22) was significantly less than that observed for the CJP welded connection (HSS-

17), 0.10 compared to 0.26.

Table 7.3.4: TTGP Brace Compression Capacity Degradation Ratio

7.3.2.2 TTGP Brace Displacement and Gusset Plate Rotation Comparisons

The thin tapered gusset plates provide a more flexible end support for the brace and typically result

in brace behavior more closely resembling a pinned-pinned compressive member. This section

evaluates the behaviors of the brace and gusset plates in compression by comparing the brace

CJP Welded Beam/Col

Conn 203.25 53.3 0.26

Shear Plate Beam/Col

Conn 160.57 15.9 0.10

* critical buckling load calculated by subtracting column shears

from applied lateral load for horizontal componet resisted by

brace

Design Parameter Pcr* Pmin Pmin/*Pcr

305

center displacement and gusset plate rotations for the two TTGP specimens. Also included in the

comparison is the AISC reference specimen with CJP interface welds (HSS-12) to illustrate

behavioral differences between rectangular gusset plates design to simulate current AISC

procedures and thin tapered plates designed following the BDP. Brace out-of-plane displacement

as a percent of the total brace length are compared in Figure 7.3.4.

Figure 7.3.4: TTGP Brace Out-of-Plane Displacement Comparison

Both TTGP specimens exhibited similar out-of-plane behavior at a given drift range compared to

the AISC reference specimen regardless of gusset plate detail. The displacement of the brace for

the AISC reference specimen does jump at approximately 2.5% total drift, where as the TTGP

continue to increase steadily until failure. The specimen with bolted shear plate connections (HSS-

22) did exhibited slightly larger displacements than the CJP welded connections (HSS-17) for a

given drift range, which achieved a larger maximum out-of-plane displacement prior to brace

fracture at a greater maximum drift range. The gusset plate rotations at the NE connection are

compared in Figure 7.3.5.

306

Figure 7.3.5: TTGP NE Gusset Rotation Comparison

As expected, the gusset plate rotations for both TTGP were greater than those achieved the AISC

reference specimen at a given drift range. The specimen with bolted shear plate connections

achieved larger rotations at smaller drift ranges than the CJP welded connection but the response of

the specimens with thin tapered gusset plates was similar at drift levels greater than1.5%. It should

be noted that interface welds at both gusset plates of the specimen with CJP welded beam-to-

column connection exhibited severe tearing of the base material prior to failure, which would have

reduced the rotational stiffness of the connection. The bolted shear plate connection only

experience moderate tearing of the interface welds. The residual brace out-of-plane displacement

and NE gusset plate rotations are compared in Figure 7.3.6 and Figure 7.3.7.

307

Figure 7.3.6: TTGP Residual Brace OOP Displacement Comparison

Figure 7.3.7: TTGP Residual NE Gusset Plate Rotation Comparison

Similar to the behavior of the brace with thin rectangular gusset plates, the specimen with bolted

shear plate connections exhibited larger out-of-plane displacement at the brace center and larger

gusset plate rotations. Greater residual out-of-plane displacements and gusset plate rotations

suggest increased the strain demands over the brace diagonal to straighten in tension, causing local

deformation of the brace plastic hinge to occur at smaller drift ranges, shortening the life of the

brace.

308

7.3.2.3 TTGP Brace and Gusset Plate Elongation Comparisons

Thicker gusset plates increase yielding in the brace section by concentrating all the elongation to the

brace section. Thin gusset plates connections designed using the Balanced Design Procedure and

a β factor of 1.0, regardless of geometry, have typically exhibited yielding in both the brace and the

gusset plates. Brace elongations for the two TTGP specimens (HSS-17 and HSS-22) and the AISC

reference specimen with CJP interface welds (HSS-12) are compared in Figure 7.3.8.

Figure 7.3.8: TTGP Brace Elongation vs. Total Drift Range

The two TTGP specimens show similar elongation behavior but CJP welded beam-to-column

connection achieved a greater maximum value by reaching a larger total drift range. The AISC

reference specimen exhibited the greatest brace elongation for a given drift range beyond 1.5% in

this comparison because of the increased axial stiffness of the thicker gusset plates.

Gusset plate elongations for TTGP specimens and AISC reference specimen HSS-12 are compared

in Figure 7.3.9. At smaller total drift ranges, the gusset plates for bolted shear plate beam-to-

column showed larger gusset plate elongation but at approximately 2.5% drift range, gusset plate

yielding was capped. The gusset plates for the CJP welded beam-to-column connections, however,

continued to yield as total drift range increased. Both TTGP specimens experienced greater gusset

plate elongation for a given drift range than the ½” AISC designed gusset plate from HSS-12.

309

Figure 7.3.9: TTGP Gusset Plate Elongation Comparison

The behavior of the brace diagonal in tension was evaluated by comparing the brace elongations

verses the gusset plate elongations. A response with yielding concentrated in the gusset plates

rather than brace is essentially placing greater inelastic demand over a shorter length and not

utilizing the ductility of the brace. This limits the ductility of the system and typically, specimens

with a balance between brace and gusset plate yielding achieved greater drift ranges. The

comparison of tensile elongation between the brace and the gusset plates for the two TTGP

specimens and the AISC reference specimen with CJP interface welds is shown in Figure 7.3.10.

Figure 7.3.10: TTGP Brace Strain vs. Gusset Plate Strain Comparison

310

The response of the specimen with CJP welded beam-to-column connections (HSS-17) was well

balanced following closely along the “One-to-One” line. Elongation over the diagonal was well

distribution and continued to increase in both the brace and gusset until failure occurred.

However, the specimen with bolted shear plate connections (HSS-22) exhibited similar behavior to

HSS-17 before limiting gusset plate yielding and relying completely on brace yielding to reach the

ultimate drift range. HSS-12 exhibited a response concentrating elongation to the brace and

achieved only a moderate drift range.

7.3.3 TTGP Frame Response Comparison

The response of the frame is evaluated for the thin tapered gusset plate specimens with this section.

Comparison of the frame resistance over drift ratio is provided in the positive and negative

directions in Figure 7.3.11 and Figure 7.3.12, respectively. The lateral force resisted by the frame

was calculated using strain gauge data from the columns as describe in Chapter 6, Section 6.4.2.

The frame resistance, or total column shear, is normalized by , the theoretical shear force to

achieve the plastic moment of the column, , at the beam-to-column intersection.

Figure 7.3.11: TTGP Frame Resistance vs. Drift Comparison (+ Direction)

311

Figure 7.3.12: TTGP Frame Resistance vs. Drift Comparison (- Direction)

The AISC reference specimen exhibited the largest frame resistance in both directions which is

expected considering the thickness and geometry of the gusset plate attributes to the in-plane

stiffness of the frame. The bolted shear plate connection of specimen HSS-22 exhibited lower

resistance at a given positive drift level but achieve larger resistance than the CJP welded

connection in the negative direction. This is somewhat surprising considering the bolted shear

plate connection is inherently less stiff.

The distribution of resistance was compared between the lateral loads resisted by the brace and by

the frame for the two TTGP specimens in Table 7.3.5. The percentages of total shear resisted by

the brace and frame are summarized at brace yielding and at the ultimate drift levels in Table 7.3.6.

Table 7.3.5: TTGP Distribution of Resistance

HS

S-2

2

312

Table 7.3.6: Summary of TTGP Resistance Distribution

When looking at the distribution of resistance for TTGP specimen with the bolted shear plate

beam-to-column connection (HSS-22) in Table 7.3.5, it can be seen that once the brace buckled in

compression, the contribution of brace resistance decreased sharply as the negative drift level

increased. The specimen with CJP welded connections shows a more balanced response with 29%

of the total story shear resisted by the brace and 71% by the frame, compared to HSS-22 which

retained only 9.4% of the total resistance through brace compression prior to fracture. Again, this

suggests that the frame connection had a significant influence on not only the critical buckling force

of the brace, but also the nonlinear response in compression after buckling.

7.3.4 TTGP Energy Dissipation

The total energy dissipated by the system, brace diagonal and by the frame are presented for the

two TTGP specimens in Table 7.3.7. Energy dissipated by the components and the percent

dissipated by the brace diagonal and by the frame are summarized in Table 7.3.8. The specimen

with bolted shear plate beam-to-column connections (HSS-22) dissipated 4.5% less total energy by

reducing the in-plane stiffness of the frame compare to the CJP welded connection for HSS-17.

This is a relatively small reduction considering HSS-17 achieved significantly larger drift levels and

exhibited greater resistance in both directions. The specimen with CJP welded beam-to-column

connections dissipated 85% of its energy through the response over the brace and gusset plates

while the specimen with bolted shear plate connections saw a larger contribution of energy

dissipation from the frame, 27%. As seen above in Figure 7.3.12, the frame resistance in the

positive direction was reduced with the more flexible connection but the total frame response was

capable of dissipation 73.1% more energy than that of the frame having CJP welded connections.


AISC w/ Fillet 78.0 22.1 80.3 19.7 36.0 64.0 73.2 26.8CJP Welded

Beam/Col

Conn 90.6 9.5 84.2 15.8 29.0 71.0 67.2 32.8Shear Plate

Beam/Col

Conn 86.5 13.5 87.0 13.0 9.4 90.6 79.2 20.8

% Resistance at Ultimate

-Drift% +Drift% -Drift% +Drift%

% Resistance at Brace Yield/Buckling

Specimen

313

Table 7.3.7: TTGP Energy Dissipation Comparison

HS

S-1

7

HS

S-2

2

Table 7.3.8: Summary of Energy Dissipation for TTGP

During the testing of the TTGP specimen with CJP welded beam-to-column connections (HSS-

17), the LVDT calibration factor used in controlling the induce displacement was incorrect. This

was not corrected until after the completion of the test when it was realized that the displacements

induced were 1.45 times larger than intended through the loading protocol (Kotulka 2007). As a

result, specimen HSS-17 only completed 31 cycles, whereas HSS-22 achieved 34 complete cycles

prior to brace fracture. Energy dissipation is related to the number of cycles completed. It is

suspected that the more rapidly escalating amplitude from the unintentionally modified loading

protocol for HSS-17 could result in less energy dissipated while still achieving a greater ultimate

drift capacity.

7.3.5 TTGP Summary

The reduction in ultimate drift range and total resistance for TTGP specimen was significant by

utilizing bolted shear plate beam-to-column connection rather than CJP welded moment

connection adjacent to the gusset plates. Reducing the in-plane stiffness of the frame appears to

limit the buckling capacity of the brace and increases the degradation of brace resistance post

buckling. The increase in residual brace displacement and gusset plate rotations seems to attribute

kip-in % kip-in %

AISC w/ Fillet 3334 3181 95 153 5 2.65

AISC w/ CJP 3219 N/A N/A N/A N/A 3.49

CJP Welded

Beam/Col Conn 4807 4086 85 721 154.94

Shear Plate

Beam/Col Conn 4589 3341 73 1248 273.98

Specimen

Total Energy

Dissipated,

kip-in



Range, %

Frame Diagonal

Energy Dissipation

314

to early plastic deformation at the brace center and early fracture at smaller drift ranges. For both

specimen, large out-of-plane rotations were exhibited by the gusset plates in compression and

severe yielding was observed in the gusset plates in tension. Severe cracking was observed in the

based material along the interface welds for HSS-17, while moderate cracking was observed in

HSS-22. In both cases, the increase deformation demands and shorter interface weld length

increase the potential for weld fracture.

