Economical Structural Steel Work 1

economical structuralsteelwork

student edition - 2004

(ABN)/ACN (94) 000 973 839

AUSTRALIAN STEEL INSTITUTE

1

economical structuralsteelwork

student edition - 2004

(ABN)/ACN (94) 000 973 839

AUSTRALIAN STEEL INSTITUTE

2

3

ECONOMICAL STRUCTURAL STEELWORK

Student Edition 2004Limit States Edition to AS 4100 -- 1990

AUSTRALIAN STEEL INSTITUTE --1979, 1984, 1991, 1996, 2004

NATIONAL LIBRARY OF AUSTRALIA CARD NUMBER AND ISBN 0909945 77 2

FIRST EDITION 1979

SECOND EDITION 1984

THIRD EDITION 1991

REPRINTED 1992, 1995

FOURTH EDITION 1996

E Australian Institute of Steel Construction 1996

While every effort has been made and all reasonable care taken to ensurethe accuracy of the material contained herein the Authors, Editors andPublishers of this Publication shall not be held to be liable or responsible inany waywhatsoever for any loss or damage costs or expenses howsoeverincurred by any person whether the purchaser of this work or otherwiseincluding but without in any way limiting any loss or damage costs orexpenses incurred as a result of or in connection with the reliance whetherwhole or partial by any person as aforesaid upon any part of the contents ofthis publication.

Should expert assistance be required, the services of a competentprofessional person should be sought.

4

FOREWORD

’Economical Structural Steelwork’ was first published in 1979 and quicklybecame the Institute’s most popular publication among practisingdesigners and students.

This fourth edition has been up--dated in its references to AustralianStandards, practices and cost concepts, and has other amendments. Itcontinues to provide the useful practical advice given by its previouseditions towards the achievement of the optimum result in structuralsteelwork.

E Australian Institute of Steel Construction 1996

While every effort has been made and all reasonable care taken to ensurethe accuracy of the material contained herein the Authors, Editors andPublishers of this Publication shall not be held to be liable or responsible inany waywhatsoever for any loss or damage costs or expenses howsoeverincurred by any person whether the purchaser of this work or otherwiseincluding but without in any way limiting any loss or damage costs orexpenses incurred as a result of or in connection with the reliance whetherwhole or partial by any person as aforesaid upon any part of the contents ofthis publication.

Should expert assistance be required, the services of a competentprofessional person should be sought.

5

Preface

When considering steel structures it is not difficult to obtain information onengineering and technological aspects, but very little is available on how tochoose steelwork economically. Yet more and more the viability of abuilding project depends upon critical financial considerations. Thus it isimportant for designers to have a good general appreciation of thecomponents thatmake up the cost of fabricated steel, and of howdecisionsmade at the design stage can influence these costs.

This publication aims to supply some of this information. It is not a designmanual, but rather a publication that discusses from a cost point of view allof the matters that a structural steel designer should consider. It takes intoaccount current fabrication practices and material/labour relationships,both of which have changed markedly over the last few years.

Adherence to the principles outlined in this publication will do much toassist designers in reaching decisions that will lead to effective andeconomic structures.

It should be noted that this edition has substantially adopted therationalised approach to the costing of fabricated steel by using a cost per

metre for sections and cost per square metre for plates, depending onthe size, in lieu of cost per tonne. The reasoning behind this is presentedin a paper entitled: ”A Rational Approach to Costing Steelwork” by T Main,K B Watson and S Dallas. This paper was presented at the InternationalCost Engineering Council/The Australian Institute of Quantity SurveyorsInternational Symposium, Construction Economics -- The EssentialManagement Tool, Australia, May 1995.

We wish to thank all those who have contributed to this publication andspecial acknowledgment goes to all AISC Staff who submitted commentson the technical and editorial content of this publication.

edited by: Jose’ R ZaragozaBScCE,

MScCE, CPEng.

AISC StateManager--N

SW

6

ECONOMICAL STRUCTURAL STEELWORKContents

FOREWORD 4. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Preface 5. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1. Preliminary Considerations 1. . . . . . . . . . . . . . . . .

1.1. Introduction 1. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1.2. Factors influencing Framing Cost 1. . . . . . . . . . . .1.3. Integrated Design 2. . . . . . . . . . . . . . . . . . . . . . . . . .

2. General Factors Affecting Economy 3. . . . . . . . .

2.1. Steel Grades 3. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .2.1.1. STRUCTURAL STEEL 3. . . . . . . . . . . . . . . . . . . . . . .2.1.2. WEATHERING STEEL 3. . . . . . . . . . . . . . . . . . . . . . .2.1.3. HOLLOW SECTIONS 3. . . . . . . . . . . . . . . . . . . . . . . .2.1.4. QUENCHED AND TEMPERED STEEL 4. . . . . . . . .2.1.5. CHOICE OF STEEL GRADE 4. . . . . . . . . . . . . . . . . .

2.2. Economy in use of Material 6. . . . . . . . . . . . . . . . .2.2.1. STEEL PRICING 6. . . . . . . . . . . . . . . . . . . . . . . . . . . .2.2.2. PLATES 6. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .2.2.3. SECTIONS 7. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .2.2.4. SCRAP AND WASTE 7. . . . . . . . . . . . . . . . . . . . . . . .

2.3. Fabrication 7. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .2.3.1. GENERAL 7. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .2.3.2. BEAM AND COLUMN FABRICATION 8. . . . . . . . . .2.3.3. GIRDER AND TRUSS FABRICATION 8. . . . . . . . . .2.3.4. SUMMARY FOR ECONOMIC FABRICATION 9. . .

2.4. Erection 9. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

2.4.1. GENERAL CONSIDERATIONS 9. . . . . . . . . . . . . . .2.4.2. HANDLING AND TRANSPORT 10. . . . . . . . . . . . . . .2.4.3. CONNECTIONS 10. . . . . . . . . . . . . . . . . . . . . . . . . . . .2.4.4. FIELD BOLTING 11. . . . . . . . . . . . . . . . . . . . . . . . . . . .2.4.5. FIELD WELDING 12. . . . . . . . . . . . . . . . . . . . . . . . . . . .2.4.6. BRACING 12. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

2.5. Surface Treatment 12. . . . . . . . . . . . . . . . . . . . . . . . .2.5.1. GENERAL CONSIDERATIONS 12. . . . . . . . . . . . . . .2.5.2. STEEL PERFORMANCE 13. . . . . . . . . . . . . . . . . . . . .2.5.3. SURFACE PREPARATION 13. . . . . . . . . . . . . . . . . . .2.5.4. PAINT SYSTEMS 13. . . . . . . . . . . . . . . . . . . . . . . . . . .2.5.5. HOT--DIP GALVANIZING 14. . . . . . . . . . . . . . . . . . . . .2.5.6. DESIGN AND DETAILS FOR CORROSION RESISTANCE .

142.5.7. SUMMARY CHECKLIST FOR SURFACE TREATMENT . . . .

14

2.6. Fire Resistance 15. . . . . . . . . . . . . . . . . . . . . . . . . . . .2.6.1. GENERAL CONSIDERATIONS 15. . . . . . . . . . . . . . .2.6.2. REGULATORY REQUIREMENTS 15. . . . . . . . . . . . .2.6.3. MATERIALS FOR FIRE PROTECTION 16. . . . . . . .

2.7. Specifications 16. . . . . . . . . . . . . . . . . . . . . . . . . . . . .2.7.1. GENERAL CONSIDERATIONS 16. . . . . . . . . . . . . . .2.7.2. WORKMANSHIP STANDARDS 17. . . . . . . . . . . . . . .2.7.3. TOLERANCES 17. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .2.7.4. CAMBERING 18. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .2.7.5. TEMPORARY BRACING 19. . . . . . . . . . . . . . . . . . . . .2.7.6. INSPECTION 19. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .2.7.7. SUMMARY FOR SPECIFICATION WRITERS 20. . .

7

3. Framing Concepts and Connection Types 21. . . .

3.1. Introduction 21. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .3.2. Connection Types 21. . . . . . . . . . . . . . . . . . . . . . . . .

3.2.1. DESIGN METHODS IN AS 4100 21. . . . . . . . . . . . . .3.2.2. FLEXIBLE CONNECTIONS 22. . . . . . . . . . . . . . . . . . .3.2.3. RIGID CONNECTIONS 23. . . . . . . . . . . . . . . . . . . . . .

3.3. Basic Framing Systems 24. . . . . . . . . . . . . . . . . . . .3.3.1. TWO--WAY RIGID FRAMEWORK (Fig 3.3) 24. . . . .3.3.2. ONE--WAY RIGID FRAMEWORK (Fig 3.4) 24. . . . . .3.3.3. TWO--WAY BRACED FRAMEWORK (Fig 3.5) 25. .3.3.4. SUMMARY OF FRAMING SYSTEMS 26. . . . . . . . . .3.3.5. STABILISING ELEMENTS 27. . . . . . . . . . . . . . . . . . . .

3.4. Cost and Framing System 29. . . . . . . . . . . . . . . . . .3.4.1. MULTI--STOREY BUILDING 30. . . . . . . . . . . . . . . . . .3.4.2. SINGLE--STOREY INDUSTRIAL BUILDING 31. . . .

3.5. Framing Details 31. . . . . . . . . . . . . . . . . . . . . . . . . . .3.5.1. SYMMETRY 31. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .3.5.2. RATIONALISATION OF MEMBERS 32. . . . . . . . . . .3.5.3. STANDARDIZATION 32. . . . . . . . . . . . . . . . . . . . . . . . .3.5.4. SIMPLICITY 32. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

3.6. Conclusion 33. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

4. Industrial Buildings 34. . . . . . . . . . . . . . . . . . . . . . . .

4.1. Introduction 34. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .4.2. Warehouse and Factory Buildings 34. . . . . . . . . . .

4.2.1. GENERAL 34. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .4.2.2. STANDARDIZED PORTAL FRAMES 35. . . . . . . . . . .4.2.3. CUSTOM DESIGNED PORTAL FRAMES 36. . . . . .4.2.4. BRACING OF PORTAL FRAMES 37. . . . . . . . . . . . .4.2.5. CRANES IN PORTAL FRAME BUILDINGS 40. . . . .4.2.6. PURLINS 41. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .4.2.7. FLY BRACING 42. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .4.2.8. SHEETING 43. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

4.3. Large Span Storage Buildings 43. . . . . . . . . . . . . . .4.3.1. SPANS OF 45--70 METRES 43. . . . . . . . . . . . . . . . . .

4.3.2. SPANS IN EXCESS OF 70 METRES 44. . . . . . . . . .4.4. Heavy Industrial Structures 44. . . . . . . . . . . . . . . . .

4.4.1. ERECTION 45. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .4.4.2. SITE WELDING 45. . . . . . . . . . . . . . . . . . . . . . . . . . . . .4.4.3. BOLTED CONNECTIONS 45. . . . . . . . . . . . . . . . . . . .4.4.4. FUNCTIONAL CONSTRAINTS 45. . . . . . . . . . . . . . . .

5. Commercial Buildings 46. . . . . . . . . . . . . . . . . . . . . .

5.1. Introduction 46. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .5.2. Low--Rise Commercial Buildings 46. . . . . . . . . . . . .

5.2.1. FULLY STEEL--FRAMED 46. . . . . . . . . . . . . . . . . . . . .5.2.2. COMPOSITE FRAMES 47. . . . . . . . . . . . . . . . . . . . . .

5.3. High--Rise Commercial Buildings 47. . . . . . . . . . . .5.3.1. GENERAL 47. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .5.3.2. FULLY RIGID FRAME 50. . . . . . . . . . . . . . . . . . . . . . .5.3.3. FULLY BRACED FRAMES 50. . . . . . . . . . . . . . . . . . .5.3.4. STABILITY BY MEANS OF SERVICE CORES 51. .

5.4. Floor Support Systems 52. . . . . . . . . . . . . . . . . . . . .5.5. Composite Construction 54. . . . . . . . . . . . . . . . . . . .

5.5.1. FLOOR SYSTEMS 54. . . . . . . . . . . . . . . . . . . . . . . . . .5.5.2. COLUMNS 55. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

5.6. Summary 56. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

6. Bolting 57. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

6.1. Introduction 57. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .6.2. Bolt Types 57. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

6.2.1. COMMERCIAL BOLTS 57. . . . . . . . . . . . . . . . . . . . . . .6.2.2. HIGH--STRENGTH STRUCTURAL BOLTS 57. . . . .

6.3. Bolting Categories 58. . . . . . . . . . . . . . . . . . . . . . . . .6.4. Factors Affecting Bolting Economy 59. . . . . . . . . . .

6.4.1. BOLT GRADE 59. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .6.4.2. BOLT DIAMETER 59. . . . . . . . . . . . . . . . . . . . . . . . . . .6.4.3. BOLTING CATEGORY 59. . . . . . . . . . . . . . . . . . . . . . .6.4.4. THREADS IN OR OUT OF SHEAR PLANE 60. . . . .6.4.5. BOLT FINISH 60. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .6.4.6. INSPECTION 60. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

8

6.5. Summary for Economic Bolting 61. . . . . . . . . . . . . .6.5.1. CHECKLIST 61. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .6.5.2. BOLT USAGE -- FLEXIBLE JOINTS 62. . . . . . . . . . .6.5.3. BOLT USAGE--RIGID JOINTS 63. . . . . . . . . . . . . . . .

7. Welding 64. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

7.1. Introduction 64. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .7.1.1. PRINCIPLES FOR ECONOMY 64. . . . . . . . . . . . . . . .7.1.2. COST COMPONENTS 64. . . . . . . . . . . . . . . . . . . . . . .

7.2. Types of Welds 65. . . . . . . . . . . . . . . . . . . . . . . . . . . .7.2.1. FILLET WELDS (see Fig 7.1) 65. . . . . . . . . . . . . . . . .7.2.2. BUTT WELDS (see Fig 7.2) 66. . . . . . . . . . . . . . . . . . .7.2.3. BUTT WELDS vs. FILLET WELDS 66. . . . . . . . . . . .

7.3. Welding Processes 67. . . . . . . . . . . . . . . . . . . . . . . .7.4. Other Cost Factors 68. . . . . . . . . . . . . . . . . . . . . . . .

7.4.1. WELD CATEGORIES 68. . . . . . . . . . . . . . . . . . . . . . . .7.4.2. WELDING SPECIFICATIONS 69. . . . . . . . . . . . . . . . .7.4.3. WELDING INSPECTION 69. . . . . . . . . . . . . . . . . . . . .

7.5. Economical Design and Detailing 70. . . . . . . . . . . .

8. Detailing for Economy 75. . . . . . . . . . . . . . . . . . . . . .

8.1. Detailing on Design Engineer’s Drawings 75. . . . .8.2. Beams 76. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

8.2.1. GENERAL 76. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .8.2.2. PLATED SECTIONS 76. . . . . . . . . . . . . . . . . . . . . . . . .8.2.3. WEB PENETRATIONS IN BEAMS 76. . . . . . . . . . . . .8.2.4. CASTELLATED BEAMS 77. . . . . . . . . . . . . . . . . . . . . .8.2.5. THREE--PLATE GIRDERS 78. . . . . . . . . . . . . . . . . . . .

8.3. Columns 80. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .8.3.1. GENERAL 80. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .8.3.2. COLUMN BASE PLATES 80. . . . . . . . . . . . . . . . . . . . .8.3.3. HOLDING--DOWN BOLTS 82. . . . . . . . . . . . . . . . . . . .8.3.4. COLUMN SPLICES 82. . . . . . . . . . . . . . . . . . . . . . . . . .8.3.5. COLUMN STIFFENERS 84. . . . . . . . . . . . . . . . . . . . . .8.3.6. BUILT--UP COLUMNS 85. . . . . . . . . . . . . . . . . . . . . . .

8.4. Trusses 87. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .8.5. Portal Frames 90. . . . . . . . . . . . . . . . . . . . . . . . . . . . .

8.5.1. CONNECTIONS 90. . . . . . . . . . . . . . . . . . . . . . . . . . . .8.5.2. PORTAL FRAME PRE--SET 91. . . . . . . . . . . . . . . . . .

8.6. Connection Detailing 92. . . . . . . . . . . . . . . . . . . . . . .8.6.1. GENERAL 92. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .8.6.2. SPECIFIC CONNECTIONS 93. . . . . . . . . . . . . . . . . . .

9. References & Further Reading 103. . . . . . . . . . . . . .

10. Standards 105. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

9

List of Tables

TABLE 2.1 -- AVAILABILITY OF PRODUCTS BY GRADE(check currency of information with steel suppliers) 4. . . .

TABLE 2.2 INDICATIVE COST RATIOS FOR DIFFERENTGRADES OF STRUCTURAL STEEL(per tonne, supply only) 5. . . . . . . . . . . . . . . . . . . . . . . . . . . . .

TABLE 2.3 PREFERRED STEEL PLATE THICKNESSES(in mm) 7. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

TABLE 2.4 TRANSPORTATION COSTS 10. . . . . . . . . . . . . . . . .TABLE 2.5 SURFACE TREATMENT COSTS 13. . . . . . . . . . . . .

TABLE 2.6 PASSIVE FIRE PROTECTION COSTS 16. . . . . . .TABLE 5.1 SHEAR WALL vs LATTICE BRACING 52. . . . . . . . .TABLE 6.1 BOLT TYPES AND BOLTING CATEGORIES 58. . .TABLE 6.2 INDICATIVE COST RATIOS OF

DIFFERENT BOLT DIAMETERS 59. . . . . . . . . . . . . . . . . . . .Table 6.3 INDICATIVE COST RATIOS OF

DIFFERENT BOLTING CATEGORIES 60. . . . . . . . . . . . . . . .TABLE 7.1 FILLET WELD COMPARISON 65. . . . . . . . . . . . . . . .TABLE 8.1 WRENCH CLEARANCES 89. . . . . . . . . . . . . . . . . . .

10

List of Figures

Fig. 2.1 Deep end plates can cause jamming 11. . . . . . . . . . . . .Fig. 2.2 Two ways of avoiding the problem

of access to column web connections 11. . . . . . . . . . . . . . . .Fig 2.3 Typical connections where allowance for

mill tolerance is needed 18. . . . . . . . . . . . . . . . . . . . . . . . . . . .Fig 3.1 Flexible connections 22. . . . . . . . . . . . . . . . . . . . . . . . . . . .Fig 3.2 Rigid connections 24. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Fig 3.3 Two--way rigid framework 24. . . . . . . . . . . . . . . . . . . . . . . .Fig 3.4 One--way braced, one--way rigid framework 25. . . . . . . .Fig 3.5 Two--way braced framework 26. . . . . . . . . . . . . . . . . . . . .Fig 3.6 Stabilising elements built in steel 27. . . . . . . . . . . . . . . . .Fig 3.7 Stabilising elements built in reinforced concrete

or masonry 28. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Fig. 3.8 Floor deck bracing systems 28. . . . . . . . . . . . . . . . . . . . .Fig 3.9 Action of lateral force resisting systems (from Ref 5.2) 29Fig. 3.10 Frame example 30. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Fig 3.11 Relationship between mass/unit area and span 31. . .Fig 3.12 Relationship between cost/unit area and span 31. . . .Fig. 3.13 Beams for economic fabrication 32. . . . . . . . . . . . . . . .Fig 4.1 Steelwork Cost Components for Warehouses 35. . . . .Fig 4.2 Configuration of framing systems for a

factory building 35. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Fig 4.3 Details of bolted portal frame 36. . . . . . . . . . . . . . . . . . . .Fig 4.4 Details for welded portal frame

(with bolted rafter splice for field erection) 37. . . . . . . . . . . . .Fig 4.5 Transportation limitations for portal frames 37. . . . . . . . .Fig 4.6 Bracing panels 38. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Fig 4.7 Bracing for long buildings 38. . . . . . . . . . . . . . . . . . . . . . . .Fig 4.8 Details for rod bracing 39. . . . . . . . . . . . . . . . . . . . . . . . . . .Fig 4.9 Details for angle bracing 39. . . . . . . . . . . . . . . . . . . . . . . . .

Fig 4.10 Details for tubular bracing 40. . . . . . . . . . . . . . . . . . . . . .Fig 4.11 Types of supporting columns 40. . . . . . . . . . . . . . . . . . . .Fig 4.12 Crane runway brackets 40. . . . . . . . . . . . . . . . . . . . . . . .Fig 4.13 Commonly used sections for crane runway girders

and their relative fabrication cost 41. . . . . . . . . . . . . . . . . . . . .Fig 4.14 Standard purlin cleats 42. . . . . . . . . . . . . . . . . . . . . . . . . .Fig 4.15 Zed section purlins with lap 42. . . . . . . . . . . . . . . . . . . . .Fig 4.16 C section purlins with butt joint 42. . . . . . . . . . . . . . . . . .Fig 4.17 Method of fixing fly bracing to standard punching 43. .Fig 4.18 Three--pinned portal truss 43. . . . . . . . . . . . . . . . . . . . . .Fig 4.19 The basic square grid double layered space frame 44Fig 5.1 Framing system for low--rise commercial building 46. . .Fig 5.2 Stability by masonry 47. . . . . . . . . . . . . . . . . . . . . . . . . . . .Fig 5.3 Stability by concrete panels 47. . . . . . . . . . . . . . . . . . . . . .Fig 5.4 -- Optimum steel framing systems for buildings

of various heights 49. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Fig 5.5. Field welded connection details 50. . . . . . . . . . . . . . . . . .Fig 5.6 Shop welded connection details 50. . . . . . . . . . . . . . . . . .Fig 5.7 Forms of bracing 51. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Fig 5.8 Bracing should connect to column 51. . . . . . . . . . . . . . . .Fig 5.9 Service core 52. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Fig 5.10 Service core at end of building 52. . . . . . . . . . . . . . . . . .Fig 5.11 Floor support members 53. . . . . . . . . . . . . . . . . . . . . . . .Fig 5.12 Composite floor beam system 54. . . . . . . . . . . . . . . . . . .Fig 5.13 Welded stud shear connector 54. . . . . . . . . . . . . . . . . . .Fig. 5.14 Profiles of composite galvanized steel decking 54. . .Fig 5.15 Composite columns incorporating a

steel erection column 55. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Fig 5.16 Composite column comprising a concrete--filled

tubular section 56. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Fig 5.17 Cost Components for a Multi--Storey Building 56. . . . .

11

Fig 7.1 Types of fillet welds 66. . . . . . . . . . . . . . . . . . . . . . . . . . . . .Fig 7.2 Types of butt welds 66. . . . . . . . . . . . . . . . . . . . . . . . . . . . .Fig 7.3 Weld cost graph 67. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Fig 7.4 Welded beam--to--column moment connection 68. . . . .Fig 7.5 Stiffened web plate girder with web penetration 68. . . .Fig 7.6 Some common detailing faults resulting in

poor accessibility for welding 72. . . . . . . . . . . . . . . . . . . . . . . .Fig 7.7 Use of bending to reduce welding and

give clean corners 72. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Fig 7.8 Beam flange with many different plate thicknesses

-- avoid when steel mass saved is less than100 times mass of weld metal required 73. . . . . . . . . . . . . . .

Fig 7.9 Exterior column/spandrel sub--assemblies forSears Tower, Chicago 73. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Fig 7.10 Angle seat detail -- (a) preferable to (b) 74. . . . . . . . . .Fig 7.11 These joints are difficult to weld and the welds

may be of questionable quality 74. . . . . . . . . . . . . . . . . . . . . .Fig 8.1 Chain of communication 75. . . . . . . . . . . . . . . . . . . . . . . . .Fig 8.2 Plated sections 76. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Fig 8.3 Web penetrations in beams

(in descending order of cost, (d) being least costly) 77. . . .Fig 8.4 Typical castellated beam geometry 77. . . . . . . . . . . . . . .Fig 8.5 Evaluation of economics of castellated beam 78. . . . . .Fig 8.6 One--sided intermediate web stiffener 79. . . . . . . . . . . . .Fig 8.7 Stiffened and unstiffened webs in three plate girders 80Fig 8.8 Column base plate details (moment resisting or fixed) 81Fig 8.9 Typical pinned base plates

(full dimensional details can be found in Ref l) 81. . . . . . . . .Fig 8.10 Holding--down bolt details 82. . . . . . . . . . . . . . . . . . . . . .Fig 8.11 Typical holding--down bolt cage 83. . . . . . . . . . . . . . . . .Fig 8.12 Minimise number of column splices

-- 1 is preferable to 3 83. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Fig 8.13 Preferred column splice locations 83. . . . . . . . . . . . . . .

Fig 8.14 Economic details for built--up columnsin ascending order of fabrication cost 85. . . . . . . . . . . . . . . . .

Fig 8.15 Welded corner details for box columns 85. . . . . . . . . . .Fig 8.16 Connections to box columns 86. . . . . . . . . . . . . . . . . . . .Fig 8.17 Spreading of columns to allow for weld shrinkage 86.Fig 8.18 Equivalent truss detailing 87. . . . . . . . . . . . . . . . . . . . . . .Fig 8.19 Single angle welded truss 87. . . . . . . . . . . . . . . . . . . . . .Fig 8.20 Split tee welded truss 87. . . . . . . . . . . . . . . . . . . . . . . . . .Fig 8.21 Use of universal sections in welded trusses 88. . . . . . .Fig 8.22 Use of rectangular hollow sections

in welded trusses 89. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Fig 8.23 Types of open web joist 89. . . . . . . . . . . . . . . . . . . . . . . .Fig 8.24 End plate details 89. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Fig 8.25 Clearance at apex joint 90. . . . . . . . . . . . . . . . . . . . . . . .Fig 8.26 Termination of haunch 91. . . . . . . . . . . . . . . . . . . . . . . . .Fig 8.27 Attachment of purlins and girts 91. . . . . . . . . . . . . . . . . .Fig 8.28 Precambering details of a rigid frame 91. . . . . . . . . . . .Fig 8.29 Typical beam details for fabrication economy 92. . . . .Fig 8.30 Typical floor beam layout 92. . . . . . . . . . . . . . . . . . . . . . .Fig 8.31 Angle seat connection 94. . . . . . . . . . . . . . . . . . . . . . . . .Fig 8.32 Flexible end plate connection 94. . . . . . . . . . . . . . . . . . .Fig 8.33 Angle cleat connection 95. . . . . . . . . . . . . . . . . . . . . . . . .Fig 8.34 Web side plate connection 96. . . . . . . . . . . . . . . . . . . . .Fig 8.35 Bearing pad connection 96. . . . . . . . . . . . . . . . . . . . . . . .Fig 8.36 Welded moment connection 97. . . . . . . . . . . . . . . . . . . .Fig 8.37 Moment end plate connection 98. . . . . . . . . . . . . . . . . . .Fig 8.38 Welded splice connection 99. . . . . . . . . . . . . . . . . . . . . .Fig 8.39 Bolted splice connection 100. . . . . . . . . . . . . . . . . . . . . . .Fig 8.40 Stiffener connections 101. . . . . . . . . . . . . . . . . . . . . . . . . .Fig 8.41 Bracing connections 102. . . . . . . . . . . . . . . . . . . . . . . . . . .Fig 8.42 Connections to concrete cores 102. . . . . . . . . . . . . . . . . .

1

1. Preliminary Considerations

1.1. Introduction

It is generally accepted that the objective of engineering design is theachievement of an acceptable probability that the structure being designedwill retain its fitness for purpose during its planned lifetime. It is also ofutmost importance that the initial costs plus the maintenance costs of thecompleted structure be within the limits provided by the Client.

For the design to be successful in the sense just outlined, the designershould search for design alternatives which consider strength andserviceability on the one hand, and economic feasibility on the other. Inother words, out of a number of alternative structural solutions whichcomply with accepted design criteria for strength and serviceability, thedesigner should select the alternative likely to be the lowest overall cost. Todo this successfully, the designer should develop an appreciation of thebasic sources of expenditure in building construction and their effect on theoverall cost of construction.

In practice, the design problem is an optimisation problem. The solution toany optimisation problem involves having some means of judging theoverall merit of alternatives. With regard to a building, the measure ofoverall merit, usually provided by the Client, will involve one or more of thefollowing criteria:

(a) Functional requirements

(b) Strength and serviceability

(c) Aesthetic satisfaction

(d) Economy in relation to capital and maintenance costs.