Both specimens achieved similar total energy dissipation but HSS-17 with CJP welded connections

dissipated most of the energy through the brace while HSS-22 with bolted shear plate connections

relied heavily on the inelastic response of the less stiff frame to dissipate energy. The component

behaviors in compression and tension were similar with both connections.

7.4 Wide-Flange Verses HSS Tubular Brace Sections

Prior to specimen WF-23, all of the specimens had been designed and tested with HSS tubular

brace sections. The behavior of A500 B/C HSS sections under large cyclic loading has shown that

once local deformation associated with the formation of the hinge is observed, tensile strain is

concentrated within the hinge region. This strain accumulation leads to the initiation of cracking

and eventual fracture of the brace section and moderate drift levels. It is thought that A992 wide-

flange brace sections could increase the ultimate drift capacity of the system by exhibiting a more

ductile response from the brace.

This section discusses and compares the performance of WF-23 which utilized a W6x25 brace

section with the performance of HSS-05, “best” performing HSS tubular specimen, and the AISC

reference specimens, HSS-01 and HSS-12. The global system responses are compared in

Section7.4.1 followed by discussion and comparison of the brace and gusset plate behaviors in

Section Error! Reference source not found..

7.4.1 Wide-Flange vs. HSS System Response Comparison

Both HSS-05 and WF-23 utilized 3/8” gusset plate connections designed following the Balance

Design Procedure with similar elliptical clearance method using an 8t offset. The frames consisted

of the same beam and column sizes and CJP welded beam-to-column connections were used

adjacent to the gusset plates. The ultimate drift capacity and maximum resistance are summarized

in Table 7.3.8. Backbone curves using the peak resistance and drifts for the system response are

compared in Figure 7.4.1. The lateral resistances of the specimens were normalized by the lateral

load associated with brace yielding from Equation (1.2.1).

315

Table 7.4.1: Peak Performance Summary for WF Comparison

Figure 7.4.1: WF Backbone Curve Comparison

The W6x25 brace section was selected because the maximum expected brace force in tension,

, was similar to that of the typical HSS5x5x3/8 tubular section, 403.7 kip compared to

398.0 kip. The calculated nominal buckling capacity of the wide-flange section was slightly lower

than that of the HSS brace based on AISC Specification Equation E3-2, 168.0 kips verses 176.2

kip. However, the system response for WF-23 showed that buckling of the brace in compression

and tensile yielding of the brace occurred and considerably lower lateral loads than previous

specimens with HSS tubes including the AISC reference specimens. Also a response characteristic

observed that was unique to the wide-flange brace section was the reduction of system resistance

immediately following brace buckling. The specimen with HSS tubular brace section typically

leveled but maintain system resistance after buckling of the brace occurred.

Range Min Max

HSS-01 2.65 -1.64 1.01 -0.95 1.64

Interface

Weld

Fracture


with Fillet Welds

HSS-12 3.49 -2.10 1.39 -0.90 1.86Brace

Fracture


with CJP Interface

Welds

HSS-05 4.96 -3.09 1.87 -0.80 1.76Brace

Fracture

"Best" HSS Tubular

Brace Section

WF-23 5.56 -3.21 2.35 -0.58 1.30

Interface

Weld

Fracture

WF Brace Section

SpecimenDrift Ratio, % Resistance

P/Py

Failure


316

WF-23 did achieve a 12.1% larger total drift range than HSS-05 and 59.3% larger than AISC

reference specimen with CJP interface welds (HSS-12). The wide-flange specimen ultimately failed

due to interface weld fracture at the NE gusset plate connection where as HSS-05 was able to

achieve the desired failure mechanisms of brace fracture at the plastic hinge.

7.4.2 Wide-Flange vs. HSS Brace and Gusset Plate Response Comparison

The response of the brace and gusset plate are compared in this section for the wide-flange brace

specimen (WF-23) and the “best” performing HSS brace specimen (HSS-05). The brace buckling

capacity and nonlinear response, forces verses elongation over the brace diagonal work-point to

work-point, are compared in Section 7.4.2.1. The behavior of the brace and gusset plates in

compression are compared in Sections 7.4.2.2 and in tension in Section 7.4.2.3. Evaluation of the

energy dissipation by the wide-flange brace specimen is provided in Section 7.4.3.

7.4.2.1 Wide-Flange vs. HSS Brace Force Comparison

The experimental buckling capacities of the specimens are compared in Table 7.4.2. The critical

buckling load, , was calculated using strain gauge data described in Chapter 6, Section 6.3.1.1 for

the wide-flange brace (WF-23), and compared to the nominal buckling capacity of the brace per

AISC Design Specifications. The experimental effective length factor, , and slenderness ratio,

, were calculated and included in the table.

Table 7.4.2: Wide-Flange vs. HSS Experimental Buckling Capacity Comparison

The W6x25 brace section buckled at a compressive load slightly above the design value, 167.0 kip

compared to 165.4 kip. This resulted in an effective length factor of 0.99 and slenderness ratio of

103.8. The full nonlinear responses of the braces, brace force verses diagonal elongation, are

compared in Table 7.4.3. Brace force was normalized by the nominal tensile yield force calculated

in Equation (7.2.2). The wide-flange brace force was calculated from the brace strain gauge data as

describe in Chapter 6, Section 6.3.1.2. This method determined the brace stresses from the

Experimental AISC

Pcr Pn=FcrAg

AISC w/ Fillet 209.5 200.7 1.04 0.94 67.4

AISC w/ CJP 195.7 200.7 0.98 1.04 74.5HSS Brace

Section (HSS-05)174.8 176.2 0.99 1.01 85.0

Wide-Flange

Brace Section 167.0 165.41.01

0.99 103.8


Pcr/Pn

Experimental

Effective Length

Factor, K

Experimental

Slenderness

Ratio (Kl/r)

317

nonlinear strain response recorded in the brace. The brace force used for the HSS-05 response was

calculated by subtracting total column shears from the applied lateral load to obtain the horizontal

components of the brace force. The method for determining column shears is described in

Chapter 6, Section 6.4.2.

Table 7.4.3: Wide-Flange vs. HSS Brace Response

HS

S B

race S

ect

ion

(HSS

-05)

Wid

e-F

lan

ge B

race

Secti

on

Although the W6x25 wide-flange brace section showed greater ductility than the HSS5x5x3/8

brace from HSS-05, both the compressive and tensile resistances of the wide-flange section were

less than that of the HSS tube. The brace yielded at a tension force lower than the calculated

nominal yield strength, . The wide-flange brace also exhibited severe compression capacity

degradation post buckling and retained little capacity at maximum negative drift levels.

7.4.2.2 Wide-Flange vs. HSS Brace Displacement and Gusset Plate Rotation Comparison

The brace displacement and gusset plate rotations are evaluated for the wide-flange brace specimen

by comparing the behavior to those of HSS-05 and the AISC reference specimen with CJP

interface welds (HSS-12). The observed behavior of the W6x25 wide-flange brace sections was

inherently different in compression than the typical HSS5x5x3/8 tube. As typically seen in

previous test with HSS tubular section with flexible gusset plate connection, initial buckling

occurred and the brace shape resembled a half sinusoidal wave. As horizontal displacement

increased, the brace began to buckle in a more triangular shape with the majority of inelastic

deformation near the brace midpoint. Once the formation of the plastic hinge occurred at the

brace midpoint, the concentration of local deformation was over a relatively short length of brace,

approximately twice the depth of the member.

The wide-flange brace section exhibited similar half sine wave shape after initial buckling, but

maintained a more gradually curving shape as horizontal displacement increased. The

concentration of yielding as the plastic hinge formed was over a much great length of brace,

approximately three times the depth of the section as shown in Figure 7.4.2. Also, WF-23 failure

318

due to interface weld fracture before developing the local deformation associated with the brace

plastic hinge.

Figure 7.4.2: WF Brace Yielding at Plastic Hinge (-3.21%)

Out-of-plane displacement response of the WF brace midpoint is compared in Figure 7.4.3 with

the brace displacement of HSS-05 and AISC reference specimen HSS-12. From the onset of

buckling, the wide-flange brace achieved larger out-of-plane displacements than the specimens with

HSS brace sections. This is partly because the larger area of the W6x25 and resulting increase in

axial stiffness, . The force in the brace after buckling was smaller or comparable to

the HSS brace at similar negative drift levels, and there for, resulted in less axial shortening of the

brace. This requires a larger out-of-plane displacement to accommodate the change in the diagonal

length associated with a given horizontal drift level. The displacement of the wide-flange brace

increased less at a given drift range than either HSS-05 with the HSS tubular brace or AISC

reference specimen HSS-12 beyond approximately 1.5%. However WF-23 achieved a maximum

displacement prior to system failure by reaching a greater total drift range.

≈ 3d

319

Figure 7.4.3: WF Brace OOP Displacement Comparison

Specimen WF-23 exhibited large gusset plate rotations and severe damage to the interface welds,

ultimately leading to fracture. Figure 7.4.4 compares the NE gusset plate rotations of WF-23 to

those of HSS-05 and AISC reference specimen HSS-12. It can be seen that significantly larger

rotations occurred to accommodate the out of plane displacement of the brace section of WF-23.

In this case, the rotations lead to fracture of the interface weld prior to the development of local

deformation at the brace plastic hinge, which could be an issue of concern when designing SCBFs

with thin flexible gusset plates and wide-flange sections.

Figure 7.4.4: NE Gusset Plate Rotation Comparison for WF

320

The residual out-of-plane displacement at the brace midpoint and residual gusset plate rotations

from the NE connection are compared with the residual displacement and rotations of HSS-05 and

AISC reference specimen (HSS-12) in Figure 7.4.5 and Figure 7.4.6, respectively. The residual

behavior for the displacement at the brace midpoint was similar to the maximum displacement.

The W6x25 brace section exhibited larger residual displacements than HSS-05 and the AISC

reference specimen for a given drift range below 4.30%, when the residual displacement of the HSS

brace section from HSS-05 equaled that of the wide-flange brace.

Figure 7.4.5: WF Residual Brace OOP Displacement Comparison

The NE gusset plate in WF-23 experienced larger residual rotation at a given drift range compared

to the gusset plates from HSS-05 or the AISC reference specimen (HSS-12). The larger out-of-

plane brace displacement and gusset plate rotations increased inelastic demand on the gusset plates.

This also puts larger stress demands on the interface welds and increases the potential for weld

tearing.

321

Figure 7.4.6: WF Residual NE Gusset Plate Rotation Comparison

7.4.2.3 Wide-Flange vs. HSS Brace and Gusset Plate Elongation Comparison

The behavior of the W6x25 brace section and gusset plates in tension are compared in this section

by comparing brace and gusset plate elongation to the HSS brace of HSS-05 and the HSS brace of

AISC reference specimen HSS-12. The brace elongation verses total drift range is compared in

Figure 7.4.7. Elongation is given as a percent of the original brace length from edge of gusset plate

to edge of gusset plate.

Figure 7.4.7: Brace Elongation Comparison for WF

322

The WF-23 wide-flange brace section exhibited significantly greater tensile elongation than HSS-05.

The W6x25 achieved more elongation at a given drift range and continue to elongation linearly as

drift range increased. Comparing the elongation between HSS-05 and WF-23 at 4.30% total drift

(the maximum recorded elongation level for HSS-05) the wide-flange brace achieved 28.9% more

elongation than the HSS tubular brace.

The majority of tensile elongation for WF-23 over the brace diagonal, work-point to work-point,

occurred in the brace. The WF specimen gusset plates exhibited minor yielding and elongation in

tension, which could be expected after observing the lower maximum brace force in tension, 351.5

kip, as calculated in Section 6.3.1 of the previous chapter. The maximum brace from HSS-05 was

taken from the previous thesis as 400.9 kip (Johnson 2005). Gusset plate elongation is compared in

Figure 7.4.8.

Figure 7.4.8: Gusset Plate Elongation for WF

The distribution of tensile yielding between the brace and the gusset plates is compared in Figure

7.4.9 by plotting the brace strain verses the gusset plate strain. The wide-flange brace shows a

response dominated by tensile elongation. It could possible to further increase the ultimate drift

capacity of the system by increasing gusset plate elongation.

323

Figure 7.4.9: WF Brace Strain vs. Gusset Plate Strain Comparison

7.4.3 Wide-Flange vs. HSS Energy Dissipation Comparison

The energy dissipation of WF-23 with the wide-flange brace section, HSS-05 with the HSS brace

section and thin rectangular gusset plate, and the two AISC reference specimens (HSS-01 and HSS-

12) were compared in Table 7.4.4 including total energy dissipated by the brace diagonal and by the

framing elements. WF-23 dissipated 19.6% more total energy than HSS-05 and 149% more energy

than AISC reference specimen HSS-01. The energy dissipations of the total system, and

individually by the brace diagonal and frame, are shown in Figure 7.4.10 for WF-23. At the

completion of the test, the wide-flange specimen saw 63% of total energy dissipated by the brace

and 36% by the frame, which varies greatly compared to that of the HSS-05.

Table 7.4.4: WF Energy Dissipation Comparison

kip-in % kip-in %

AISC w/ Fillet 3334 3181 95 153 5 2.65

AISC w/ CJP 3219 N/A N/A N/A N/A 3.49HSS Brace

(HSS-05)6941 5451 79 1490 21 4.96

WF Brace

(WF-23) 8300 5259 63 2963 36 5.56

Specimen

Total Energy

Dissipated,

kip-in



Range, %

Frame Diagonal

Energy Dissipation

324

Figure 7.4.10: WF-23 Energy Dissipation

It can be seen that the rate of dissipation for the frame increased considerably at severe drift ranges,

greater than approximately 4.5%, but the energy dissipated through the brace diagonal continued to

increase linearly with drift range. It is likely that any specimen regardless of brace section type

would see significantly larger contribution of energy dissipated through frame action by achieving

drift ranges greater than 5.0% within this test setup.

7.4.4 Wide-Flange Brace Summary

Specimen WF-23 achieved a larger total drift range than any specimen with HSS brace sections and

similar gusset plate designs. The ductile behavior of the brace resulted in delayed formation of

local deformation at the brace plastic hinge in compression and extreme extremely large tensile

elongation. However, the large rotational demand on the gusset plates resulted in interface weld

damage and an undesirable failure mode to control for the system.

Another consideration when designing SCBFs with wide-flange brace sections is the increased cost

due to additional material, fabrication and construction time. Steel is typically purchased by the ton

which means that in our case, the W6x25 wide-flange weighs 12.1% more than the HSS5x5x3/8

tube and resulted in lower resistance. The brace to gusset plate connection are typically more

complicated and require fabrication time to prepare the ends, as well as more welding to connect

both flange and the web to the gusset plate. For the connection implemented in WF-23 shown in

Figure 7.4.11, an additional web plate was also required to connect the web. Tubular section

connected to gusset plates only required two slots and 4 fillet welds to connect.

325

Figure 7.4.11: WF Brace to Gusset Plate Connection

7.5 Bolted Connections for SCBFs

The designs of HSS-19, HSS-20, HSS-21 and HSS-24 had considerable input from representative

with AISC to evaluate SCBF connections that utilize bolted connections. Bolted connections in

lieu of welding are more economical and provide ease of installation during erection. This section

discusses each of these specimens and evaluates the benefit and detriment of each in terms of

seismic performance and constructability.

7.5.1 HSS-19: Bolted WT Brace to Gusset Plate Connection

Specimen HSS-19 was the only test within this test series that utilized a bolted brace to gusset plate

connection. The brace did not behave as intended by design and buckling of the extension plate

occurred and ultimately fractured before fully developing buckling over the brace length. This was

due to the length between the ends of HSS tubular brace and the start of the WTs being too large,

creating a reduction in stiffness and instability in compression. However, the idea of a bolted brace

to gusset plate does show some promise with regards to constructability and should not be

abandoned without further research. Specimen HSS-19 showed the importance of assuring

All Sides

Typ

326

consistent buckling resistance over the entire brace length if intending to achieve buckling and

eventual hinge formation at the two gusset plates and the brace midpoint.

Benefits

Ease of installation during construction

Draw-backs

Undesirable failure mechanism controlled

Reached only moderate drift levels prior to fracture

Minimal energy dissipated

Additional material needed for connection: HSS tube, ¾” extension plate and 2-WT4x17.5

The bolted connection between the brace and gusset plates used 1-1/8” A490 slip-critical bolts.

Under the AISC Seismic Provision, bolted connections used to resist seismic loading are required

to be slip-critical. This is intended to avoid racking of the bolted connections during moderate

level earthquakes. However, during a severe event, the initial frictional force within the slip-critical

bolts would mostly likely be exceeded and bolt slip would occur, as seen in HSS-19. The shear

force in the connection was then transferred through bolt bearing. This behavior of bolt slip

through the brace should be investigated further to determine what affect it has on brace resistance

and ductility.

7.5.2 HSS-20 and HSS-21: Bolted Beam End-plate Connection

This section discusses the performance of the two specimens with bolted beam end-plate to

column connections adjacent to the gusset plate connections. HSS-20 was tested first and utilized

an 18 bolt configuration while HSS-21 used a reduced configuration of 14 bolts. The concept of a

bolted beam-to-column connection for SCBFs that eliminates the requirement of field welding the

gusset plate to either framing element is very attractive to designers and erectors. This connection

also provides moment resistance similar to a CJP welded beam-to-column connection.

For both specimens, the gusset plates exhibited significant yielding in tension but lower gusset plate

rotations without the occurrence of weld damage. At the end of HSS-21, severe yielding covered

the entire area of the gusset plate as shown in Figure 7.5.1. The brace behavior and buckled shape

showed larger curvature toward the brace midpoint at lower drifts, which inherently lead to earlier

local deformation at the plastic hinge and short brace life. The thickness of the beam end-plate

being 1”, essentially increased the calculated in-plane frame stiffness as shown in Section 6.4.2 and

327

the behavior of the gusset plate rotations suggest larger rotational stiffness out-of-plane. Previous

research within this test program have shown that having increasing the in-plane stiffness of the

framing elements and rotational out-of-plane stiffness of the gusset plates results in smaller ultimate

drift capacity and short brace life (Herman 2007).

Figure 7.5.1: HSS-21 NE Gusset Plate at End of Test

HSS-20 (18 Bolt Configuration) Benefits

Improved ultimate drift capacity by 13.8% compared to AISC reference specimen HSS-12

50.5% increase in energy dissipation over AISC reference specimen HSS-01

Similar total resistance of system compared to HSS-05

Achieved significant yielding of both brace and gusset plates in tension

Bolting beam and gusset plate to column reduces installation time

More shop welding required rather than field welding improves weld quality

HSS-20 (18 Bolt Configuration) Draw-backs

19.8% less drift capacity than achieved by HSS-05

27.7% less energy dissipated than by HSS-05

Early onset of local deformation at plastic hinge

Early brace fracture

HSS-21 (14 Bolt Configuration) Benefits

328

Improved ultimate drift capacity by 18.6% over AISC reference specimen HSS-12

44.7% increase in energy dissipation compared to AISC reference specimen HSS-01

Similar system resistance to HSS-05

Significant yielding extending beyond brace into gusset plates

14 bolt connection further reduces installation time during erection

Higher quality assurance by welding interface welds in shop rather than in the field

HSS-21 (14 Bolt Configuration) Draw-backs

16.5% less drift capacity than achieved by HSS-05

30.5% less drift energy dissipated than by HSS-05

Increased potential for tensile bolt fracture as seen at NE connection

The performances of HSS-20 and HSS-21 showed that improved ductility was possible with the

bolted beam end-plate connection and gusset plates designed following the Balanced Design

Procedure and the 8t elliptical clearance. The increased ease of installation and shop and

elimination of field welding make this type of connection attractive with regards to constructability.

Nevertheless, the specimen did not perform as well as HSS-05 and the 1” thick endplate appeared

the modify the brace behavior, resulting in formation of the plastic hinge at an earlier drift range

and fracture at lower drift than what was achieved with a CJP welded beam-to-column connection.

7.5.3 HSS-24: Welded Flanges, Bolted Web Beam-to-Column Connection

Specimen HSS-24 implemented a modified beam-to-column connection that reduced the total

amount of CJP welds required by connecting the beam web to the column with a bolted shear plate

but CJP welding the flanges. It is recognized in the AISC Seismic Provisions that the combinations

of welds and bolts should not be used to resist seismic loads because of the differential stiffness

behavior between bolts and welds. Welds is a more rigid connection method while bolts can slip or

move within the bolt holes. This could cause force to be concentrated on the more rigid connector

rather than a distribution as intended by design.