This publication deals almost entirely with item (d) above.

In the preliminary and final design, the designer often deals primarily withmember design and consequently tends to consider theminimisation of themass of the structure as a guiding criterion towards achieving minimum

cost. That is, the designer substitutes the more straightforward criterion ofmass minimisation for the more involved criterion of minimum cost.

In regard to steel structures, aminimummass solution does not nec-essarily result in aminimum cost solution. Connection detailing andthe resulting cost of fabrication and erection aremore often themajorinfluences affecting overall cost. Undue preoccupationwith themini-misation of themass of a steel structure can lead to serious errors ofjudgement.

This publication is intended to highlight the manner in which a number offactors affect the cost of fabrication and erection. It will also highlight theinfluence these costs have on the total final cost of a steel structure.

1.2. Factors influencing Framing Cost

Fabricated steel has been traditionally costed on a per tonne basis.Consequently in discussing the cost of fabricated steel the question oftenraised relates to howmuch is the cost per tonne of fabricated steel. Such aquestion usually ignores the fact that a large number of factors have asignificant influence on the final cost of fabricated steel.

A more rationalised approach to the costing of fabricated steel is based ona cost per metre for sections and cost per square metre for platesdepending on the size of the member. Fabrication costs for connectionsand erection costs, etc can then be added on a component by componentbasis (Ref 3.).

In the design, detailing, fabrication and erection of a steel structure, thefollowing factors influence the cost of the framing:

(a) Selection of the framing system

(b) Design of the individual members(c) Design and detailing of the connections

(d) Fabrication processes used

2

(e) Erection techniques used

(f) Specification for fabrication and erection

(g) Other items such as corrosion protection, fire protection, etc.

The selection of the most efficient framing system is fundamental toachieving an economical framing solution and aspects relating to this itemare discussed in Sections 3. 4. and 5.

Efficient member design remains an important cost factor tempered by thecomments made in Clause 1.1. Detailed consideration of this item doesnot fall within the scope of this publication. One point that does deservemention however is the avoidance of the individual design of every beamand column in an attempt to achieve least mass. The aim should be togroup similar members (e.g. similar main beams in a floor grid) and adoptthe one size for all members of the group. An experienced designer willoptimise the design by being aware that if toomuch grouping is done, therewill be material wastage. However, if little grouping is done, then there is agreat waste of time on the part of the draftsperson and the erector.

Economic fabrication and erection are significantly affected by economicalconnection details. This publication is very concerned with economicdetailing of steelwork and themanner in which detailing influences the costof fabrication and erection. Sections 6. 7. and 8. deal with a variety of pointswhich need consideration.

The specification (item (f) above) is amajor influence on the cost of both thefabrication and erection since it specifies the quality of materials andworkmanship required.

Similarly, the costs of both corrosion protection and fire protection (item (g)above) are important influences on the final cost. All these items arediscussed in greater detail in Section 2.

1.3. Integrated Design

One of the obstacles to achieving maximum economy is that three of themost important activities in steel frame construction, namely structuraldesign, detailing and fabrication, are usually done in isolation from oneanother. This is partly due to specialisation in each of the disciplines and

partly because of a lack of an effective dialogue among the peopleinvolved.

As a result of this, there often occurs a total preoccupation with theanalytical phase of the design, and a complete absence of rational thinkingabout the detailing phase. Consequently, the problems that arise duringthe detailing phase are solved by complicating the detail rather than bymodifying the design concept. When the job reaches the fabrication shop,there is little alternative but to carry out whatever happens to be shown onthe drawings.

A much improved situation results when the design effort is integrated sothat the framework, its members and its connections are considered as awhole. In this way, it becomes possible to modify the structural framingconcept to allow the use of simpler and less costly connections in theinterest of overall economy.

The cost factors listed in Clause 1.2 should be considered in an integratedmanner so that interactions between the framework, its members and itsconnections are considered during the design process. In this way, oneaspect can bealtered to enable another to be improved. This enhances theoverall cost efficiency of the final structure.

Obviously, such an approach ideally requires an extensive and up--to--dateknowledge of the steel fabrication and erection industries. Since suchknowledge is not always easily achieved, communication with fabricatorsis a useful method of establishing the optimum practical solution. Aninterchange of ideas among fabricators, erectors and designers is an idealsituation for achieving optimisation.

It should be appreciated that what constitutes ”design” and ”good (i.e.economical) design” will vary depending on whose viewpoint is beingconsidered. To the designer, an economical design is usually the lightestmember to carry the load. To the fabricator, a ”good design” means hightonnage output with minimum amount of labour. To the erector a ’good’design is one where most members are the same size and can beinterchanged without any problems.

Clearly such different viewpoints are best resolved by an integrated andinteractive approach on the part of the steelwork designer.

3

2. General Factors Affecting Economy

2.1. Steel Grades

2.1.1. STRUCTURAL STEEL

Throughout the world the least costly and most commonly used grades ofsteel for structural purposes are those generally referred to as normalstrength structural steel.

In Australia such steel is covered by AS 3678 or AS 3679 (Parts 1 & 2). Ithas a typical design yield strength of 250/300 MPa (varying above andbelow this figure depending on thickness), a tensile strength of at least410/430 MPa, a minimum elongation of 22% and a carbon equivalent of0.43/0.44 so as to assure good weldability.

AS 3678 and AS 3679 (Parts 1 & 2) are omnibus standards covering afamily of structural steel grades including variants of the main gradeshaving superior low temperature toughness.

Plates, rolled sections, welded sections and bars are all produced to thesestandards, although not every product is available in every grade. This isexplained more fully in Table 2.1.

2.1.2. WEATHERING STEEL

AS 3678 and AS 3679 (Parts 1 & 2) also deal with so--called ’weatheringsteel’. Weathering Steel contains alloying elements which cause it toweather to a uniform patina after which no further corrosion takes place. Bynature of the chemical composition the steel is high strength (Grade 350)steel. However in Australia it is available in only a limited number ofproducts----see Table 2.1.

2.1.3. HOLLOW SECTIONS

InAustralia structural hollowsections are produced to the product standardAS 1163. This standard covers a number of cold--formed (C) grades.Rectangular hollow sections are available inGradeC350andGradeC450.Circular hollow sections (CHS) are available in Grade C250 and GradeC350.

4

TABLE 2.1 -- AVAILABILITY OF PRODUCTS BY GRADE (check cur-rency of information with steel suppliers)

Steel Grade Plates (orFloor plates)

RolledSections

WeldedSections

StructuralHollow Sections

Grade AS3678

AS3679.1

AS3679.2

AS1163

200 Y -- --

250 Y Y (1) --

250L0 -- X --

250L15 + -- --

300 + Y (2) Y

300L15 + -- +

350 Y + --

350L0 -- X --

350L15 + -- --

400 + -- Y

400L15 X -- +

WR350/1 + -- --

WR350/1 L0 + -- --

Grade C250 Y

C350 Y

C450 Y

Quenched &Tempered Structural

Steel

AS 3597

60 Y

70 Y

80 Y

Y = regular grade commonly produced, readily available from stockists+ = regular grade not commonly produced, availability subject to time limitations and order sizeX = non--regular grade, availability subject to time limitations and order size

-- = not manufacturedNotes: 1. Applies to TFB, TFC and small angles.

2. Sections not considered in Note 1 above.

2.1.4. QUENCHED AND TEMPERED STEEL

Steel plates are produced in Australia in very high strength heat--treatedgrades known as ’quenched and tempered steel’. These steel plates areuseful in special applications wheremass reduction is important (eg. cranebooms) or where their high wear resistance is needed (e.g. dump truckbodies).

Australian Standard AS 3597 covers these steel plates for structural steelapplications and for use in pressure vessels.

2.1.5. CHOICE OF STEEL GRADE

Table 2.1 lists the availability of various products by steel grade. Theindicative relative cost of grades is shown in Table 2.2. For most structuresthe greatest economy will be achieved by the selection of the least costlyand most readily available steel, i.e. Grade 300.

In large structures with longer lead times the use of higher grades will oftenbe worth considering at least for parts of the frame.

Heavy plate members such as bridge girders are one instance wherehigher grades may prove economical. Other applications include:

(a) Multi--storey structures, particularly with composite steel beams;also in maintaining the same column size down a building byvarying steel grades;

(b) Trusses and lattice girders.

Grade 350 steel costs around 10% more than Grade 300, and generallyabout 5%more to fabricate. To offset these cost extras, it provides greateryield strength but no increase in stiffness.

In some frames, significant reduction in steel mass may overcome theincrease in material cost and fabrication cost by the use of higher grades.Each individual frame must be assessed on its merits, but there areundoubtedly applications where the use of higher grades is economical.

While the information presented in Table 2.1 is indicative of the generalsituation, it must be remembered that the steel suppliers are always willing

5

to discuss special cases where, for example, the economics of a highstrength steel has been considered by the designer and the sectionsrequired are not normally manufactured in that grade. For a projectrequiring large tonnage of specific sections, it may be possible to negotiatea special order with the supplier, provided that an arrangement has beenagreed at an early enough phase in the design.

Conversely, on average projects the designer should always be careful tokeep within the range of readily available products so as to ensure that noproblems of steel procurement occur at the fabrication stage.

TABLE 2.2 INDICATIVE COST RATIOS FOR DIFFERENT GRADESOF STRUCTURAL STEEL (per tonne, supply only)

Grade Plates RolledSections

WeldedSections

AS 3678, AS3679.1 & AS3679.2

Grade 250 100 100 --

250L0 -- 105 --

250L15 105 105 --

300 100 110 130

300L15 105 -- 140

350 110 120 --

350L0 -- 130 --

350L15 120 -- --

400 115 -- 150

400L15 120 -- 155

WR350/1 125 -- --

WR350/1 L0 135 -- --

AS 1163

Grade C250 130

C350 130

C450 130

AS 3597 Quenched& Tempered Steel

Grade 60 150

70 160

80 160

6

2.2. Economy in use of Material

As well as having a knowledge of the factors affecting the choice of steelgrade, the designer should also be aware of how design decisions canavoid unnecessarymaterial cost or wastage. This will involve a study of thefactors discussed below.

2.2.1. STEEL PRICING

Mill prices are expressed in terms of a base price and various extras. Thebase price relates to the type ofmill product such as plate or sections, whileextras relate to specifics of the particular product or section.

The most common extras for structural quality steel include the size ordesignation, standard or non--standard lengths, quantity extras ordiscounts related to the total mass of individual order items, and the gradeextras which apply to the quality specification for the material chosen.

Quality extras for structural steel relate to the material specifications andreflect the costs of alloying elements, of tighter controls on such elementsas carbon, manganese, phosphorus and silicon, and of tighter controls onmanufacturing techniques to meet the specified chemical and mechanicalproperties. The cost of additional tests and greater frequency of testing,necessary for increased stringency of yield strength and notch ductility, arealso reflected in increased quality and testing extras.

Designers should recognise that the more exotic the requirements of thesteel specification, the greater is the probability that other costs associatedwith its use, ranging fromprocurement through all stages of fabrication, willalso be increased. Unnecessary demands by specifiers for mill heatcertificates for standard sections to be used on routine projects is anotherexample of unnecessary costs added onto projects.

The foregoing relates to purchases made direct from the steel mill, but inAustralia most fabricators obtain their steel through steel merchants.These steel merchants aim to carry comprehensive stocks and are thus

able to offer prompter delivery than would be available through the normalsteelmaker’s rolling programs.

On the other hand, the larger fabricators are able to meet the millrequirements for direct purchase and prefer to procure their material thisway. In recent years there has been a trend among some larger fabricatorsto limit their mill purchases to main material, and to obtain low--volumeancillary material from steel merchants as and when required.

In Australia there is not much difference in the cost of steel procured in oneway or the other if all expenses are truly accounted.

The argument about using, wherever possible, the preferred grades andsizes applies equally to steel obtained from a steel merchant, becausethese firms naturally tend to concentrate their stockholding on popular,fast--moving, items.

2.2.2. PLATES

In Australia there is a rationalised series of ’preferred’ plate thicknesses aslisted in Table 2.3.

For practically all structures the designer should operate within thisstandard range. Non--preferred thicknesses incur cost premiums andextended delivery times, and should only be considered on major projectswhere the overall saving in using a special thickness is greater than thedirect and indirect cost penalties.

Similarly there are preferred lengths and widths of plates which should beborne in mind. Major plate elements should be dimensioned as far aspossible so that they can be cut from standard plates with a minimum ofscrap. Smaller plate details such as brackets and gussets should beconsidered in the same way, especially when there is a large number ofthem. The most common sizes for plates up to 25mm thick are 1.8m x 6m,2.4m x 6m, 2.4m x 9m, and 3m x 9m.

(Note: Small plate components may be substituted by flat bars which areconsidered as Sections.)

7

TABLE 2.3 PREFERRED STEEL PLATE THICKNESSES (in mm)

3 25 70

4 28 80

5 32 90

6 36 100

8 40 110

10 45 120

12 50 140

16 55 160

20 60 180

2.2.3. SECTIONS

Australia produces a range of welded products, universal sections,channels, angles, and hollow sections which provides the designer with areasonable choice without the proliferation which can lead to problems ofavailability.

The lowest weight in each nominal size of universal section is the moststructurally efficient and they account for over two--thirds of all UB sales.The designer should thereforemakeevery endeavour to keep to the lowestweights, although this will not always be possible.

Very long lengths of sections become difficult to keep straight and tohandle, and the mills impose a price extra for them. It should be especiallynoted that although universal sections are listed as being available up to18m long (and up to 27m by inquiry), the usual maximum length found instock is around 15m.

The available lengths of structural hollow sections are usually restricted to6.5m (circulars) or 12m (rectangulars and squares).

2.2.4. SCRAP AND WASTE

The real cost of material is affected by the quantity of scrap and waste, anddesigners should be receptive to suggestions for minimising andcontrolling the generation of waste. This may include greaterstandardisation of structural sizes, or of plate widths and thicknesses, inorder to take advantage of size and quantity discounts. Itmight also includeamore liberal approach to the splicing of beamsor other structural sectionsusing standard lengths.