The ultimate drift capacity of HSS-24, which achieved a 4.44% total drift range, was the largest of

any specimen tested within this series with HSS5x5x3/8 brace sections. The specimen exhibited

large out-of-plane displacements of the brace midpoint and large gusset plate rotations with the

brace in compression. Only minor tearing of the interface welds was observed at large negative

drift levels. In tension, the brace and the gusset plates displayed an even distribution of yielding to

increase elongation over the entire diagonal length. The pattern of yielding for the gusset plates

329

suggests an innately different stress distribution compared to those observed in previous tests. The

typical gusset plate design by the Balance Design Procedure and with elliptical clearances showed

the arching elliptical shape of yielding as the plate rotated out-of-plane and a fairly even distribution

of tensile yielding along the brace to gusset welds and at the brace end. A photo of the SW gusset

plate for HSS-24 in Figure 7.5.2 shows that concentration of yielding occurred between the brace

and the gusset to column interface weld. This could be because of the reduce shear stiffness of the

web at the beam-to-column connection caused more of the vertical component of the brace force

to be transferred through the gusset plate to column interface weld.

Figure 7.5.2: HSS-24 SW Gusset Plate Yielding at End of Test

Overall, the specimen performed well, but still was unable to achieve the ultimate drift capacity

shown possible through HSS-05. It appears that even with a minor adjustment to the in-plane

stiffness of the frame, some loss in ductility is evident. Also, the observed level of damage to the

column webs, especially near the frame connection was significantly more than what had been

observed in previous test. Figure 7.5.3 shows the level of yielding observed in the column webs at

the end of the test.

330

Benefits

Achieved 27.2% greater total drift range than AISC reference specimen HSS-12 and only

10.4% less than HSS-05

Dissipated 57.1% more total energy than AISC reference specimen HSS-01

Desired brace behavior in compression and tension with yielding occurring over brace and

gusset plates

Less field welding required by bolting web to column

Draw-backs

Combination of bolted and welded connections not currently permitted under AISC

Dissipated 24.5% less energy than HSS-05

Increased demand to columns

Specimen HSS-24 showed the most similarity in performance to HSS-05 but also utilized a

modified connection that only provides minimal gains in economy and constructability. Also,

increasing demands and inelastic deformation to the columns is less desirable because of increased

risk to the gravity load carrying capacity of the system.

Figure 7.5.3: HSS-24 Column Damage at Beam-to-Column Intersection (End of Test)

331

7.6 Net Section Reinforcing Requirement

This section discusses the requirement of net section reinforcing by evaluating the behaviors of

HSS-11, HSS-25 and HSS-26. Nine of the 24 specimens tested with HSS within this research

program prior to HSS-25 did not include net section reinforcing. The necessity of additional plates

at the brace to gusset connection based on AISC Specifications (Equation D2-2) and Table D3.1 to

account for shear lag had come into question considering no damage, yielding or tearing, was

observed at the brace net section (Kotulka 2007). Nonlinear finite element analysis had shown that

the resulting stress levels at the brace net section increased with frame and gusset plate stiffness

(Yoo 2006). The increased in-plane stiffness of the heavier beam and larger gusset plate increased

the in-plane moment of the brace in tension, which results in the reduced brace area experiencing

the highest stress and strain demands. Therefore, HSS-25 was proposed to evaluate the potential

for net section fracture with a heavier W16x89 beam section, 7/8” thick gusset plates and no net

section reinforcement.

Specimen HSS-25 was tested using the same cyclic loading protocol with increasing amplitude as

used in HSS-11. Initial tearing of the brace material at the northeast brace end occurred at initial

drift range of 0.59%. At the end of the test, the cracking had propagated to approximately 5/8”

above and 9/16” below the slot in the gusset plate but did not result in fracture. The damage to

the net section for HSS-25 is shown in Figure 7.5.3.

Figure 7.6.1: HSS-25 Net Section Tearing (0.88% Drift Level)

332

HSS-25 ultimately failed due to brace fracture at the plastic hinge but the results from the test

demonstrated the potential for net section fracture as the controlling failure mode. A study

conducted at UC Berkeley and included in the Steel Tips publication (April 2005) Limiting Net

Section Fracture in Slotted Tube Braces experimentally evaluated the affect of different drift histories and

effective net section area on the potential for net section fracture. The results showed that the

requirement for net section reinforcing is more critical for tension near-fault drift histories (Yang

and Mahin, 2005). HSS-26 tested an identical test specimen, also without net section

reinforcement with a tension dominated near-fault loading protocol and resulted in brace net

fracture at a drift level of 0.99%.

The performance HSS-25 did not appear limited when subjected to the typical loading protocol and

the brace was able to develop its full capacity in tension and yield over the length. However,

damage was observed at the net section of HSS-25 and net section fracture controlled in HSS-26 by

applying the alternate loading protocol. The design of SCBFs must address the demands from any

expected seismic ground excitation in order to assure the desired response and meet the

expectations of Life Safety and Collapse Prevention performance levels. These results show that it

would be negligible to eliminated net section reinforcing for SCBFs with HSS tubular braces and

that further research is required to determine if lessening the requirement is feasible.

333

Chapter 8: Conclusions

8.1 Introduction

A summary of the nine tests specimen within this test series is provided in Section 8.2.

Conclusions based on the results from the analysis and performance comparisons in Chapter 6 and

Chapter 7 are presented in Section 0. Finally, recommendations to future work on SCBFs are

discussed in Section .

8.2 Research Summary

Specimens HSS-18 through HSS-26 were selected to evaluate both new directions in connection

methods using the Balance Design Procedure and to further investigation potential failure modes

currently address by AISC strength design specification and φ factors. All of the test specimens

were successfully tested within the designed test setup. The test observations and recorded data

were effective in capturing global and local behavior for comparison between the nine tests and

with results from previous test series.

HSS-18: Specimen HSS-18 reduced the required welding at the beam-to-column connection by

utilizing a web shear plate rather than the CJP welded moment connection. Gusset plate design

was identical to what was consider the “best” performing specimen, HSS-05, with rectangular

gusset plates design with the BDP and 8t elliptical clearance. The 3/8” thin gusset plate was

designed for the full plastic capacity of the plate and connected to the beam and column with 3/8”

fillet welds each side. The specimen performed well in that it achieved the desired behavior of

yielding in the brace, gusset plates and framing elements and eventually failed due to fracture of the

HSS5x5x3/8 brace section at the plastic hinge.

The total drift range achieved was 4.19% and resistance was slightly less but comparable to

specimens with similar gusset plate design. The effect of the unwelded shear plate beam-to-column

connection on performance was the reduction in ultimate total drift capacity and loss in resistance

with the brace in both tension and compression. Damage to the gusset plates was also noticeably

more severe by comparison with similar tests and included severe edge deformations from in-plane

rotation at the beam/column joint. Also, the more flexible beam-to-column connection placed less

rotational demand on the framing elements and more on the gusset plates, resulting in only minor

damage to the beams and columns at larger drift levels.

334

HSS-19: The brace configuration implemented for Specimen HSS-19 was intended create a more

constructible brace to gusset connections for SCBFs requiring less installation time than welded

connections. The brace was bolted to the gusset plates using 2-WT section which were spliced to

the HSS5x5x3/8 tubular brace with 3/4” extension plates. The large number of connection points,

elements to element, along the brace created instability over the full brace length and hinging

occurred in the southwest extension plate at initial drift levels and lower resistance than anticipated.

Inelastic buckling damage was concentration to the extension place with very little yielding

occurring beyond the brace. The extension plate fractured at a total drift range of 1.31%.

HSS-20: Specimen HSS-20 was the first of two specimens designed using bolted beam end plate

connections to the columns adjacent to the gusset plates. This specimen utilized an 18 bolt

configuration at the end plate and 3/8” gusset plates designed using the BDP and 8t elliptical

clearance. The specimen exhibited a total drift range of 3.97% and resistances similar to specimen

with similar gusset plate design and the CJP welded beam-to-column moment connection. The

response of the specimen exhibited the desired brace behavior of inelastic buckling and tensile

yielding, and achieved the desired failure mechanism of brace fracture. Inelastic damage was

observed beyond the brace and into the gusset plates and framing elements. Notably, initial

yielding of the gusset plates occurred at very small initial drift levels and the severity of tensile

yielding observed over the gusset plates was more than what was typically seen for similar designs.

HSS-21: The design of specimen HSS-21 was identical to HSS-20 with the exception of a reduced

number of bolts at the beam end plate connection. A more economical 14 bolt configuration was

implemented increasing the in-plane flexibility of the beam-to-column connection. The specimen

achieved a total drift range of 4.14% and showed similar resistance. The desired inelastic brace

behavior was achieved and fracture occurred at the brace plastic hinge. Tensile yielding and out-of-

plane deformation of the gusset plates was observed as well as yielding and deformation of the

framing elements. Again, with the bolted end plate connection, initial gusset plate yielding was

observed at early drift levels and severe yielding over the entire gusset plate areas was seen at the

end of the test. The bolted end plate connections decreased the gusset plate rotational capacity,

essentially decreasing the life of the brace and resulted in a lower ultimate drift capacity than what

was achieved with the CJP welded beam-to-column moment connection.

HSS-22: Similar to HSS-18, Specimen HSS-22 investigated the effects of eliminating the CJP

welded beam-to-column moment connection in lieu of the bolted web shear plate. The gusset

plates of this specimen utilized an identical tapered geometry and BDP design as HSS-17, which

335

achieved very large drift ranges, 4.94% total, and is considered the “best” performing specimen

with tapered gusset plates design following the BDP.

Specimen HSS-22 exhibited the desired inelastic brace behavior and ultimate controlling failure

mode, but reached only 3.96% total drift range and saw a significant reduction in compressive

resistance. Severe damage to the gusset plate including moderate damage to the interface welds at

the completion of the tests. The beams and columns exhibited minimal inelastic behavior and

lower resistance than seen in any other specimen. The in-plane flexibility of the beam-to-column

connection and tapered geometry resulted in very large rotations in the gusset plates. More flexible

gusset plates also increased the effective slenderness of the brace, degrading resistance post-

buckling more rapidly. The formation of local deformations at the plastic hinge and brace fracture

occurred at a smaller total drift range than what was observed for HSS-17.

HSS-23: Specimen HSS-23 investigated the performance of BDP gusset plates with an A992

W6x25 wide-flange section. The specimen reached very large total drift ranges, 5.56%, but failed

due to interface weld fracture prior to fully developing the plastic hinge at the brace midpoint. The

response of the brace section showed considerably more ductility than what was typically seen of

A500 B/C HSS sections. Inelastic damage was also observed throughout the gusset plates and into

the framing elements at the completion of test. The behavior of the brace resulted in larger out-of-

plane displacements at the midpoint and therefore larger rotations at the gusset plates. These

rotations induced severe damage to the interface welds which eventually propagated to fracture.

HSS-24: Specimen HSS-24 was chosen to further evaluate the potential for reducing the required

welding at the beam-to-column connection while maintaining improved seismic performance. The

connection consisted of CJP welded beam flanges to the column but a bolted shear plate

connection for the web. Again, the 3/8” rectangular gusset plate with 8t elliptical clearance

following the BDP was implemented. The specimen reached a large total drift range, 4.44%, and

achieved the desired inelastic response of the brace and of the gusset plates. The connection

resulted in more damage to the columns that what was seen in previous tests. Connecting the

beam web with bolts rather than CJP welds showed minor difference in system stiffness and

resistance, but did reduce the ultimate drift capacity.

HSS-25: The influence of the heavier beam sections and thick gusset plates on the brace behavior

and the potential for net section fracture were evaluated with this specimen. The specimen design

was identical to HSS-11 with heavier W16x89 beam sections and 7/8” gusset plates but eliminated

net section reinforcement of the brace at the brace to gusset connection. The specimen achieved a

total drift range of 3.30% and exhibited greater resistance by increasing the brace buckling capacity

336

and frame resistance. Minimal tensile yielding occurred in the gusset plates but significant

deformation of the gussets was observed from out-of-plane rotation as the brace buckled. Inelastic

damage to the columns was significant at larger drift ranges and the specimen ultimately failed due

to fracture of the brace at the plastic hinge. However, tearing of the brace net section was observed

and, as expected, the life of the brace was diminished because of the increased demands from the

stiffer connections. This test showed that net section fracture was a possibility and set the ground

for further investigation into the failure mode.

HSS26: Specimen HSS-26 was identical to HSS-25 by design, but was subjected to an alternate

near-fault loading protocol. Research had shown that potential for net section fracture of the brace

in SCBFs was more severe when a large tension story drift was induced prior to the cyclic loading

of the system because inelastic buckling of the brace in compression concentrates strain at the

brace midpoint rather than over the full length. The specimen fractured the brace at the net section

at a positive drift ratio of 0.99%.

8.3 Conclusions

8.3.1 Balance Design Procedure and Elliptical Clearance

Gusset plate connections designed following the BDP developed the full tensile capacity of

the brace and induced yielding over both the brace and the gusset plates as intended

The severity of the gusset plate yielding for specimen with BDP tapered designs limited

brace yielding resulting in less total ductility

The 8t elliptical clearance effectively allowed inelastic deformation for gusset plates to

rotate out-of-plane for specimen with varying plate thicknesses and the HSS5x5x3/8 brace

section

The 8t elliptical clearance used with the W6x25 wide-flange brace section resulted in severe

interface weld damage and fracture as the controlling failure mode

8.3.2 Beam-to-Column Connection

The specimen with CJP welded moment connection achieved a larger ultimate drift

capacity than any of the alternate connection methods

Connections with greater in-plane flexibility of the beam-to-column connections increased

inelastic deformation demands on the gusset plates

337

Bolted shear plate connections resulted in increased residual out-of-plane deformation of

the brace and increased plastic strain at brace hinge resulting in earlier fracture

Specimens with bolted shear plate connections resulted in less contribution of frame

resistance to the total system resistance post brace buckling

Bolted shear plate connections with thin tapered gussets increased post buckling brace

degradation, increased potential for interface weld damage and resulted in earlier brace

fracture

8.3.3 Bolted SCBF Connections

Bolted end plate connections for SCBFs provided the greatest benefits for constructability

while maintaining improved performance for SCBFs

Multiple elements over the brace length for bolted brace to gusset plate connections where

not able to develop the desired buckling capacity or the desired behavior in compression

8.3.4 Net Section Reinforcement Requirement

Relative frame and gusset plate stiffness increased stress demands at brace net section and

increase the potential for net section fracture

Elimination of net section reinforcement would be inadvisable but the AISC design

requirement could be lessened with further study resulting in more economical

connections

8.3.5 Wide-flange Brace Sections

A992 wide-flange sections exhibit increased ductility compared to A500 B/C tubular

sections as brace member in SCBFs

Larger out-of-plane displacements achieved by WF brace result in increased deformation

demands on gusset plate and interface welds

Wide-flange brace section exhibited increased degradation in resistance post-buckling

compared to HSS brace with similar gusset plate connections

8.4 Future Recommendation

Continuation of experimental work would be beneficial for creating the statistical data necessary for

further developing the Balance Design Procedure. The recommended balance factors for yield

mechanism, such as whitmore yielding, are well supported with result showing increased inelastic

338

deformation and strength in SCBFs. However, β factors for failure modes, such as block shear and

whitmore fracture, need further investigation to determine at what design value the failure mode

controls. Tests design to induce a specific failure mode would help finalize the recommendation

for the BDP as a comprehensive design guide for SCBF gusset plate connections.

With regards to net section fracture, multiple test specimens with gusset plates design following the

BDP exhibited no evidence of net section yielding or fracture. HSS-25 and HSS-26 have shown

that completely eliminating reinforcement opens the potential for occurrence of fracture during a

very specific and rare induced ground motion. However, the current AISC specifications for

designing net section reinforcement, including the 0.75 φ factor for rupture and the shear lag

factor , should be further investigated for relevance in SCBF design. Economically, reducing the

amount of time and material required for welding net section reinforcement would result in

substantial reduction in the fabrication of SCBF braces. Experimentally, system or components

test could be conducted investigating brace sections with varying designs for net section reinforcing

to determine an appropriate β for a more economical design.

The bolted end plate connection showed significant promise with regards to both constructability

and performance. This connection is especially beneficial in that considerably less field welding is

required to provide a beam-to-column connection with larger in-plane stiffness and similar

performance as the CJP welded moment connection. Variations of this type of connection

including thinner end plates and different bolt configurations would make valuable test specimens.

339

Appendix A: Specimen Design Calculations

A.1 General

This appendix provides design calculations and commentary for specimens tested within this

sequence. The gusset plate designs were similar for the all of the specimens, however, variations

in the beam-to-gusset connections presented additional design requirements that are also include

within this appendix.

Section 1.2 presents the design calculations for specimen HSS-18 including the bolted shear plate

beam-to-column connection. The design of WF-23 is presented in Section 1.3 to illustrate the

procedure used for the wide-flange brace to gusset plate connection.

A.2 Specimen HSS-18 Design Calculation

The gusset plate design of specimen HSS-18 was identical to HSS-05 with the exception of

interface welds designed for the full capacity of the plate. The calculations presented here reflect

the design process discussed in Chapter 3.

A.2.1 Member Selection

The HSS5x5x3/8 was the predetermined brace size based on the specimen prototype but was

checked to evaluate compactness and slenderness requirements based on the AISC Seismic

Design Manual.

The beam are used to deliver the actuator load into the specimen and required to transfer the

story shear in compression and tension. The compressive capacity was checked, as well as

compactness and slenderness requirements.

Beam compressive capacity using the unbraced length equal to half of the clear span, 66 in.,

because bracing of the weak-axis due to the out-of-plane supports:

340

This exceeds the 400 kip maximum capacity of the actuator. The check for seismic compactness

is as follows.

The W16x45 beam is acceptable. A similar check can be done for the columns. Each column is

to be loaded axial with approximately 400 kips of axial load from the gravity load system. The

unbraced length of the weak-axis is taken as half the span, work-point to work-point, 72 in.,

while the unbraced length of the strong-axis is the full length.

341

The check for seismic compactness is as follows.

The W12x72 column provides sufficient strength and meets the criteria for slenderness and

seismic compactness. These checks to not consider the demands on the framing elements from

the nonlinear response of the brace and the frame at severe drift levels.

A.2.2 Brace Forces

The required compressive and tensile strength of the gusset plates is determined from the

maximum expected brace forces in tension and compression.

The nominal buckling capacity of the brace is required to determine the maximum expected

force in compression. The effective length factor, , is taken as 1.0 and is the actual brace

length.

342

With the brace forces determined above, the design of the gusset plates for specimen HSS-18

were completed using the Balance Design Procedure with balance factors, β, described in

Chapter 3. The design calculations for the gusset plates follow the same step-wise procedure

also described Chapter 3.

A.2.3 Brace to Gusset Plate Design

The brace is connected to the gusset plates by slotting the brace end and sliding the brace over

the gusset plates. The resistance of the connection is to be greater than the maximum expected

brace force in tension. Four fillet welds secure the brace to the gusset. Calculations for sizing

the fillet welds and determining the splice length are described below.

The required length of the connection based on the resistance to shear rupture of the base

material is as follows:

The required length of the connection based on the resistance to shear rupture of the weld is

calculated below using an assumed weld thickness of 5/16”.

The actual splice length, , used in the design was 14.75” to allow for construction tolerances.

At this point, it is recognized that the reduced section of the brace should be checked for net

343

section fracture and reinforcement plates designed. However, the width of the slot depends of

the gusset plate thickness. Net section reinforcement with be design after the required thickness

of the gusset plates have been determined.

A.2.4 Gusset Plate Design

The preliminary thickness of the gusset plate is determined in order to provide resistance to the

tension limit states; whitmore yielding, whitmore rupture, and block shear rupture. The required

thickness for each is calculated below.

Required thickness for whitmore yielding:

Required thickness for whitmore rupture:

Required thickness for block shear rupture:

The preliminary gusset plate thickness was controlled by whitmore yielding and was taken as

3/8”. This thickness is now used to determine the geometry of the gusset plate. Because

determining the geometry with the elliptical clearance is an iterative process, only the final

calculation is shown here. The variables for determining the gusset plate geometry are shown in

Figure A.2.1.

344

Figure A.2.1: Gusset Plate Geometry for Elliptical Clearance Requirement (Kotulka 2007)

The height of the gusset plate, , is taken as 21” and an 7.5tp elliptical clearance is established.

The width of the gusset plate, , is then calculated as:

The location of the brace end is determined as coordinates relative to the free edges of the

gusset plate. The variable is the width of the brace.

345

The corners of the brace end are then determined:

The variables and are then calculated:

Now checks are performed to verify the location of brace end in relation to the elliptical

clearance line. The correct gusset dimensions have been chosen when both of the first

verification equations and one of either of the second verification equations are satisfied.

First verification equations which both must be satisfied are as follows:

346

Also, either of the second verification equations below must be satisfied:

For simplification of fabrication, the actual dimensions used for the gusset plate design were 21”

x 24” to be consistent with HSS-05. The next step was to check the resistance to gusset plate

buckling using the geometry established above and the preliminary gusset plate thickness. An

AutoCAD drawing of the gusset plate was used to determine the dimensions , and as

shown in Figure A.2.2. The compressive resistance is calculated using the Thornton Method

described in Chapter 3.

Figure A.2.2: Buckling Lengths for HSS-18

The required plate thickness to provide sufficient resistance to gusset plate buckling is as follows:

1'-

9"

2'-1"

Gusset Plate

Brace Location

31

4"

11916"

916"

347

The 3/8” preliminary plate thickness was acceptable for gusset plate buckling and will be used in

the final design.

A.2.5 Net Section Reinforcement

Now that the gusset plate thickness was determined, the width of the brace slot is known and the

brace to gusset plate connection can be designed for net section rupture resistance. The check to

determine if net section reinforcement is required is below.