Random splicing, which involves welded splices anywhere within thelength of a rolled structural member, can be particularly effective whenmaterial is sawn to length and fabricated on a conveyorised productionline.Whencarefully controlled, it can dramatically reduce theaccumulationof shorts and thus reduce the total cost. The only real restriction to randomsplicing applies to its use for beams subject to severe dynamic loads. Ofcourse the savings in scrap have to be balanced against the welding costs,and the designer should be receptive to this technique where it isappropriate.

2.3. Fabrication

2.3.1. GENERAL

Fabrication costs are a function of complexity and are influenced by:

(a) Size of the component

(b) Size and type of sections involved

(c) Amount of stiffening and reinforcing required

(d) Amount of repetition

(e) Shop and field details

(f) Space requirements in the shop, and

(g) Facilities available for handling, lifting and moving the structuralcomponents.

Fabrication costs are sensitive to simplicity or complexity of detail, and thedegree to which production line techniques can be applied. They arecontrolled by the quality of the shop detail drawings, whichmust reflect the

8

designer’s concept for the structure, but must also permit the optimumutilisation of the fabricator’s facilities and equipment. Shop drawingpreparation should be guided by the basic principle that they must providefor economy of fabrication and for economy of erection.

Shop operations basically involve cutting material to size, hole--making formechanical fasteners, and assembling and joining. Other operationsinclude handling, cleaning and corrosion protection. All shop operationsrequire facilities for lifting and for moving or conveying the structural steel.

Cutting operations include shearing, sawing and flame cutting;hole--making operations include punching and drilling; assemblyoperations include welding and bolting. Increased use of computernumerically controlled (CNC) fabrication processes is changing theeconomics of steel fabrication. Cutting, drilling andwelding operations cannow be undertaken by the CNC fabrication process. Information fromcomputer drafted shop drawings can be fed directly into CNC fabricationequipment to further improve operational efficiency. Some fabricators arenow bar coding steelwork to facilitate control and monitoring of projects.

Generally welding is the preferred method for shop assembly, with boltingfor field assembly. There are, however, some fabricatorswith sophisticatedhole--making equipment, who prefer shop bolting to shop welding forstandard connections. Some steel merchants also provide basic cuttingand drilling services to the steel fabricators.

2.3.2. BEAM AND COLUMN FABRICATION

A large part of structural steel fabrication consists of beam and columnwork. It embraces framing members consisting of standard rolled shapesconnected by shear or moment connections, and also includes highlyirregular framing members with custom designed built--up sections andcomplex connections designed for combinations of shear, moment anddirect tension.

Simple beam and column fabrication lends itself to production linemethods, inwhich themembers are transported ona series of conveyors tosaws which cut the material to length, and to hole--making equipmentwhich provide holes in either the web or flange or both.

Any additional requirements, such as the attachment of cleats or brackets,are off--line operations. It is important therefore that connections and otherdetails be selected so as to provide themaximumnumber of members withonly cutting and holing. Otherwise the economy of using CNC equipmentand the conveyorised beam--line system will be less apparent (see Figs3.13 and 8.29).

2.3.3. GIRDER AND TRUSS FABRICATION

Fabrication of plate girders and trusses differs frombeamand columnworkin that it involves assembly in the shop, and calls for adequate space andhandling facilities. Both girders and trusses require special fit--up jigs forassembly and welding, and the availability of heavy lifting equipment.

Just as with beam and column work, however, the key to productivity andeconomical fabrication is the use of simple standard details for stiffeners,splices, gussets, etc.

For plate girders all details should be designed for automatic welding,allowing adequate clearances for the welding machines to pass and fortermination of welds at the ends of web stiffeners. Maintaining constantwidth flanges within a shop fabricated length of girder permits splicing ofmultiple width plate and subsequent stripping to finished width. This willreduce weld set--up time, eliminate weld starts and stops, and require onlyone set of run--on and run--off tabs. Reductions of flange widths, webdepths and plate thicknesses purely to reducemass should be consideredvery carefully as they can significantly increase fabrication costs.

Control of distortion in plate girder fabrication is amajor problem,which canbe helped by designwhichminimises the amount of welding and avoids theuse of significantly non--symmetrical sections. It is false economy to designfor minimum web thickness only to require web stiffeners, therebyincreasing the amount of welding and distortion; or to use very light topflanges in composite girders only to compound the problem of cambercontrol. See also Clause 8.2.5.

Trusses can be designed in a large variety of configurations which dependon the truss span, depth and loads to be carried. Therefore, it is impossibleto make general statements regarding the most economical design forfabrication, other than to stress again the importance of simplicity of detail.Designers should avoid situations that can cause weld restraint andproblems resulting fromweld induced distortion. As far as possible trusses

9

in the one project should have the same configuration so that they can allbe fabricated from the one jig.

In trusswork, the correct selection of chordmembers can often remove theneed to turn the truss over during the fabrication (see Clause 8.4). This willenable the fabricator to complete the entire welding on the trusscomponent without further handling.

2.3.4. SUMMARY FOR ECONOMIC FABRICATION

The key to economic fabrication is the use of standards at all stages. Thisincludes standard procedures, standard schedules, standard drawings,and above all standard connections and details. Non--standard details areusually handled as ’special job standards’; however, the net effect of anyspecials is to slow production with some loss of fabrication economy.

In the selection of connections the designer should observe the followingprinciples:

(a) Select members and connections to provide a maximum of repe-tition throughout a structure. This provides the fabricator with theopportunity to make up jigs and fixtures to speed up the fabrica-tion process.

(b) As far as possible, select connections so that the assembly offitments on a member can be carried out in one position. This willreduce the number of handling or rotating operations during fab-rication.

(c) Keep the number of components in a connection to a minimum.

(d) Select connections so that assembly of components occurs onthe least number of members.

(e) As far as possible use connections that are standard in the in-dustry (see Standardized Structural Connections, Ref 1).

(f) Ensure a minimum standard of documentation in line with AISC’spublication: ” A Guide to the Requirements for Engineering Draw-ings of Structural Steelwork” (Ref 2.12).

(g) Most importantly, keep an open mind on the selection of mem-bers and connections. Before finally committing a design to thedetail design phase, communicate with the industry and try to

determine the best solution to optimise the use of material andlabour in the fabrication shop. This industry communication canoften be facilitated through the services of AISC.

2.4. Erection

2.4.1. GENERAL CONSIDERATIONS

The rate of erection of steel in a structure is controlled by five main factors:

(a) Connection simplicity

(b) Number of members

(c) Number of bolts and/or amount of field welding

(d) Size and efficiency of erection crew, and the equipment at theirdisposal

(e) Timely supply of steel.

It is interesting to note that of these factors, the first three are under thecontrol of the designer.

Connections should be simple, and of such a type that the allowabletolerances (in member size and shape, detailing and fabrication) can beaccommodated during the placing of the members.

The number of members should be kept to a practical minimum, and soshould the number of bolts or amount of field welding. There should besufficient access for welding or for tightening bolts using power wrenches.

Bolted connections should be used wherever possible, and field weldingkept to a minimum. Connection plates should be shop welded to onemember, rather than field bolted to both, unless other considerationsgovern.

Every endeavour should be made to standardise as far as possible(member sizes, bolt sizes, type of connection, gauge lines, memberspacing, etc.), and careful consideration should be given to how amemberis to be installed with minimum interference by other members, gussetplates, etc. (see Ref 1).

With an increasing awareness of the importance of employee safety in thework place, erection methods are changing. Designers and erectors havea duty of care and should consider safe erection methods. The use of

10

equipment such as cherry pickers is becoming more common duringerection. Designers need to include anchorage points for safety lines andharnesses for riggers. These issues are resulting in steelwork beingerected on the ground and then craned up to final position inmany projectsto reduce the amount of work done at great heights. This may requirealternative design and detail methods, and utilisation of additional shortterm cranage but provides a safer work site. A safer work site will lead tofaster and more economical erection.

2.4.2. HANDLING AND TRANSPORT

As a general rule it is more economical to erect fewer large pieces thanmany small pieces, due to the number of lifts involved and the number ofjoints tomake.Generally thismeans fabricating larger pieces in the shop toreduce the number of pieces and field connections. On the other hand,transportation constraintsmay limit the size of a piece for delivery to the siteand require additional field splices. For example, with long flexible trusses,the transportation length may have to be curtailed to avoid damage duringtransfer to site or to avoid obstructions along the way.

Large sub--assemblies may require to be transported using specialvehicles attended by police escort, and this may add greatly to the finalprice of the structure. However, projects outside capital cities could usethis approach as it minimises the size of the site crew required to bemobilised on a remote or semi--remote site. With greater availability oflarger mobile cranes and trucks, the balance between transport costs andsite costs is changing. Where projects require large site crews, minimisingtime spent on site is essential to economical erection. The erection or trialerection of large components in a fabricator’s yard before delivery to site isgood practice and a cost savings exercise. Trial erection guards againstfabrication errors being discovered on site which may prove expensive torectify.

Generally, transport costs are high, and therefore considerable economycan be achieved if vehicles travel fully laden. The dimensions of a typicalload of structural steelwork which requires no special escort are in theorder of 15m long x 3m wide x 2m high. It is important that like pieces areloaded together to optimise truck capacity, but also that the components bedelivered to site in the order required by the erection sequence (i.e.columns followed by beams from ground upwards). This will save double

handling on site and also reduce the cost of site storage and possibledamage.The virtue of designing for repetitive components has already beenstressed. The gains can be partly lost on site if interchangeable parts aregiven individual mark numbers. This will require the erector to search for aparticular number mark on a member when any one of a considerablenumber of members would fit. After completing a design it is worth lookingat marking plans with this idea in mind.Indicative transportation costs are given in Table 2.4. Costs include theloading of steelwork onto and off the truck.

TABLE 2.4 TRANSPORTATION COSTS

Transport Fabrication Shop toSite (see Note)

Section Mass (kg/m) $/member

60.5 and less 15

60.6 to 160 56

160.1 to 455 225Note: Allow for twice the cost of transportation if the surface treatment

is applied at premises other than the fabrication shop.

2.4.3. CONNECTIONS

It is in the final fixing of members that the greatest scope for erectioneconomy lies.Connections selected to permit flexibility in fit up should beofprime concern to designers. The use of one type of bolt and one boltingprocedure throughout a structure will allow the use of aminimum variety oftools on site and provide for speedy erection sequence (see Section 6).Similarly where site welded connections are required, cleats should beincorporated to allow mating members to be held together in place foractual welding.Angle seat, angle cleat and web side plate connections (see Clause 8.6.2)provide considerable flexibility in fit--up, and are preferred in braced framesfrom a purely erection viewpoint. The flexible end plate connection is notquite so easy to erect, although its selection may be decided by otherconsiderations.

11

In rigid frames, the following should be taken into consideration for thedesign of bolted connections:

(a) The end plate depth should be kept to a minimum to reduce thetendency to jam during installation (Fig. 2.1).

(b) The tolerance between the face of the end plate and the face ofthe column should either be tightly controlled so that the buildingplumbs itself automatically, or allowance should be made forshimming in order to plumb the building. Shimming, however, canbe expensive.

(c) In end plate connections for portal frames careful considerationshould be given to access for installing and tensioning bolts, (seeTable 8.1).

If welded connections are preferred, the following should be taken intoconsideration:

(a) Welded connections are normally erected using a bolted erectionconnection. The same criteria should apply to the design of theseconnections as described above.

(b) Substantial erection clearance between the end of the girder andcolumn face should be provided where permitted by the design ofthe connection.

(c) Field welding should be kept to a minimum and overhead weld-ing should be avoided.

(d) Attention should be paid to access for welding and welding in-spection.

(e) Consideration should be given to plumbing the building.

Themost significant time delays in the erection of a girder can be expectedto occur when it is installed with the end connection against a column web.The girder can normally only be manoeuvred in a vertical plane andfrequently jams. Gusset plates, stiffeners, and other members tend tointerfere with its installation. Access for bolting is usually difficult andsometimes impossible. Every effort should be made to get the connectionoutside the flanges of the column, or at least as far out from the web aspossible. This is especially important when the column section is compact.Consideration should always be given to excluding direct girder/web

connections even if it involves increasing columnweight, and/or fabricationcosts (see Fig. 2.2).

Fig. 2.1 Deep end plates can cause jamming

Fig. 2.2 Two ways of avoiding the problem of access to columnweb connections

2.4.4. FIELD BOLTING

In projects with a predominance of large connections, threads may beexcluded from the shear plane for bearing type connections as this will helpto reduce the number of bolts. However with Australia’s ISO metriclong--thread bolts, care should be taken that the long ’stick--through’ thatoccurs does not cause fouling or access problems. In projects with smallconnections the saving in number of bolts is not so evident, and it is more

12

economic to design for threads included in the shear plane. This thenmeans that bolt lengths can be selected so as to avoid excessivestick--through. However the two systems (threads--in, threads--out) shouldnot be mixed on the one job (see Ref 6.1).

Generally, the smaller the bolt the easier it is to install. Bolt diametersshould therefore be kept small if this canbe donewithout compromising theobjective of keeping the number of bolts to a minimum. M12 bolts arenormally adequate for stairs and girts, while M20 bolts are the maximumsize which should be considered if access for tensioning is poor; otherwiseM24 bolts are acceptable.

Bolts should be specified as ’snug--tight’ unless there are compellingreasonswhy fully tensionedbolts are necessary. The cost of full tensioning,including associated inspection, is very high and can double the cost ofeach installed bolt. Access for wrenches is also less critical where onlysnug tightening is to be carried out. Care should be exercised, however,where a project is designed to overseas codes because some of theserequire high strength structural bolts to be always fully tensioned.

It is preferable that only one bolting category (see Section 6) be used onanyone structure.Whenadeparture from thegeneral category (e.g. to fullytensioned bolts, to threads excluded from shear plane, etc.) isunavoidable, this should be highlighted on erection and detail drawings toreduce the possibility of the requirement being overlooked by erectioncrews.

More information on structural bolting is given in Section 6 and Ref 6.1.

2.4.5. FIELD WELDING

Where site welding is used for connections the total amount of welding onthe job should be sufficient to justify the cost of bringing and setting upwelding equipment on the site.

Access for welding is also important, and it should be remembered that awelder generally requires a substantial and carefully placed workingplatform.

Otherwise the normal rules for economic welding apply. Fillet welds arepreferred to butt welds, and down--hand welding to any other position. In

most structural work difficult out--of--position welds such as overhead arevery slow and costly. See also Section 7.