By design, the maximum expected brace force is greater than the resistance to net section

fracture. However, net section reinforcement was not included in the fabrication of specimen

HSS-18. The design of reinforcement plate is shown below for demonstration purposes.

Two plates 3” wide are used to allow for room to weld the plates to the brace.

348

The cross section of each plate on each side of the brace at the reduced section would be 1/4” x

3”. The length of the plates required to develop the full capacity of the plate is determined

below. The plates are connecting with a 3/16” fillet welds all around but only the longitudinal

welds are considered in the design. The material overstrength factor of the plate is included

to account for the expected tensile stress of the plates.

The final dimensions for the next section plates are ¼” x 3” x 10”. The plates are positioned

with 5” on each side of the reduced brace section and welded 3/16” fillet welds.

A.2.6 Interface Weld Design Calculation

The welds connecting the gusset plate to the framing elements are designed for the full plastic

capacity of the gusset plates as describe in Chapter 3. The calculations for sizing the interface

welds are as follows:

The size of the fillet welds required by design was 3/8” on each side.

A.3 Specimen WF-23 Design

This section documents the design of the wide-flange brace to gusset connection described in

Chapter 3, Section 3.3.3. Also included are the section checks for the W6x25 brace section for

seismic compactness and slenderness based on the actual brace length.

A.3.1 W6x25 Brace Section Check

The wide-flange brace size was selected to have a similar maximum expected tensile force as the

HSS5x5x3/8. The expected brace forces for the W6x25 are shown below.

349

The value calculated for the expected brace force of the W6x25 was similar to that of the

HSS5x5x3/8, 398 kip.

NG!

The slenderness ratio of the W6x25 brace was greater than what is specified in the AISC Seismic

Provisions. However, this was noted and considered acceptable in the design of the specimen.

Seismic compactness of the W6x25 is shown below.

A.3.2 Wide-Flange Brace to Gusset Plate Design

The required splice length for the flange to gusset plate connection is calculated below. First, the

length is determined based on the capacity of the flange base material.

350

The required length of the connection based on the resistance to shear rupture of the weld is

calculated below using an assumed weld thickness of 5/16”.

A splice length of 1’-1” was selected to achieve a similar gusset plate design and geometry to

previous test specimens with HSS5x5x3/8 brace sections.

The web plates were design using the following equation to account for the reduced net section

of the brace including shear lag effects.

The net section area of the brace was calculated.

The shear lag factor, U, was determine based on the geometry of the connection cross-section.

The distance from the gusset surface to the center of gravity of the half section, , was calculated

as 1.207 in. The induced eccentricity of the connection was also considered to determine the

most conservative value of .

Two plates, one on each side, 3/8 in. by 3 ½” wide were used.

The length of the web plate and weld are determined to develop the plastic capacity of the plate.

A weld size of ¼” was assumed based on the plate thickness.

351

The total length of the plates, , was calculated at twice the required length plus one

inch clear between the gusset plate end and the brace web.

352

Appendix B: Design Drawings

B.1 Specimen Drawings

Figure B.1.1: Specimen HSS-18

CONNECTION DETAIL AUNIVERSITY OF WASHINGTON

SCALE: 3/8" = 1'-0"

BRACE FRAME PROJECT

SK-HSS18DRAWING

DRAWN BY: JAP NUMBER

Material: Beams and Columns A992, HSS Tubes A500 Grade

B/C, Plates A572 Grade 50, Bolts A490, Welds E70XX

Notes: Beam to column connection adjacent to the gusset plates

consists of bolted shear plates.

See Shop Drawings for fabrication detials and dimensions

of beams and columns.

W16x45

W12x72

W12x72

W16x45

A

HSS5x

5x3/

8

12"x 41

2"x 1'-112" PLATE

34" Ø A490 BOLTS

38"x 1'-2"x 1'-111

2" WEBDOUBLER PLATE

1116" STIFFENER PLATEEACH SIDE TYPICAL

353


CONNECTION DETAIL A UNIVERSITY OF WASHINGTON

SCALE: 3/8" = 1'-0"

BRACE FRAME PROJECT

SK-HSS19DRAWING


Material: Beams ,Columns and WTs A992, HSS Tubes A500

Grade B/C, Plates A572 Grade 50, Bolts A490, Welds E70XX


consists of CJP welded web and flanges.



12"x 41

2"x 1'-112" PLATE

34" Ø A490 BOLTS

38"x 1'-2"x 1'-111

2" WEBDOUBLER PLATE

BEAM WEB TOCOLUM FLANGE

BOTH FLANGES


W16x45

W12x72

W12x72

W16x45

A

HSS5x

5x3/

8

END PLATE TO HSS B

34" SPLICEPLATES

2-WT4x17.5

354


UNIVERSITY OF WASHINGTON

SCALE: 3/8" = 1'-0"

BRACE FRAME PROJECT

SK-HSS20DRAWING





consists of beam endplates bolted with 18-34 " Ø A490

bolts. Only the beam web is welded to the endplate.



W16x45

W12

x72

W12

x72

W16x45

A

12"x 41

2"x 1'-112" PLATE

34" Ø A490 BOLTS

38"x 1'-2"x 1'-111

2" WEBDOUBLER PLATE


HSS5x

5x3/

8

CONNECTION DETAIL A

355



SCALE: 3/8" = 1'-0"

BRACE FRAME PROJECT

SK-HSS21DRAWING





consists of beam endplates bolted with 14-34 " Ø A490

bolts. Only the beam web is welded to the endplate.



W16x45

W12x72

W12x72

W16x45

A

12"x 41

2"x 1'-112" PLATE

34" Ø A490 BOLTS

38"x 1'-2"x 1'-111

2" WEBDOUBLER PLATE


HSS5x

5x3/

8

CONNECTION DETAIL A

356



SCALE: 3/8" = 1'-0"

BRACE FRAME PROJECT

SK-HSS22DRAWING


W16x45

W12x72

W12x72

W16x45

A1

12"x 41

2"x 1'-112" PLATE

34" Ø A490 BOLTS

38"x 1'-2"x 1'-111

2" WEBDOUBLER PLATE


HSS5x

5x3/

8

CONNECTION DETAIL A1




consists of bolted shear plate connection..



357

Figure B.1.6: Specimen WF-23



SCALE: 3/8" = 1'-0"

BRACE FRAME PROJECT

SK-WF23DRAWING





consists of CJP welded flange and web.



12"x 41

2"x 1'-112" PLATE

34" Ø A490 BOLTS

38"x 1'-2"x 1'-111

2" WEBDOUBLER PLATE


BOTH FLANGES


W16x45

W12x72

W12x72

W16x45

A1

W6x

25

SECTION 3-

358




SCALE: 3/8" = 1'-0"

BRACE FRAME PROJECT

SK-HSS24DRAWING





consists of CJP welded flange and bolted web.



W16x45

W12x72

W12x72

W16x45

A1

HSS5x

5x3/

8

12"x 41

2"x 1'-112" PLATE

34" Ø A490 BOLTS

38"x 1'-2"x 1'-111

2" WEBDOUBLER PLATE

BOTH FLANGES


359


W16x89

W12x72

W12x72

W16x89

A

1

14"x 4"x 1'-11

2" PLATE

34" Ø A490 BOLTS

78" GUSSETPLATE

516"


HSS5x

5x3/

8


SCALE: 3/8" = 1'-0"

BRACE FRAME PROJECT

SK-HSS25DRAWING




Notes: Specimen HSS26 interface welds can not be considered

CJP due to insufficient backgouging along entire weld

length. Full backgouge along first (approx.) 6" from

re-entrant corners. Partial penetration weld for remainder.



38"x 1'-2"x 1'-111

2" WEBDOUBLER PLATE



BOTH FLANGES

BACK

GOUGE

360


W16x89

W12x72

W12x72

W16x89

A1

14"x 4"x 1'-11

2" PLATE

34" Ø A490 BOLTS

78" GUSSETPLATE

516"

BACK

GOUGE


HSS5x

5x3/

8


SCALE: 3/8" = 1'-0"

BRACE FRAME PROJECT

SK-HSS26DRAWING




Notes: Specimen HSS26 interface welds can not be considered

CJP due to insufficient backgouging along entire weld

length. Full backgouge along first (approx.) 6" from

re-entrant corners. Partial penetration weld for remainder.

Specimen HSS26 loaded with alternate "Near-Fault"

Loading Protocol.



38"x 1'-2"x 1'-111

2" WEBDOUBLER PLATE



BOTH FLANGES

361

Appendix C: Analysis Plots

C.1 Brace and Gusset Plate Behavior

The appendix contains figures from the analysis not presented in Chapter 6 for the nine test

specimen in this series.

C.2 Brace and Gusset Plate Behavior

In this section, plots from the analysis results that were not included in Chapter 6 are displaced for

reference. The following are included:

Out-of-Plane Displacement at Brace Center (Section C.2.1)

Gusset Plate Rotations (Section C.2.2)

Brace Elongation (C.2.3)

C.2.1 Out-of-Plane Displacement at Brace Center

Brace midpoint out-of-plane displacements are tabulated below. The method used to determine

the displacement at the brace midpoint is described in Chapter 6, Section 6.3.3.1. The displacement

verses drift ratio hysteretic responses are shown in Table C.2.1. Displacement is given in inches.

Table C.2.1: Brace OOP Displacement Hysteresis

HS

S-1

8

HS

S-2

0

362

HS

S-2

1

HS

S-2

2

WF

-23

H

SS

-24

HS

S-2

5

The enveloped brace displacements are plotted over total drift range in Table C.2.2. The plotted

data for HSS-18 and HSS-21 are cut short of the absolute maximum displacement because the

string potentiometer was disconnected from the specimen due to large strains at the brace hinge

location. As a result, the attachment method was modified for later tests. HSS-18 is limited to the

out-of-plane displacement at 4.14%, but continued to a maximum total drift range of 4.18%. HSS-

21 is plotted only to 1.96% total drift range because of early loss of the instrumentation but the

specimen reached 4.14% before failure. The brace displacement for HSS-19 was not included in

this comparison because some of the data was erroneous and the brace did not buckle over the full

length as desired. HSS-26 was also left out because of the nature of the tension dominated

alternate loading protocol.

363

Table C.2.2: Enveloped Brace OOP Displacements

HS

S-1

8

HS

S-2

0

HS

S-2

1

H

SS

-22

WF

-23

HS

S-2

4

HS

S-2

5

C.2.2 Gusset Plate Rotation

The southwest gusset plate rotation and the northeast gusset plate rotation verses drift ratio

hysteretic responses are shown in Table C.2.3. The method of calculating gusset plate rotations are

discussed in Chapter 6, Section 6.3.3.2. The gusset plate rotations for HSS-19 are not included in

364

this comparison. The recorded data from HSS-19 was not relevant considering the brace buckled

down towards the floor and hinges at the splice/extension plate.

Table C.2.3: NE and SW Gusset Plate Rotation Hysteresis

HS

S-1

8

HS

S-1

8

HS

S-2

0

HS

S-2

0

HS

S-2

1

HS

S-2

1

N/A

HS

S-2

2

HS

S-2

2

365

WF

-23

WF

-23

N/A

HS

S-2

4

HS

S-2

4

HS

S-2

5

HS

S-2

5

N/A

The enveloped gusset plate rotations are shown in Table C.2.4 over total drift range. Erroneous

data from the instrumentation forced portions of the gusset plate rotations to be excluded.

Southwest gusset plate rotations for WF-23 are plotted to a total drift range of 1.84% but

continued to a maximum of 5.25%. HSS-25 is plotted to .64% but reached a maximum of 3.27%.

Southwest gusset plate rotation for HSS-21 is not included all together. Northeast gusset plate

rotation for WF-23 is plotted to 4.87% drift total.

366

Table C.2.4: Enveloped NE and SW Gusset Plate Rotations H

SS

-18

HS

S-1

8

HS

S-2

0

HS

S-2

0

HS

S-2

1

HS

S-2

1

N/A

HS

S-2

2

HS

S-2

2

WF

-23

WF

-23

N/A

367

HS

S-2

4

HS

S-2

4

HS

S-2

5

HS

S-2

5

N/A

C.2.3 Brace Elongation

Plots for brace elongation are provided in Table C.2.5 by plotting enveloped elongation over the

total drift range. The brace elongation for specimen HSS-26 is plotted over the positive drift ratio

because information regarding range was not available due to the tension dominated near-fault

loading history. Elongation is given as in./in. and calculated as describe in Chapter 6, Section

6.3.4.1.

Table C.2.5: Enveloped Brace Elongation

HS

S-1

8

HS

S-1

9

368

HS

S-2

0

HS

S-2

1

HS

S-2

2

W

F-2

3

HS

S-2

4

HS

S-2

5

HS

S-2

6*

C.3 Frame Response

Plots for the frame response that were not included in Chapter 6 are tabulated in this section. Plots

for the follow are available:

Column Moments (Section C.3.1)

Column Shears (Section C.3.2)

369

Beam Shear Tab Rotations (C.3.3)

C.3.1 Column Moments

Column moments calculated at the edge of the gusset plate are tabulated below. The method for

determining the moment from the strain gauge records is described in Chapter 6, Section 6.4.1.

Table C.3.1 shown the column moment hysteresis for each column of each specimen. Enveloped

response is shown in Table C.3.2.

Table C.3.1: Column Moments Hysteresis at Edge of Gusset Plate

HS

S-1

8 (

NE

)

HS

S-1

8 (

SW

)

HS

S-1

9 (

NE

)

HS

S-1

9 (

SW

)

HS

S-2

0 (

NE

)

HS

S-2

0 (

SW

)

370

HS

S-2

1 (N

E)

HS

S-2

1 (S

W)

HS

S-2

2 (

NE

)

HS

S-2

2 (

SW

)

WF

-23 (

NE

)

WF

-23 (

SW

)

HS

S-2

4 (

NE

)

HS

S-2

4 (

SW

)

HS

S-2

5 (

NE

)

HS

S-2

5 (

SW

)

371

HS

S-2

6 (

NE

)

HS

S-2

6 (

SW

)

Table C.3.2: Enveloped Column Moments at Edge of Gusset Plate

HS

S-1

8 (

NE

)

HS

S-1

8 (

SW

)

HS

S-1

9 (

NE

)

HS

S-1

9 (

SW

)

HS

S-2

0 (

NE

)

HS

S-2

0 (

SW

)

372

HS

S-2

1 (N

E)

HS

S-2

1 (S

W)

HS

S-2

2 (

NE

)

HS

S-2

2 (

SW

)

WF

-23 (

NE

)

WF

-23 (

SW

)

HS

S-2

4 (

NE

)

HS

S-2

4 (

SW

)

HS

S-2

5 (

NE

)

HS

S-2

5 (

SW

)

373

HS

S-2

6 (

NE

)

HS

S-2

6 (

SW

)

C.3.2 Column Shears

The calculated shear responses for each column are presented in the tables below. Column shears

were calculated using the method described in Chapter 6, Section 6.4.2. The hysteretic response of

the column shear over drift ratio is shown in Table C.3.3. Enveloped values of column shear are

shown in Table C.3.4. The shear value is normalized by , as described in Equation 6.4.1.

Table C.3.3: Column Shear Hysteresis

HS

S-1

8 (

NE

)

HS

S-1

8 (

SW

)

HS

S-1

9 (

NE

)

HS

S-1

9 (

SW

)

374

HS

S-2

0 (

NE

)

HS

S-2

0 (

SW

)

HS

S-2

1 (N

E)

H

SS

-21

(SW

)

HS

S-2

2 (

NE

)

HS

S-2

2 (

SW

)

WF

-23 (

NE

)

WF

-23 (

SW

)

HS

S-2

4 (

NE

)

HS

S-2

4 (

SW

)

375

HS

S-2

5 (

NE

)

HS

S-2

5 (

SW

)

HS

S-2

6 (

NE

)

HS

S-2

6 (

SW

)

Table C.3.4: Enveloped Column Shears

HS

S-1

8 (

NE

)

HS

S-1

8 (

SW

)

HS

S-1

9 (

NE

)

HS

S-1

9 (

SW

)

376

HS

S-2

0 (

NE

)

HS

S-2

0 (

SW

)

HS

S-2

1 (N

E)

H

SS

-21

(SW

)

HS

S-2

2 (

NE

)

HS

S-2

2 (

SW

)

WF

-23 (

NE

)

WF

-23 (

SW

)

HS

S-2

4 (

NE

)

HS

S-2

4 (

SW

)

377

HS

S-2

5 (

NE

)

HS

S-2

5 (

SW

)

HS

S-2

6 (

NE

)

HS

S-2

6 (

SW

)

C.3.3 Beam Shear Tab Rotations

The rotation at the shear tab connection is calculated below. The method for calculating rotation is

described in Chapter 6, Section 6.4.4. Shear tab rotations for specimen HSS-19 were not included

because of errors in the instrumentation record. Hysteretic plots of rotation over drift ratio are

shown in Table C.3.5 and enveloped values are shown in Table C.3.6.

Table C.3.5: Beam-to-Column Shear Tab Rotations

HS

S-1

8 (

NW

)

HS

S-1

8 (

SE

)

378

HS

S-2

0 (

NW

)

HS

S-2

0 (

SE

)

HS

S-2

1 (N

W)

H

SS

-21

(SE

)

HS

S-2

2 (

NW

)

HS

S-2

2 (

SE

)

WF

-23 (

NW

)

WF

-23 (

SE

)

HS

S-2

4 (

NW

)

HS

S-2

4 (

SE

)

379

HS

S-2

5 (

NW

)

HS

S-2

5 (

SE

)

HS

S-2

6 (

NW

)

HS

S-2

6 (

SE

)

Table C.3.6: Enveloped Shear Tab Rotations

HS

S-1

8 (

NW

)

HS

S-1

8 (

SE

)

HS

S-2

0 (

NW

)

HS

S-2

0 (

SE

)

380

HS

S-2

1 (N

W)

HS

S-2

1 (S

E)

HS

S-2

2 (

NW

)

H

SS

-22 (

SE

)

WF

-23 (

NW

)

WF

-23 (

SE

)

HS

S-2

4 (

NW

)

HS

S-2

4 (

SE

)

HS

S-2

5 (

NW

)

HS

S-2

5 (

SE

)

381

HS

S-2

6 (

NW

)

HS

S-2

6 (

SE

)

382

References

1. Adel, E.T., Goel, S., “Cyclic Behavior of Angle X-Bracing with Welded Connections,” Report UMCE 85-4, Department of Civil Engineering, University of Michigan, Ann Arbor, Michigan, April 1985

2. AISC “Steel Construction Manual,” 13th Edition, American Institute of Steel Construction, Chicago, Illinois, 2005

3. AISC “Manual of Steel Construction, Load and Resistance Factor Design,” 3rd Edition, American Institute of Steel Construction, Chicago, Illinois, 2001

4. AISC “Seismic Design Manual,” American Institute of Steel Construction, Chicago, Illinois, 2005

5. Aslani, F., Goel, S., “Experimental and Analytical Study of the Inelastic Behavior of Double Angle Bracing Members Under Severe Cyclic Loading,” Research Report UMCE 89-5, Department of Civil Engineering, University of Michigan, Ann Arbor, Michigan, February, 1989

6. Aslani, F., Goel, S., Xu, P., “Effect of Stitch Spacing on the Cyclic Behavior of Built-up Bracing Members,” Report UMCE 87-8, University of Michigan, Ann Arbor, Michigan, January, 1987

7. Astaneh-Asl, A., "Seismic Behavior and Design of Gusset Plates," Steel TIPS, Structural Steel Educational Council, Moraga, California, December, 1998

8. Astaneh-Asl, A., Goel, S.C., and Hanson, R.D., “Cyclic Behavior of Double Angle Bracing Members with End Gusset Plates,” Research Report UMEE 82R7, Department of Civil Engineering, University of Michigan, Ann Arbor, Michigan, August, 1982

9. ATC 24, "Guidelines for Cyclic Seismic Testing of Components of Steel Structures," Applied Technology Council, 1992

10. Becker, R., "Seismic Design of Special Concentrically Braced Steel Frames," Steel TIPS, Structural Steel Educational Council, Moraga, California, November, 1995

11. Beer, P.F., Johnston, E.R., DeWolf, J.T., “Mechanics of Materials,” Third Edition, McGraw Hill Publishers, c2001

12. Brown, V.L.S., "Stability of Gusseted Connections in Steel Structures," A thesis submitted in partial fulfillment of Doctor of Philosophy in Civil Engineering, University of Delaware, 1988

13. Bruneau, M., Uang, C.M., Whittaker, A., “Ductile Design of Steel Structures,” First Edition, McGraw-Hill Publishing, c1998

14. Celik, O.C., Berman, J.W., Bruneau, M., “Cyclic Testing of Braces Laterally Restrained by Steel Studs,” ASCE, Journal of Structural Engeering, July 2005

15. Chen, C.H., Lai, J.W., Mahin, S., “Numerical Modelling and Performance Assessment of Concentrically Brace Steel Frames,” ASCE, Structures 2008: Crossing Borders, 2008

383

16. Cheng, J.J.R., Grondin, G.Y., Yam M.C.H, "Design and Behavior of Gusset Plate

Connections," Fourth International Workshop on Connections in Steel Structures, Roanoke, VA, October 2000

17. Christopulos, A.S., "Improved Seismic Performance of Buckling Restrained Braced Frames", Department of Civil Engineering, University of Washington, Seattle, Washington, June, 2005