2.4.6. BRACING

Bracing is usually difficult and time consuming to install. To reduce erectiontime the number of braced bays should be kept to a minimum, i.e. fewerbraced bays with heavier bracing is preferred.

Wherever possible, wall bracing should be connected to columns ratherthan beams. This allows bracing to be installed before the beamabove is inposition, hence reducing any interference this beam may cause duringerection. Connecting the brace to the column at its lower end eliminatesinterference to the floor system resulting from a gusset plate on the topflange of a beam.

Connecting wall bracing to the column also usually results in lowerfabrication costs.

2.5. Surface Treatment


With the development in recent years of a large variety of surface treatmentmethods, the designer may experience considerable difficulty in selectingthe optimum system for a particular application.

Furthermore, it is often not fully realised that the cost of a sophisticatedmulti--coat treatment system can easily be more than the cost of the rawsteel itself. Thus care is needed to avoid unnecessary, and sometimesunexpected, surface treatment costs.

These costs are a function of surface area which can vary with both: thetype of section used and the class of construction.

For example, a structural hollow section has typically only one--half totwo--thirds of the surface area of an ”open” structural section (UB, UC...) ofequivalent capacity. For this reason, hollow sections arewell worth bearingin mind for applications requiring any significant amount of surfacetreatment.

Heavy steel construction such as for power stations usually averages outwith comparatively less surface area (despite the higher tonnage) than a

13

typical factory or warehouse where light trusswork may have a muchgreater surface area (despite the lower tonnage). Obviously treatmentcosts on a per square metre basis will vary widely depending on the actualsurface area to be treated.

2.5.2. STEEL PERFORMANCE

Bare steel will corrode only in the presence of both oxygen and moisture.Corrosion will be accelerated if traces of pollutants such as sulphur dioxideor chlorides are present -- the so--called ’aggressive environments’.

Steel inside a building is rarely a corrosion risk except in the occasionalcasewhere the building houses an aggressive atmosphere as a result of itspurpose, e.g. a fertiliser factory. It follows therefore that steel needs nocorrosion protection whatsoever in most interior applications such asmulti--storey buildings where the steel framing is eventually concealed.

Where the steelwork remains exposed to view as in a factory or warehousethe same negligible risk applies but in these instances the owner mayrequire a surface finish for a more attractive appearance. The designershould distinguish between treatment specified to achieve protection fromcorrosion and that specified merely to provide decoration. In practice, ofcourse, any surface finish will attempt to do both.

Detailed advice on the classification of environments and the selection ofappropriate surface treatment systems is contained in AS 2312 ’Guide tothe protection of iron and steel against exterior atmospheric corrosion’ (seeSection 10).

2.5.3. SURFACE PREPARATION

An important part of any steel treatment system is the preliminary surfacepreparation. This can range fromsimple degreasing andbrushing, to costlychemical or mechanical descaling.

The surface preparation should be matched to the applied finish.Expensive paint systems will not last if applied to only partially prepared(e.g. wire--brushed) surfaces. Conversely it is a waste of money applying alow--cost porous alkyd primer to a descaled ’white metal’ surface.

Various methods of surface preparation are covered by AS 1627 ’Metalfinishing -- preparation and pretreatment of surfaces’ (see Section 10), andadvice on their selection is contained in AS 2312 (see Section 10).Themost commonly usedmethods in Australia are wire brushing (suitablefor low cost paints) and abrasive blasting to Class 2--1/2 of AS 1627 Part 4(needed for high performance paint systems). Acid descaling (’pickling’) isencounteredmainly as part of the hot--dip galvanizing process (seeClause2.5.5).An idea of the costs of various methods of surface preparation is given inTable 2.5.

TABLE 2.5 SURFACE TREATMENT COSTS

Section

Mass

Paint Types Hot-- dip

Galvanize

ROZP ALKYDGLOSS

ZnSi MIO

(kg/m) $/sq m $/sq m $/sq m $/sq m $/sq m

60.5 and less 6 6 18 9 17

60.6 to 160 5 5 16 8 23

160.1 to 455 4 4 14 7 33

Notes: 1. Red Oxide Zinc Phosphate (ROZP) 50 µm paint thicknessincluding wire brush.2. Inorganic Zinc Silicate (ZnSi) 75 µmpaint thickness includingClass 2--1/2 blast cleaning.3. Alkyd Gloss 40 µm.4. MIO 100 µm.5. For double--dip galvanizing, add 30 % to the above rates.6. For lengths greater than 12 m, check with local galvanizers.7. Rates include cost of finish painting.

2.5.4. PAINT SYSTEMS

There is a very large selection of paint systems available for structural steel-- too many to be discussed within the scope of this publication. However,

14

excellent guidance on the performance and capabilities of various paintformulations is given in AS 2312.

Probably the most commonly used paint is ’red oxide zinc phosphateprimer’, often referred to as ROZP, which is applied over a wire brushedpreparation. Paints of this type provide an economic base for possiblefurther decorative coats of conventional oil paint. However beingpermeable, ROZP cannot be expected to last if left in the open for morethan normal construction periods.

Another regularly used paint is ’inorganic zinc silicate primer’ which isapplied over aClass 2--1/2 abrasive blast preparation. It forms an excellentbase for most high performance paint formulations, or gives good resultsas a single coat protection for steel in all but the most aggressiveenvironments.

Paint is normally applied to steel by spraying. It is sometimes suggestedthat better coating is achieved by brush application, but there is littleevidence to support this claim. Brush application costs two to three timesas much as spraying, and cannot be used at all for some modern paints;inorganic zinc silicate is an example.

Table 2.5 includes the cost of the finish painting in the surface treatmentcosts. It should be noted that transportation cost should also beconsidered if the treatment is done at premises other than the fabricationshop. Table 2.4 gives an indication of transportation costs.

2.5.5. HOT--DIP GALVANIZING

Galvanizing is carried out by specialist firms and the process requiresprecleaning and surface preparation, usually by pickling. The cost ofgalvanizing includes these preparatory processes.

Advice on the performance of hot--dip galvanizing, either as a single coatprotection or as a base for paint systems, is contained in AS 2312.

When considering galvanizing the designer should ascertain the scope oflocal facilities, and in particular the size of the available galvanizing baths.The galvanizing bath determines how big an individual component can bedipped. (Items larger than the bath can sometimes be galvanized by’double dipping’ but at extra handling cost). Information on bath sizes inAustralia is given in ’Hot--dip Galvanizing’ (Ref 2.4).

2.5.6. DESIGN AND DETAILS FOR CORROSIONRESISTANCE

In a severe environment where steelwork is exposed to aggressiveconditions the designer can vastly enhance the corrosion resistance of thestructure by careful attention to a few simple principles. Conversely astructurewith bad details will not perform satisfactorily nomatter howmuchhas been spent on elaborate multi--coat protective systems.Fortunately, the principles of good corrosion detailing are generally muchthe same as those for economic fabrication. Connections and other detailsshould be kept as simple as possible with the minimum number ofcomponents. Depressions, pockets, ledges, narrow crevices andanywhere where water and foreign matter may lodge permanently shouldbe avoided whenever possible. In really severe situations the use of boxsections,CHSorRHSmight be considered. Several examples of good andbad practice are given in AS 2312.

2.5.7. SUMMARY CHECKLIST FOR SURFACE TREATMENT

1. The required level of surface treatment and/or corrosion protec-tion should be decided at the very earliest stage of the design, sothat all design decisions can be made with this in mind.

2. In benign atmospheres such as the interiors of most buildings, orexposed steelwork in non--polluted non--marine environments,corrosion rates are generally so low as to not require corrosionprotection. Any painting carried out would therefore be only foraesthetics.

3. Where corrosion protection is required, the extent needs to becarefully evaluated to ensure that it is appropriate to the circum-stances. Too much protection is a waste of money, as also is toolittle. Obviously professional judgement is needed.

4. The degree of surface preparation should match the surfacetreatment system to be applied (see Clause 2.5.3).

5. As painting is substantially a labour intensive process, the cur-rent trend is to replace multi--coat (3 or 4 coat) systems with oneor two coat systems. Zinc--rich paint systems are consequentlyincreasingly used, particularly on blast cleaned surfaces. In thesesystems, however, film thickness build is vital to a satisfactoryperformance.

15

6. Good design practice is essential -- e.g. avoid pockets where wa-ter and debris can lodge and accelerate coating failure, (seeClause 2.5.6).

7. Allowance should be made for easy future repainting.

8. Shop painting is always cheaper and more effective than sitepainting, but no steel can be handled, transported and erectedwithout damage to the coating from crane slings, etc. Touchingup of the base coats and the final top coat must therefore bedone on site.

9. Hot--dip galvanizing is a high performance protective systemwhich is not prone to damage during transport and handling. Insome circumstances it may cost the same as an alternative paintsystem (see Table 2.5).

10. Recent developments in the field of corrosion protection haveevolved protective systems greatly superior to those availablesome years ago. These systems are expensive but are invalu-able when appropriate, as in exposed structures in severe indus-trial or marine environments. However, this has led to waste ofmoney by the specification of such sophisticated treatments incircumstances where they are not necessary.

11. Some paint systems require special application techniques, con-trolled temperature and humidity when being applied, long dryingtimes or may have a tightly constrained time interval betweensuccessive coats. Designers should be careful of such sensitivesystems, as experience has shown that they are almost impossi-ble to apply correctly in normal construction industry conditions.

2.6. Fire Resistance


All structural material can be damaged in severe fire conditions and steel,although non--combustible and making no contribution to a fire, can haveits function impaired. For this reason, building regulations require it to beprotected, usually by a non--combustible insulation, when used for certainelements of construction in some types of building. Building regulations

prescribe statutory levels of fire resistance for structural steel members inmany types of applications.

The fire resistance level of a building element or structure is determined byconstructing a truly representative prototype of that element or structureincorporating fire protection materials, systems or coatings wherenecessary and submitting that prototype element or structure to theStandard Fire Test. The Australian Standard Fire Test is given in AS 1530Part 4 which enables a fire tested element or structure to be assigned a fireresistance level in accordance with the criteria laid down in the fire teststandard. Fire resistance ratings are expressed inminutes such as 30min,60 min, 90 min, 120 min, 180 min or 240 min.

Traditionally, building regulations have been based on the trial--and--errorconcept of the practical fire test. This is administratively convenient, buthas two main disadvantages. Firstly, until recently it has been difficult topredict from a particular test the fire performance of a similar but slightlydifferent configuration -- calling perhaps for further expensive tests.Secondly, it has been shown that the conditions of the standard fire test donot replicate the observed behaviour of actual building fires. The presentday trend is toward the development of fire engineering design ruleswhereby the engineer can design for fire performance in the same way ashe or she does for structural performance. The Australian design code AS4100 contains a comprehensive section on design for fire, and it seemslikely this approach will soon become standard procedure.

2.6.2. REGULATORY REQUIREMENTS

Australian Building Regulations require that elements of a structureachieve specified fire resistance levels (FRL). The level of fire resistancerequired for a particular application is related to the expected fire loadwithin the building (which is in turn related to type of occupancy), to thebuilding height and area, and to the fire zoning of the building locality andthe on--site positioning. It is not within the scope of this publication to repeatthe requirements of the various Building Regulations.

The fire ratings of common building elements have become wellestablished by virtue of accumulated testing and accepted values arespecified in the various Codes and Regulations. Unprotected steelworkdoes not normally attract any FRL, except where specialised approachesare adopted. One example is in open car parks where full scale tests have

16

demonstrated that bare steel will not reach a critical temperature should acar catch fire (Ref 2.5).

Another example is composite steel deck floor systems utilising fireemergency reinforcement (Refs 2.6, 5.4, 5.5).

2.6.3. MATERIALS FOR FIRE PROTECTION

Where steel has to be protected, the most practicable way is to cover it orencase it in a protective material. Such material should be:

(a) Non--combustible

(b) Unable to produce smoke or toxic gases at elevated temperature

(c) Able to be efficiently and uniformly applied

(d) Durable to prevent dislodgment, and

(e) Thermally protective.

Table 2.6 compares passive fire protection costs and gives an approximateindication of their costs. These costs may not tell the whole story where aprotectedmember is exposed to view andwill be given a decorative finish --some systems are less costly than others to decorate.

Another important factor to be borne inmind is that dry systems cause lessdisruption to other trades and the building schedule, and therefore canbring significant indirect cost savings in terms of shorter overallconstruction time.

Commercially available materials must be able to demonstrate theircapability of achieving a fire resistance level as part of building systems.The various manufacturers can supply the necessary accreditation andtechnical data by reference to tests conducted at recognised fire testingstations. See also Ref 2.11.

TABLE 2.6 PASSIVE FIRE PROTECTION COSTS

Section Mass Intumescent Paint Vermiculite Spray Vermiculite Spray

Fire Rating Level Fire Rating Level Fire Rating Level

60 min 120 min 180 min

(kg/m) $/sq m $/sq m $/sq m

60.5 and less 108 27 37

60.6 to 160 106 23 35

160.1 to 455 104 17 33

Notes: 1. Rates include supply and paint/spray.2. Intumescent paint cost includes ZnSi and Class 2--1/2 blastcleaning costs.

2.7. Specifications


The specification is important because it forms part of the tenderdocuments and ultimately becomes part of the contract documents. Itspurpose is to cover aspects of the work that fall between the legal contractclauses and the technical data shown on drawings.Such aspects may include:(a) Workmanship standards(b) Tolerances(c) Inspection levels etc.

In past years the specification was essential for the designer to convey tothe contractor exactly what was wanted. Nowadays so many of thesematters have been codified that a detailed specification has become lessnecessary.

17

The specification should not repeat material that is already in the relevantcodes or standards. Nor should it become a repository for informationwhich should more properly be shown on the drawings -- nowadays mostdesign offices use standard notes on their drawings in order to handle thisaspect more efficiently. A set of guideline notes are provided in AISC’sSteel Construction Journal, Volume 29, Number 3, September 1995 (Ref2.1). However, such standard notes should always be checked as eachdrawing is prepared to ensure that they are relevant.

A specification should be precise so that both parties to a contract knowwhat is required, and should clearly state what the contractor is required todo andwhat he/she is to refrain from doing. Great caremust be taken in thewording, with definitive requirements being stated and all allowablealternatives clearly specified. Vaguegeneral statementswhich couldmeandifferent things to different people should be avoided.

The requirements specified should be designed only to produce work ofappropriate quality to the building requirements, while avoidingunnecessarily tight requirements which only add to the cost.

Experience has shown that short and precise specifications helpconsiderably in the smooth flow of the work, and thus have a beneficialinfluence on costs. Conversely, long and repetitious documents can easilylead to misunderstanding, contractual arguments and expensive delays.

2.7.2. WORKMANSHIP STANDARDS

Standards of workmanship and quality are extremely difficult to define inwords. In the past many specifications attempted to do so by incorporatingsuch phrases as ’workmanship shall be of first class quality’ or ’membersshall be true to line and neatly finished’. However, when tested suchclauses are meaningless and fortunately are becoming rare in modernspecifications.

In practice the owner’s and designer’s interests are best protected byobserving these three principles:

(a) Use the tolerance and workmanship standards specified in theappropriate Code, e.g. AS 4100.

(b) Select inspection procedures and frequencies appropriate to theclass of work, using Code guidance (e.g. AS 1554) where avail-able.

(c) Select the fabrication and/or erection contractors on the basis ofproven capability, using their previous work as the most reliableindicator of their quality. Check that they have quality assuranceprograms.

2.7.3. TOLERANCES

Tolerances on theex--mill dimensions of steel sections andplates are listedin AS 3678 and AS 3679 (Parts 1 & 2). The necessity for these tolerancesarises because of factors in the steel--rolling process, including rollingspeed, roll wear, roll adjustment and differential cooling.

A study of the Standards shows that these dimensional tolerances can besignificant enough to warrant consideration in detailing and fabrication; Fig2.3 gives some examples.

(a) allow for variation in beam depth in flange splice and for off--centre of webs in web splice.

(b) any connection to column web or column flaange must make al-lowance for out of square, especially end plate connections -- al-low for shimming where necessary (may involve tapered shims).

18

(c) web side plate connection -- allow for out of square of columnflange and off centre of beam web.

Fig 2.3 Typical connections where allowance for mill tolerance isneeded

Experienced fabricators are aware of the possibility of dimensionalvariations, and it is normal practice to match members at splices in such away as to minimise the effect of these variations.

Tolerances on the dimensions of fabricated members and erected framesare given in AS 4100.

The tolerances specified can be considered as related to the designprovisions of the Code. Thus for structures designed in accordance withAS 4100, there is no case for specifying tighter tolerances since the tightertolerances are not then consistent with the design assumptions, nor withthe manufacturing tolerances of the raw steel.

These fabrication and erection tolerances can be realistically andeconomically achieved and are consistent with world wide practice. Theyshould not be varied without compelling reason.

It must be particularly noted that the specifying of tighter tolerances can bea costly decision which, in most applications, will serve no purpose anddestroy consistency. It is also recommended that tolerances be specifiedby simple reference to the provisions of AS 4100.

Where dimensional tolerances are not defined, there is plenty of room forargument and contractual dispute, as most experienced designers andfabricators know. Conversely, where allowable tolerances are clearlystated, it is a simple matter to decide whether a component or structurecomplies or not.

2.7.4. CAMBERING

The practice of cambering beams is intended to provide an upward ’set’that will counteract the downward deflection due to normal working loads.Several obvious problems present themselves with this procedure:

(a) It is difficult to calculate accurately the true deflection of a mem-ber under working loads;

(b) It is difficult to control accurately the degree of camber induced ina member; and

(c) Cambering requires the fabricator to perform a difficult, andhence expensive, fabrication operation.

There are two main methods by which rolled sections are cambered. Thefirst involves the use of some form of heavy press, such as a hydraulicside--press. These machines are massive and costly, and are found in theshops of only the largest companies.

Most fabricators employ the alternative method of controlled heating andshrinking using a standard flame--cutting torch.

Both of these methods involve a degree of trial--and--error in the setting ofthe member so that cambering is a slow, labour--intensive and thereforerather costly procedure in the fabrication process.On simple, well--detailedbeams it can more than double the actual fabrication cost.

It is therefore an operation to be called for only when absolutely necessary.

Generally, where members are ultimately concealed from view, or ifexposed are unlikely to cause visual offence, cambering is pointless. Anexception is sometimes found in steel beam/metal deck composite floorsystems where it is desirable to camber against the deflection due to thewet concrete because of the ’springiness’ of the whole system duringpouring.

If the requirement to camber is based on a need to offset increaseddeflections in light members, consideration should be given to using astiffer member without camber. There is certainly scope to do this, as thesaving on cambering costs would, to a large extent, offset the increase inthe cost of the heavier member.

Camber is measured with the member flat on the floor with the webhorizontal. Where a member is specified to be cambered, it is reasonable

19

to accept a tolerance on the specified camber similar to theout--of--straightness tolerance of AS 4100. To maintain tolerances closerthan this can be very costly indeed (Ref 2.10).

2.7.5. TEMPORARY BRACING

Problems often arise when the specification requires the erector to supplytemporary bracing for a structure. Sometimes the erector is required todesign this bracing and be responsible for its performance. In line with newoccupational health and safety regulations, erectors should developerection plans including temporary bracing requirements with the principalcontractor. These plans may need to be checked by the design engineer.

So--called ’temporary bracing’ actually falls into two categories:

(a) Erection Bracing -- the bracing or guys required to support indi-vidual members during their erection.

(b) Temporary Bracing -- required in order that the steel skeleton re-mains plumb and in a safe condition after erection is completed,until permanent bracing elements such as shear walls are built.

Erection bracing is the principal contractor’s and erector’s responsibility inrelation to the supply and its removal on completion.

However, temporary bracing which is to be left in place until otherstabilising elements are built is a different matter. Its design requiresknowledge of the building sequence and of other factors. Normal prudencewould suggest that it must be designed by the Engineer. Any special orunusual features of the structural design that may limit or affect stabilityduring erection should be emphasised on the construction drawings.

2.7.6. INSPECTION

Whilst some level of routine inspection is obviously necessary in theowner’s interest, it should always be remembered that inspection in itself isa non--productive expense. It should therefore be specified with discretion.

In most contracts most of the inspection is directed at high--strengthbolting, welding and surface treatment. Guidance on inspection levels andmethods is given in the relevant codes and standards:

AS 1554 Structural Steel Welding

AS 2312 Guide to the Protection of Iron and Steel against ExteriorAtmospheric Corrosion

AS 4100 Steel Structures

The specification should define the nature of inspection to be carried outand themethods to beused. This latter is especially important in the caseofnon--destructive weld testing where there is a range of methods availablewith widely varying costs. Specifications requiring 100% x--ray testing onall butt welds in standard industrial buildings impose significant andwasteful costs on projects. The welding test requirements for the oil andgas industry should not be applied on everyday industrial or commercialstructures. Appropriate testing levels are essential for economicalstructures.

Where an independent inspection authority is to be engaged it should bemade clear in the tender documents whether or not the fabricator is tocover the cost in his price quotation.

The following guidelines will assist in setting up effective and economicinspection procedures:

(a) Inspection methods and levels should be compatible with thequality and tolerance requirements of the codes applying to theparticular class of work. Inspectors should not seek to imposehigher standards.

(b) Early inspection efforts should be directed towards checking thatthe fabricator’s procedures will produce the required results.Thus inspection will be more intensive at the start of the job, andcan be relaxed to a nominal level when production methods areproven.

(c) The inspectors themselves should not only be experienced intheir particular fields but should also have a steel fabricationbackground. This allows the inspector and fabricator to come toagreement quickly on many day--to--day matters on the basis of

20

common experience, rather than hold up the work unnecessarilyon minor details.

2.7.7. SUMMARY FOR SPECIFICATION WRITERS

1. Specifications are not as important as in previous years becauseso much has now been codified.

2. Omit meaningless clauses, no matter how well--sounding. Theycan achieve nothing but may exacerbate disputes.

3. Do not include information in specifications that should be moreproperly shown on drawings.

4. Call up AS 4100 and associated documents.

5. Keep it brief.

21

3. Framing Concepts and Connection Types

3.1. Introduction

The framing system and framing layout chosen for a particular applicationwill be influenced by:

(a) Nature and level of the loads to be resisted

(b) Requirements and restrictions on useable space within theframework, and

(c) Constraints imposed by architectural requirements.

One advantage of steel framing is the diversity of solutions that arepossible for any given application.

There are available to the designer two basic connection types, namely:

• Rigid connections, and

• Flexible connections.

The above connections may be used in the three basic framing systemsavailable:

1. Two--way rigid frameworks

2. One--way rigid/one--way braced frameworks, or

3. Two--way braced frameworks.

Judicious selection of the appropriate framing system and connectiontypes is a prerequisite to an economical structural design. Once a framingsystem is selected, the connection types to be used follow directly, thussetting bounds to the final cost of the structure. Economy in detailing,fabrication anderection canonly serve tomove the final design towards thelower bound of cost established by the framing system.

In the discussions of connection types and framing systems which follow,no distinction will be made between single or multi--storey buildings sincethe basic principles apply to most buildings.

3.2. Connection Types

3.2.1. DESIGN METHODS IN AS 4100

AS 4100 allows the use of three different design methods, wherein thebehaviour of the connections is fundamental to the design method. Thesemethods are:

(a) Rigid Construction, in which it is assumed that the connectionshave sufficient rigidity to hold the original angles between themembers unchanged.

(b) Semi--Rigid Construction, in which the connections may not havesufficient rigidity to hold the original angles between the mem-bers unchanged, but are assumed to have a capacity to furnish adependable and known degree of flexural restraint.

(c) Simple Construction, in which the connections are assumed notto develop bending moments. The stability of the structure istherefore provided by triangulation (i.e. bracing) or by separateshear walls -- see Section 3.3 et seq.

Clearly from these brief descriptions it is seen that connection behaviourhas a significant influence on design.

Allied to design methods (a) and (c) above are the basic connection typesnoted in Clause 3.1, namely:

• Rigid connections, and

• Flexible connections.

Design method (b), Semi--Rigid Construction, will not be consideredfurther in this publication.

22

3.2.2. FLEXIBLE CONNECTIONS

Flexible connectionsare used in steel structures designedusing the simpledesign method of AS 4100. These connections offer low restraint to beamrotation, being close in behaviour to that of an ideal pin.

Typical flexible connections are shown in Fig 3.1. The most commonflexible connections in use in Australia are the flexible end plate (Fig 3.1(c)), the angle cleat (Fig 3.1 (d)), and the web side plate (Fig 3.1 (e)).

Such connections are:

(a) Assumed to behave as a simple support

(b) Simple to fabricate

(c) Simple to erect, and

(d) Less costly of the two connection types.

Flexible connections shown in Fig 3.1 are totally standardised in the AISCStandardized Structural Connections publication (Ref 1).

(a) Angle Seat (b) Bearing Pad

(c) Flexible End Plate (d) Angle Cleat(single or double)

(e) Web Side Plate

Fig 3.1 Flexible connections

23

3.2.3. RIGID CONNECTIONS

Rigid connections are used in steel structures designed using the rigiddesign method of AS 4100. These connections offer very high restraint tobeam rotation, being close in behaviour to fully fixed (or encastre)connections.

Typical rigid connections are shown in Fig 3.2. The most common rigidconnections in use in Australia are the stub girder connection (Fig 3.2 (b))and the bolted moment end plate connection (Fig 3.2 (c)). These are alsocovered in the AISC Standardized Structural Connections publication (Ref1).

Rigid connections are:

(a) More complex in fabrication

(b) More difficult to erect where tight tolerances are involved, and

(c) More costly of the two connection types.

(a) Field welded moment connection -- with erection cleat (also use filletwelded web cleats in lieu of beam web welds).

(b) Stub girder connection -- fully shop welded beam stub, pliced on site.

24

(c) Bolted Moment End Plate Connection

Fig 3.2 Rigid connections

3.3. Basic Framing Systems

3.3.1. TWO--WAY RIGID FRAMEWORK (Fig 3.3)

Two--way rigid frameworks comprise two planes of rigid framesintersecting at right angles using common columns at their intersection.Such frameworks resist lateral forces in both planes by frame actionwithout the need for any separate stabilising elements. All thebeam--to--column connections must of necessity be of the rigid type, andthe columns may need to have approximately equal stiffness in bothdirections, so that boxed or tubular columns may be employed due to theirhigh stiffness about both principal axes. Under the action of lateral forces,there is always some sway as a result of the elastic deformation of theframework, but there is no problem in designing the structure in such awaythat this sway is kept within an acceptable limit.

Themain advantage of the two--way rigid framing system is in the completefreedom in planning it offers. On the minus side is the necessity for themore costly rigid connections and columns.

Since the rigid design method of AS 4100 is used for this framework, theanalysis can be either by the elastic or the plastic method, the latter beingmore mass economical due to a better utilisation of material. It does,however, require slightly more costly connections.

The main design advantage of a rigid beam--to--column connection lies inthe reduction in the sizes of the floor beams due to the end fixity. Increasedcolumn section mass may, however, counterbalance this saving sincelarger bending moments need to be considered in the columns. Theresulting increase in material cost should not exceed the extra costinvolved in the rigid connections for the resulting framework to be aneconomical selection.

Fig 3.3 Two--way rigid framework

Typical applications that may use this type of framing include:

(a) Multi--storey frames

(b) Iow--rise rectangular frames (especially where architectural re-quirements restrict the use of bracing elements)

(c) Heavy industrial structures (especially where planning needs re-strict the use of bracing elements), and

(d) Architectural structures that can be modelled as two--way rigidframes.

3.3.2. ONE--WAY RIGID FRAMEWORK (Fig 3.4)

One--way rigid framework has been used quite extensively for the simplereason that the most commonly employed structural sections, (namely,universal sections) exhibit high bending resistance about the x--axis andinferior bending resistance about the y--axis.

The relatively more expensive rigid beam--to--column connection isrequired in the unbraced plane, while simple connections of the flexibletype can be utilised in the braced plane. In comparison with the two--wayrigid framing system, there is slightly more restriction in planning the floorlayout since spacemust be reserved for the stabilising elements. However,this is seldom a problem since the bracing can be arranged within thethickness of the perimeter walls or alternatively be tied back to a bracingelement.