18. Clark. K.A., “Experimental Performance of Multi-Story X-Brace Systems,” Department of Civil Engineering, University of Washington, Seattle, Washington, December 2009

19. Cochran, M., Honeck, W.C., "Design of Special Concentric Braced Frames," Steel TIPS, Structural Steel Educational Council, Moraga, California, May, 2004

20. Foutch, D.A., Goel, S.C., Roeder, C.W., “Seismic Testing of Full-Scale Steel Building-Part I,” ASCE, Journal of Structural Engeering, p 2111-2129, October 1986

21. Fahnestock, L.A., Stoakes, C.D., “Cyclic Behavior and Perofrmance of Beam-Column Connections in Concentrically Braced Frames,” ASCE, Structures 2008: Crossing Borders, 2008

22. Foutch, D.A., Goel, S.C., Roeder, C.W., “Preliminary Report on Seismic Testing of a Full-Scale Six Story Steel Building,” University of Illinois, Urbana-Champaign, Urbana, Illinois, November 1986

23. Grondin, G.Y., Nast, T.E., and Cheng, J.J.R., "Strength and Stability of Corner Gusset Plates Under Cyclic Loading," Proceedings of Annual Technical Session and Meeting, Structural Stability Research Council, 2000

24. Gugerli, H., Goel, S.C., “Inelastic Cyclic Behavior of Steel Bracing Frames,” Report UMEE 82R1, University of Michigan, Ann Arbor, Michigan, January, 1982

25. Gunnerson, I., "Numerical Performance Evaluation of Braced Frame Systems", Department of Civil Engineering, University of Washington, Seattle, Washington, December, 2004

26. Haddad, M., Tremblay, R., “Influence of connection design on the inelastic seismic response of HSS steel bracing members,” Tubular Structures XI, p639-646, 2006

27. Hardash, S, Bjorhovde, R., "New Design Criteria for Gusset Plates in Tension," Engineering Journal, AISC, Vol. 22, No. 2, Second Quarter, 1985

28. Hu, S.Z., Cheng, J.J.R., “Compressive Behavior of Gusset Plate Connections,” University of Alberta, Department of Civil Engineering, Structural Engineering Report, n153, July, 1987

29. Herman, D.J., “Further Improvements on and Understanding of Special Concentrically Brace Frame Systems,” Department of Civil Engineering, University of Washington, Seattle, Washington, 2007

30. Hsiao, P.C., “Simulation Methods for Special Concentrically Braced Frames”, General Examination, Department of Civil Engineering, University of Washington, Seattle, Washington, December 2009

384

31. Jain, A.K., Goel, S.C., Hanson, R.D, "Hysteresis Behavior of Bracing Members and Seismic Response of Braced Frames with Different Proportions," Research Report UMEE 78R3, Department of Civil Engineering, University of Michigan, Ann Arbor, Michigan, .July, 1978

32. Johnson, S.M., “Improved Seismic Performance of Special Concentrically Brace Frames,” Department of Civil Engineering, University of Washington, Seattle, WA, June 2005

33. Kotulka, B.A., “Analysis for a Design Guide on Gusset Plates used in Special Concentrically Brace Frames,” Department of Civil Engineering, University of Washington, Seattle, Washington, 2007

34. Lehman, D.E., Roeder, C.W., Herman, D., Johnson, S., Kotulka, B., “Improved Seismic Performance of Gusset Plate Connections,” ASCE, Journal of Structural Engeering, p 890-901, June 2008

35. Lehman, D., Roeder, C., Jung H. Y., Johnson, S., "Seismic Response of Braced Frame Connections," 13th World Conference on Earthquake Engineering, Vancouver, B.C., Canada, Paper No. 1459, August 2004

36. Lesik, D.F., Kennedy, D.J.L., "Ultimate Strength of Fillet Welded Connections Loaded in Plane," Canadian Journal of Civil Engineering, Vol. 17, No. 1, National Research Council of Canada, Ottawa, Canada, 1990

37. Lin, M.E., Tsai, K.C., Hsiao, P.C., Tsai, C.Y., “Compressive Behavior of Buckling Restrained Brace Gusset Connections,” The First International Conference on Advances in Experimental Structural Engineering, Nagoya, Japan, July 2005

38. Liu, J., Astaneh-Asl, A., “Moment-Rotation Parameters for Composite Shear Tab Connections,” ASCE, Journal of Structural Engineering, p 1371-1380, September 2004

39. MacRae, G.A., Kimura, Y., Roeder, C.W., “Effect of Column Stiffness on Braced Frame Seismic Behavior,” ASCE, Journal of Structural Engineering, p 381-391, March 2004

40. Murphy, G., "Advanced Mechanics of Materials," New York and London, McGraw-Hill Book Company, Inc., c1946

41. Nast, T., Grondin, G., Cheng, R., “Cyclic Behavior of Stiffened Gusset Plate-Brace Member Assemblies,” University of Alberta, Department of Civil Engineering, Structural Engineering Report, n229, December, 1999

42. Powell, J.A., “HSS-29: Test Summary”, University of Washington, Seattle, Washington, December 2009

43. Powell, J.A., “Evaluation of Special Concentrically Braced Frames for Improved Seismic Performance and Constructability”, University of Washington, Seattle, Washington, December 2009

44. Powell, J.A., Clark, K.A., “Test of a Full Scale Concentrically Brace Frame with Multi-Story X-Bracing,” ASCE, Structures 2008: Crossing Borders, 2008

45. Rabinovitch, J., Cheng, R., “Cyclic behavior of steel gusset plate connections,” University of Alberta, Department of Civil Engineering, Structural Engineering Report, n 191, August, 1993

385

46. Richards, P.W., “Seismic Column Demands in Ductile Brace Frames,” ASCE, Journal of

Structural Engineering, p 33-41, January 2009

47. Roeder C.W., "Connection Performance for Seismic Design of Steel Moment Frames," ASCE, Journal of Structural Engineering, p517-525, April, 2002

48. Roeder, C.W., MacRae, G., Leland, A., Rospo, A. "Extending the Fatigue Life of Riveted Coped Stringer Connections." Journal of Bridge Engineering, ASCE, v 10, n 1, p 69-76 January/February 2005.

49. Roeder, C.W., Lehman, D.E., Yoo, J.H., “Improved Seismic Design of Steel Frame Connections”, Steel Structures, p 141-153, 2005

50. Roeder, C.W., Lehman, D.E. "Performance-Based Seismic Design of Concentrically Braced Frames", National Science Foundation, Grant CMS-0301792

51. Roeder, C.W., “Seismic Behavior of Concentrically Brace Frame,” ASCE, Journal of Structural Engineering, p1837-1856, August 1989

52. SAC Steel Project, "Protocol for Fabrication, Inspection, Testing and Documentation of Beam-Column Connection Tests and Other Experiments," Report No. SAC/BD-97/02, SAC Joint Venture, October 1997

53. Salmon, C.G., Johnson, J.E., "Steel Structures Design and Behavior," 4th edition, HarperCollins College Publishers, c1996

54. Segui, W.T., “Steel Design,” 4th Edition, Thomson Publishers, c2007

55. Shaback, B., Brown, T., "Behaviour of square hollow structural steel braces with end connections under reversed cyclic axial loading," Canadian Journal of Civil Engineering, v 30, n 4, p 745-753, August, 2003

56. Tam M.C.H., Cheng J.J.R., "Behavior and Design of Gusset Plate Connections in Compression," Journal of Constructional Steel Research, Vol 58, No. 5-8, Elsevier, pgs 1143-59, 2002

57. Tamboli, A.R., "Handbook of structural steel connection design and details," New York, McGraw-Hill, c1999

58. Timoshenko, S.P., Gere, J.M., "Theory of Elastic Stability," New York, McGraw-Hill Book Company, Inc., c1961

59. Thornton, W.A., "Bracing Connections for Heavy Construction," Engineering Journal, AISC, Vol. 21, No. 3, pp. 139-148., 1984

60. Tremblay R., "Inelastic seismic response of steel bracing members," Journal of Constructional Steel Research, 58, 665-701, 2002

61. Uriz, P., "Summary Of Test Results For UC Berkeley Special Concentric Braced Frame Specimen No. 1 (Scbf-1)", Retrieved May 5, 2005, from http://www.ce.berkeley.edu/~patxi/

http://www.engineeringvillage2.org./controller/servlet/Controller?CID=quickSearchCitationFormat&searchWord1=%7bShaback%2C+Brad%7d&section1=AU&database=1&startYear=1884&endYear=2005&yearselect=yearrange&sort=yr

386

62. Uriz, P., “Towards Earthquake Resistant Design of Concentrically Brace Steel Structures”, Department of Civil and Environmental Engineering, University of California, Berkeley, Berkeley, California, Fall 2005

63. Wakabayshi, M., Nakamura, T., “Experimental Studies on the Elastic-Plastic Behavior of Brace Frames under Repeated Horizontal Loading, Part 1” Kyoto University, Kyoto, Japan, September 1977

64. Wakabayshi, M., Nakamura, T., Yoshida, N., “Experimental Studies on the Elastic-Plastic Behavior of Brace Frames under Repeated Horizontal Loading, Part 2” Kyoto University, Kyoto, Japan, March 1980

65. Walpole, W. R. “Behaviour of cold-formed steel RHS members under cyclic loading.” Dept. of Civil Engineering, Univ. of Canterbury, Christchurch, New Zealand, 1996

66. Whitmore, R.E., "Experimental Investigation of Stresses in Gusset Plates," Bulletin No. 16, Engineering Experiment Station, University of Tennessee, 1952

67. Yam M.C.H., “Compressive Behavior and Strength of Steel Gusset Plate Connections,” University of Alberta, Department of Civil Engineering, Fall, 1987

68. Yang, F., Mahin, S., "Limiting Net Section Fracture in Slotted Tube Braces," Steel TIPS, Structural Steel Educational Council, Moraga, California, April 2005

69. Yoo, J.H., "Analytical Investigation on the Seismic Performance of Special Concentrically Braced Frames," Department of Civil Engineering, University of Washington, Seattle, Washington, June 2006

70. Yoo, J.H., Lehman, D.E., Roeder, C.W., “Influence of Connection Design Parameters on the Seismic Performance of Braced Frames,” Journal of Constructional Steel Research, p607-623, June 2008

71. Yoo, J.H., “Simulated Behavior of Multi-Story X-Braced Frames,” Engineering Structures, p182-197, 2008

72. Zhiyuan, L., Goel, S.C., "Investigation of Concrete-Filled Steel Tubes Under Cyclic Bending and Buckling," Research Report UMEE 87-3, Department of Civil Engineering, University of Michigan, Ann Arbor, Michigan, April, 1987

387

Jake Powell Thesis (HSS18-HSS26)

Documents

braced frames scbfs

gusset plate design

design engineers

entire system

masters thesis

improved seismic performance

moment frames

system ductility