25

As a general rule, it is necessary with this arrangement to construct a rigidsystem consisting of either wind girders or a diaphragm having greatrigidity in its own plane and being properly connected to the framingsystem. With such a system, it becomes possible to distribute the lateralforces to the individual stabilising elements. A reinforced concrete floorslab resting on steel beams is one example of a reliable diaphragm action.

In the unbraced plane, the frame can be analysed as a rigid frame using themethods outlined in Clause 3.3.1. In the braced plane, ’pinned’ connectingbeams are usually assumed, although rigid connectionsmay be employedin order to provide beam continuity and/or reduce the lateral deflection ofthe frame in this direction. Such a procedure, however, may not be aneconomical overall solution.

Typical applications that may use this type of framing include:

(a) Low--rise industrial frames (portal frames)

(b) Rectangular frames (especially where bracing can be accommo-dated within the perimeter)

(c) Industrial structures, and

(d) Architectural structures (bracing elements are often used as partof the architectural feature).

Fig 3.4 One--way braced, one--way rigid framework

3.3.3. TWO--WAY BRACED FRAMEWORK (Fig 3.5)

Two--way braced frameworks depend on stabilising elements arranged sothat lateral forces from all directions can be effectively resisted. Theframework itself can be constructed in the form of beams ’pin’ connected tothe columns, in which case the beams are designed as simply supported,and the columns as essentially axially loaded members, with beamreactions acting at small eccentricities off the column face. It is mostimportant with this system to have a relatively rigid floor system capable ofpreventing distortion of the framework in plan.From the design engineer’s point of view this is the easiest framing systemto analyse since there is very little interaction between the framingmembers. Not surprisingly the two--way braced system is also veryattractive from the cost point of view, since the simplicity of the memberconnections can offset the cost of the somewhat heavier floor beamsrequired with this system.The stabilising elements can be orthogonally arranged shear walls, bracedpanels or cores (Clause 3.3.5). These stabilising elements have to belocated to give a well balanced system and the floor plan mustaccommodate this. In most cases it is possible to utilise the walls aroundservice blocks or external walls (Clause 3.3.5). External bracing, formingpart of the architectural feature, can also be utilised.In this type of design, all beams are assumed to be pinned at theirconnections to the columns. In fact the connections are not pins but aflexible type so that free end--rotation can be assumed. The design of thebeams can be carried out without reference to the framing as a whole.However since the beams, designed as pin--ended, tend to be larger in sizethan if fixed connections are used, it is imperative to design them to be asefficient as possible.One of the ways of securing economy is by making use of any concretefloor slab present to achieve composite action. The main advantage ofcomposite action is that it augments the beam with a ’concrete flange’ andalso increases its depth. Ref 5.3 contains a full discussion of compositesteel beam design.The columns carry only the gravity loads. Some bending is present due tothe eccentric application of the beam reactions, but the effect of thisbending is usually small. The bracing system is usually assumed to takemost of the lateral forces.

26

Typical applications that may use this type of framing are Iow tomedium--rise rectangular frames (up to 50--storeys -- especially usingcores, either steel framed or slip--formed concrete).

Fig 3.5 Two--way braced framework

3.3.4. SUMMARY OF FRAMING SYSTEMS

FramingSystem

Advantages Disadvantages

Two--way rigid No stabilizing elementsrequired for lateralforces in any plane.Freedom of layout plan-ning.Plastic design methodscan be used if desired --economical in material.Continuous beam de-sign leads to reducedbeam size.

Requires the use of rig-id connections, whichare more costly thansimple connections.Columns ideally shouldhave near equal stiff-ness in both directions --hence fabricated boxcolumns may be need-ed.Large column move-ments.

One--way rigid /One--waybraced

Simple connections(least costly type) usedin the braced plane.Can use I columns --usually rolled sections.Can use plastic designmethods and continu-ous beam design inplane of rigid connec-tions -- saving in materi-al.

Rigid connections usedin unbraced plane.Some restriction onplanning layout; stabiliz-ing elements required inone plane.

Two--waybraced

Simple connectionspossible -- least costlytype.Usually use I columns.Beams assumed simplysupported for design;columns designed foraxial load only at smalleccentricity.

Restriction on planninglayout because of re-quirement for provisionof stabilizing elements.Little interaction be-tween elements.Heavier beam sizes.

27

3.3.5. STABILISING ELEMENTS

Construction elements whose function is to provide a means of stabilisingthe framework in either one or two planes may be divided into the followingcategories:

(a) Triangulated steel bracing panels using the X, K, or diamond pat-tern of diagonal members -- Fig 3.6(a);

(b) Vertical Vierendeel cantilevers in steel -- Fig 3.6(b);

(c) Triangulated steel core -- Fig 3.6(c);

(d) Reinforced concrete or masonry shear walls -- Fig 3.7(a);

(e) Reinforced concrete or masonry cores or shear tubes -- Figs3.7(c) and (d);

(f) Brick in--fill panels and walls -- Fig 3.7(e);

(g) Light metal cladding used on the stressed skin principle.

(a) Triangulated bracing systems

(b) Vertical Vierendeel cantilever

(c) Triangulated core

Fig 3.6 Stabilising elements built in steel

When stabilising elements are constructed of concrete ormasonry, it iswellto remember that some means of temporary bracing may be requiredduring the early construction phase, since the steelwork may not havesufficient in--built resistance to withstand lateral forces prior to constructionof the stabilising elements. Rigid systems of wind girders or diaphragms(Fig 3.8) may also be required to distribute lateral forces to the stabilisingelements.

Openings can readily be incorporated in all types of stabilising elements,although there is some restriction on the maximum size of openings. It isimportant, however, to distinguish between the low--rise building whichdoes not require large stabilising elements, and tall building where thestabilising elements are required to carry very large forces and have arelatively high stiffness.

(a) Shear Wall (b) Opening may beaccommodated in shear wall

28

(c) Shear tube (d) Corner walls

(e) Brick in--fill wall

Fig 3.7 Stabilising elements built in reinforced concrete ormasonry

(a) Wind girders as sole means of transfer of wind forces

(b) Concrete floor slab as diaphragm

Fig. 3.8 Floor deck bracing systems

(i) Rigid frame action

(ii) Steel lattice bracing

(iii) In--fill wall panel

29

(iv) Transverse wall

(v) Stairwell walls

(a) Vertical systems

(i) Lateral force transmitted to foundation at every column -- no horizontalbracing

(ii) Horizontal wind girder

(iii) Use of floor as diaphragm

(b) Horizontal bracing systems

Fig 3.9 Action of lateral force resisting systems (from Ref 5.2)

3.4. Cost and Framing System

The type of framing system selected to satisfy all the design constraints willhave a profound effect on the structural cost. The labour cost in thefabrication of a fully braced system employing simple flexible connectionsis much less than the labour cost in fabricating a fully rigid system usingmore complex moment connections. On average the rigid frameworkrequires about 2.5 times the labour cost input in the fabrication process.

To achieve the most economical final structure the designer has to find asolution which, within the various constraints, will provide for maximumcost effect in both material and fabrication labour input.

30

3.4.1. MULTI--STOREY BUILDING

The following example illustrates the way in which cost effective solutionscan be achieved and the importance of selecting a framing system of leastcost to serve function. A minimum mass solution may not always producethe best cost effect-- in this case the minimum mass fully rigid framerequires substantial additional labour input for connections in comparisonwith the simpler flexible connections used in the braced system. Thus theapparent savings inmaterial cost are less than the increase in labour costs.The adoption of a fully rigid frame, although of significantly lower mass ofmaterial, will not produce the best economical solution unless such asystem is demanded by constraints such as freedom of layout orarchitectural bias against cross bracing.In structures such as city buildings even greater benefit in cost is achievedby using the service core as a stabilising element in lieu of cross bracing.

Braced Frame Elastic Design48.1 tonnes of beams and columnsat cost ratio 1.0 = 48.15.5 tonnes of bracingat cost ratio 0.8 = 4.4

53.6 tonnes, Total Cost = 52.5

Unraced Frame Plastic Design38.2 tonnes of beams and columnsat cost ratio 2.0 = 76.4

38.2 tonnes, Total Cost = 76.4

Fig. 3.10 Frame example

31

3.4.2. SINGLE--STOREY INDUSTRIAL BUILDING

Similarly in other types of structure the framing system will influence finalcost. In typical factory buildings, for instance, which were once framed bycolumn--and--truss systems, it is quite clear that the rigid portal frame is themost economical system. Fig 3.11 shows that truss systems are obviouslymore efficient on a mass/ unit area basis. However, on a cost basis, theinherent simplicity of the portal frame renders it less costly to fabricate andshows up as the economical solution within the range shown (see Fig3.12).

Fig 3.11 Relationship between mass/unit area and span

Fig 3.12 Relationship between cost/unit area and span

These examples are intended to illustrate the importance of carrying out anexamination of framing system costs at the earliest design concept stage.The best end result will be obtained by selecting the framing systemwhichwill satisfy function and economy.

3.5. Framing Details

Having thus selected the framing system as previously discussed, it isimportant to consider framing details for that particular system so that thebest cost effect will be achieved.In general the following points must be considered.

3.5.1. SYMMETRY

In many cases symmetry is available in framing systems simply as a resultof functional requirement e.g. city building frames. However in other typesof structure it is often possible to arrange symmetrical layout withoutprejudice to function. Symmetry will invariably lead to the possibility ofrepetition and this will provide for the most economical fabrication anderection.

32

3.5.2. RATIONALISATION OF MEMBERS

The grouping of members in a framework with respect to type and size willalso have advantages in fabrication and erection economy. Series ofmembers of the same size and length will be processed more efficiently inthe shop. At the erection stage the greater number of identical items willprovide for speedy erection.

Obviously in grouping of members considerable skill is required of thedesigner. Too much grouping of member size can be wasteful of materialand too littlewill add to detailing, fabrication and erection costs. In general itis advisable to minimise the number of highly individualised members andthus provide for maximum repetition and interchangeability.

3.5.3. STANDARDIZATION

Connections

The AISC publication Standardized Structural Connections (Ref 1)contains highly standardized data for both simple flexible connections andrigid connections. The use of such a system, with constant dimensionalcriteria, allows for efficient fabrication by optimising the use of modernautomated equipment in the fabrication shop.

It is also recommended that the designer consider the various suitablealternativeswithin a particular connection group (i.e. either flexible or rigid).This will allow the fabricator to select from the Standardized StructuralConnections publication the connection typewhich canmost economicallybe fabricated with the equipment available and which will satisfy thedesigner’s requirements.

The important thing to remember is that the greater part of the fabricationprocess is involved in preparing members to be connected to one anotherand the more standardization, especially with respect to connectiongeometry, which can be incorporated in a design, the better will be the finaleconomy.

Finally, in selecting connection types, try to consider groups of membersrequiring only one operation in the shop. This can be accomplished byarranging for a series ofmembers (e.g. primary floor beams) to require only

cutting to length and holing, while another series (e.g. beams connecting toprimary beams) to require only cutting and welded fitments (see Fig 3.13).

Group 1 Cutting and holing only

Group 2 Cutting and welding only

Fig. 3.13 Beams for economic fabrication

Bolts and Welds (Fasteners)

It is advisable to consider the standardization of fasteners within a givenstructure.

Where possible, adopt the use of one bolt size, grade and procedure withinthe structure. See Section 6. Similarly, use one electrode strength grade,one weld category and if possible one weld size (in the case of fillet welds).See Section 7.

3.5.4. SIMPLICITY

Simple detailing for such things as stiffeners, bracing gussets, attachmentcleats and base plates, will produce the greatest economy in fabricatedwork. The number of man--hours spent can increase dramatically if suchdetails become complex. See Section 8.

The following general examples show how cost extras can be incurred:

Structure A Commercial Building

A relatively simple beam and column framework with repetition of bay sizeandminimumbracing components; standard connections (two types) usedthroughout with snug--tightened bolts.

Structure B Similar Building

This example is considerably more complex having varying bay sizes,spandrel periphery trusses and extensive bracing in the wall planes;

33

connections are of several types and custom designed, some usingfully--tensioned bolts.

Cost Index

Structure A Structure BMaterial 1.00 1.00Shop Labour 1.00 2.08Painting 1.00 1.22Shop Detailing 1.00 1.67Erection 1.00 1.25

Notes: 1. Cost indices are presented for the purpose of comparisononly.2. Some common items such as administrative overheads,profit and builder’s mark--up have been excluded from thiscomparison.

It can be seen therefore that for two structures performing similar functionthe final cost of structural steel is sensitive to the complexity of workrequired. For example, the introduction of truss work into the framingsystem together with more complex connections has more than doubledthe shop labour component for Structure B. Also costs are higher for shop

detailing (increased complexity required additional time), painting(increased surface area for truss work) and erection (complex connectionsand fully--tensioned bolts add to cost).

3.6. Conclusion

The selection of the system for a steel framework is the most fundamentaldeterminant of the final cost of the erected structure. Once the basicframing system is selected, the connection types which may be used arechosen. Thus, the basic cost of the erected framework is predetermined,recognising that this cost may vary within a certain range. Economicdetailing, fabrication and erection can only move the final cost towards theminimum possible within this range.

It is essential that at the preliminary design stage the full range ofalternative framing systems are evaluated and compared before makingthe final selection. This comparison of alternatives must be done on thebasis of erected cost -- not on the basis of mass.

Good design i.e. economical design, should take into account all theinfluences which have an effect on the form and cost of the final structure.The economics of designmust be considered in this context since the clientis mainly concerned with what he/she pays for-- a complete building whichmeets his/her needs at least cost.

34

4. Industrial Buildings

4.1. Introduction

Steel--framed buildings in common use for industrial purposes can beclassified into three broad categories:

(a) Warehouse and factory buildings

(b) Large span storage buildings, and

(c) Heavy industrial process plant structures.

In the design of industrial buildings, functionmore than any other factor willdictate the degree of complication and hence the economy possible.Towards this end, the designer should obtain as much knowledge aspossible of the industrial process or purpose for which the building isintended, and of the limitations this might force on the structure.

In this way, an optimum balance between function and economy can beachieved.

The main dimensions of an industrial building are usually determined froma combination of functional and design considerations.

Its width is derived first froman owner’s study of the space required to carryout the processing or storage operations. The designer then needs toconsider whether this width can be provided economically by a single clearspan, or whether multi--bay spans are feasible.

Likewise the overall length is usually readily determined by the owner, butthe designer should give thought to the optimum bay length. Some of thefactors affecting the choice are:

1. Foundation conditions, and their ability to accept the columnloads

2. Crane runway girder considerations (see Clause 4.2.5)

3. Purlin and girt capacities (see Clause 4.2.6)

4. Masonry bond dimensions, and

5. Tilt--up concrete panel size and available cranage.

The building height is again usually a functional consideration; for buildingswith overhead travelling cranes the critical dimension is the clearancerequired under the hook.

In most areas of Australia there is no snow, and therefore fairly low roofpitches are practicable. The steeper the slope the better the structuralaction, but this benefit is usually outweighed by additional sheeting costs.In practice, roof pitches between 5_ and 10_ are preferred. These pitchesare suitable for any of the continuous length steel sheet roofing profiles,some of which are adequate for pitches down to 1_.

4.2. Warehouse and Factory Buildings

4.2.1. GENERAL

In the early days of steel--framed industrial buildings the economic solutionwas a column--and--truss configuration (Fig 4.2 (a)). However, since trussfabrication is inherently labour intensive, rising labour costs have excludedthese truss systems from normal factory or warehouse applications.

Presently, rigid ’portal frames’ fabricated from universal beams offer themost economic structural solution in the usual span range of 15 to 45metres. For very large spans, portal trusses (see Fig 4.18) are often usedin lieu of the portal frame.

Although the portal frame may require a greater mass of steel than theequivalent column--and--truss arrangement, the savings in the cost offabrication and erection due to the relative simplicity of the work almostalways make it the optimum system in the span range given above.

Tominimise the overall cost of warehouse and factory buildings, designersshould be aware of the major steelwork cost components. Effort can then

35

be focused on cost components that can reduce the overall cost. Figure4.1 shows the various cost components in relation to a warehouse.

4.2.2. STANDARDIZED PORTAL FRAMES

Overseas, particularly in North America, the portal frame structure hasbeen developed to the stage where many companies offer a standardrange of buildings in spans up to asmuch as 50m. Economies of scale andproduction line manufacture have made these ’catalogue’ buildings acost--effective choice for many industrial as well as commercialapplications.The samemanufacturing and marketing techniques have been attemptedin Australia, but with limited success, probably due to ourmuch smaller andmore widespread demand. As a consequence, practically all larger portalframe structures built in Australia today are custom designed andmanufactured. This is not as inefficient as it may sound, because there aremany standardized routines in both the design office and the fabricationshop.On the other hand, smaller buildings (sheds, garages, etc) are widelyavailable in Australia as standard catalogue items. Nowadays these areoften manufactured entirely from cold--formed sections rather than fromtraditional hot--rolled sections.

Roof & WallSheeting sup-ply & fix =37%

Purlins & GirtsSupply & Fix =24%

Steel Supply =20%

Fabrication = 15%

Surface treatment 2%

Steel erection 2%

Fig 4.1 Steelwork Cost Components for Warehouses

(a) Column and Truss

(b) Portal Frame

Fig 4.2 Configuration of framing systems for a factory building

36

Fig 4.3 Details of bolted portal frame

4.2.3. CUSTOM DESIGNED PORTAL FRAMES

In this case, a client engages an Architect and Consulting Engineer whoprepare design drawings and submit the project to tender. The contract isusually awarded to a builder who then sub--contracts the structuralsteelwork to a steel fabricator on the basis of the Consulting Engineer’sdrawings.

The portal frames will usually consist of universal sections in order to beeconomic in fabrication -- see Fig 4.3. A variety of connection details areencountered, but only a limited number are truly economic for such frames.Fig 4.4 shows examples of economic details using bolted knee and apexjoints, while Fig 4.5 shows examples of economic details for frames usingshop welded knee and apex joints and bolted rafter splices.

For spans up to 20m a uniform column and rafter section is the mosteconomic but for greater spans haunching of the rafter may provide amoreeconomical system. Haunching is most economically achieved by using acut universal beamsection in themanner shown inFig 4.3,with the depthofthe section at the haunch about twice the rafter depth. The haunch length isusually of the order of 10%--15% of the span of the rafter.

37

Fig 4.4 Details for welded portal frame (with bolted rafter splice forfield erection)

The selection of either bolted or shop--welded knee and apex joints will begoverned by the span of the frame and the transport and erection facilitiesavailable for a particular job.

In general the dimensions given in Fig 4.5 are a guide to limitations onmaximum size imposed by transportation considerations.

For frames of larger dimensions than those indicated in Fig 4.5,consideration would have to be given either to special transport facilities oradditional field splices.

A further discussion on portal frame details can be found in Clause 8.5.

Fig 4.5 Transportation limitations for portal frames

4.2.4. BRACING OF PORTAL FRAMES

Bracing Disposition

The typical disposition of bracing panels for portal frames buildings isshown in Fig 4.6.

For shorter buildings (up to 60--80m), a single end braced bay is all that isnecessary to stabilise the building structure. However, this arrangementrequires wind forces on the opposite end to the braced bay to betransferred along the building length by way of longitudinal eave and ridgestruts. This may require heavy struts, and it is often more economic toprovide braced panels in each end bay and remove the necessity toprovide these substantial struts.

38

Temperature expansion of buildings 60--80m long is most commonly andeconomically catered for either by allowing the expansion force to act onthe end bay bracing or by the use of slotted holes (or oversize holes) in theconnections of the longitudinal struts to the columns.

In longer buildings (over 60--80m), corner bracing can be a disadvantagesince the expansion involved is too much to be accommodated by theabove methods. In such cases, a central expansion joint can be provided(thus effectively making two buildings -- Fig 4.7(a)) or alternatively, thebracing can be provided near the central interior bays -- Fig 4.7(b). For thelatter alternative, substantial longitudinal strutsmay be required to transmitwind forces from the end walls through to the braced bays. Whether thissolution is economic depends on the increase in size of the longitudinalstruts required for the latter solution compared to the additional cost of theextra column in the expansion joint solution.

Elevation

Plan

Fig 4.6 Bracing panels

(a) Use of central expansion joint for buildings over 60--80m long

(b) Alternative bracing system for buildings over 60--80m long

Fig 4.7 Bracing for long buildings

Bracing Details

For sheds and small buildings rod bracing, tensioned by turnbuckle or bydeliberately ’detailing short’, is the most economic solution, although thereis an alternative school of thought which uses angle bracing. With rodbracing, the ability to plumb frames and square the buildings by using theturnbuckle adjustment makes for easier erection.

For wide frame spacing, rod bracing will tend to sag over the longer spaninvolved andmay present someproblems in effectively bracing the roof. Aswell, rod bracing in the walls may become subject to physical damageduring occupancy. Angle bracing can overcome these difficulties.

Tubular sections are efficient members for bracing in larger structures.Their inherent properties provide high load carrying capacities for lowmass of material and make circular and rectangular hollow sections (CHSand RHS) very attractive from a design point of view. However, for theseadvantages to be reflected in the overall economy of the fabricatedstructure attention should be paid to the end connections since theirpreparation involves the largest part of the fabrication cost, (see Ref 4.7).

Economic connection details for bracing members are shown in Figs 4.8,4.9 and 4.10.

39

(a) End connection(b) Simple crossover intersection

(c) Intersection using a pipe piece(no turnbuckles needed)

Fig 4.8 Details for rod bracing

(a) End connection (b) Typicalintersection

Fig 4.9 Details for angle bracing

(a) Flattened end (CHS only)

(b) Welded tee end

40

(c) Slotted end plate

(d) Typical intersection

Fig 4.10 Details for tubular bracing

4.2.5. CRANES IN PORTAL FRAME BUILDINGS

The most common crane type used in portal frame industrial buildings isthe electric overhead travelling crane. The crane bridge travels on twolongitudinal girders which are supported at each portal frame of thebuilding structure. The design of a crane runway girdermust be consideredas an integral part of the whole building. At the same time, it must berecognised that because of the dynamic forces imposed on the runwaygirder, extreme economy in member and connection design is notrecommended and is considered unwise. The best solution may be a

heavier structure providing lower maintenance cost in the future operationof the crane.

Fig 4.11 Types of supporting columns

Fig 4.12 Crane runway brackets

41

Fig 4.13 Commonly used sections for crane runway girders andtheir relative fabrication cost

The method of supporting the crane runway girder depends on themagnitude of the crane wheel reactions (i.e. on the crane capacity and thecrane classification) and upon the structural characteristics of the portalframe column. Fig 4.11 shows some typical arrangements as follows:

(a) Separate crane column, acting with the frame column

(b) Combined frame and crane column

(c) Separate crane column, acting separately from the frame column

(d) Light frame column bracket, with the frame column acting as bothframe and crane column.

Generally types (a), (b) and (c) in Fig 4.11 will be chosen for heaviercapacity cranes as classified in AS 1418. In most factory type buildings,cranes will be of low to medium capacity (up to 5 tonnes) in which case thecrane runway girders could be supported on a column bracket (type (d)).This bracket should be proportioned to minimise stiffening of the framecolumn (see Fig 4.12).

Ref 4.4 is a publication on the design of crane runway girders and outlinesthe factors which affect the overall economy of both the crane girder andthe enclosing structure. Fig 4.13 shows the most commonly used cranegirder sections in portal frame industrial buildings, and gives an indicationof their relative fabrication cost. Ref 4.4 gives more detail and discussesother types of runway girders.

The cost of continuous girders is usually higher than for simply supportedgirders since the efficiency of themember is offset by higher erection costs.However, the most economical compromise is often to design and detailthe girder as continuous over two frame spans. This allows the fabricationof either rolled members or plate girders from stock material and thereforeminimises fabrication costs while still reducing the total number of girdersto be erected.

4.2.6. PURLINS

The sheet cladding of industrial buildings is attached to a framework ofsecondary members which is itself connected to the main frame. Thesesecondary members are known as purlins (for roof sheeting) or girts (forwall sheeting); the term purlin is used when referring generally to bothtypes.

In Australia industrial purlins consist almost exclusively of cold--formedmembers -- usually Zed or C sections, often formed from hot--dipgalvanized strip. These members are available from severalmanufacturers and in a variety of depths ranging from 100mm up to350mm in 50mm increments. The availability of section depths varies ineach State. Availability of the larger sections should be confirmed withsuppliers before being specified to avoid unnecessary delays and cost tothe project.

For average industrial buildings a purlin 200 mm deep appears torepresent an economic optimum, and it is the capacity of this size that oftenfixes the frame spacing typically 6 to 8m. The supply and fixing of purlinsand girts represent about 24% of the total steelwork cost for a warehouse.Judicious selection of purlins, and attention to design loads and details cancontribute to a significant reduction in overall project cost (see Fig 4.1).

Purlins are bolted to the rafters by means of a simple welded cleat (Fig4.14). Most manufacturers specify M12 bolts and some provide specialpurlin bolts having an M12 thread and an M16 shank. Purlin and cleatbolt--hole geometry has been standardized by the Australian Institute ofSteel Construction in the ’Standardized Structural Connections’ and mostmanufacturers conform to these standards (Ref 1).

Zed section purlins are shaped so that they can be lapped, and this featureallows the designer to take advantage of partial or complete continuity atthe splices (Fig 4.15). However in some cases the structural advantages of

42

continuity may be off--set by extra cost and complication in the purlinsthemselves.

C section purlins are normally used simply supported at the ends (Fig 4.16)or continuous over two spans.

For shorter bay lengths purlins can be obtained long enough to be usedcontinuously over two spans. This reduces deflection compared withsimple spans but does not give the same structural performance as a fullylapped system.

The performance of purlin systems requires in most cases the provision ofadequate lateral stability bymeans of ties or bridging. Purlinmanufacturerssupply such systems, and some also offer accessory items such as rakinggirts, fascias, etc.

Details of proprietary purlin systems, design information and load tablescan be obtained from manufacturers’ literature.

Fig 4.14 Standard purlin cleats

Fig 4.15 Zed section purlins with lap

Fig 4.16 C section purlins with butt joint

4.2.7. FLY BRACING

In a portal frame building either flange of both the rafters and the columnscanbea compression flange depending upon the assumedmagnitude anddirection of wind loading. The exterior flanges are normally adequatelylaterally braced by the purlins and girts, but sometimes the design mayrequire the provision of bracing to the otherwise unrestrained interiorflanges.

43

This is most conveniently accomplished by the inclusion of so--called ’flybracing’ at purlin intersections (see Fig 4.17). This can easily become avery costly detail, and unnecessary expense can be avoided by the use ofthe simple flat bar arrangement as shown.

Fig 4.17 Method of fixing fly bracing to standard punching

4.2.8. SHEETING

Coated steel sheeting are the most popular and economic claddingmaterial for both the roof and walls of industrial buildings. (There may insome circumstances be regulatory constraints on its use in walling).

A variety of profiles is available, ranging from traditional corrugatedsheeting to sophisticated ’concealed fix’ products. All of these sheets aremanufactured from continuous strip and therefore can be supplied in mostcases so as to eliminate end laps. It is usual practice for sheeting to be’custom cut’ by the manufacturer in the precise quantities and lengthsneeded for each particular project.

Except in cyclonic areas, steel roofing is capable of spanning about 1200mm in the case of corrugated sheeting up to as much as 2700 mm forstronger and deeper profiles. These figures relate to interior spans. Endspans for screw--fixed products should normally be limited to aboutthree--quarters of these figures. For walling, spans can be 25% to 50%greater.

It can be seen that the choice of cladding determines the purlin spacingwhich in turn can influence some of the basic design parameters such aspurlin size and bay length.

Steel sheeting is readily fixed to cold--formed purlins by means ofself--tapping screws. Special heavy duty self--drilling self--tapping screwswith in--built neoprene seals are normally used.

Concealed--fix profiles are secured by separate clips or straps which arenormally attached to the purlins. On the finished job these straps arehidden and there is no piercing of the cladding surface.

Where sheeting is to be painted for decorative purposes or to provideadded protection, considerable economy can be gained by the use ofpre--painted cladding. The factory--applied finish avoids costly site paintingand provides far superior paint adhesion and quality.

Full details of steel sheet cladding profiles, accessories, design and fixingdata etc., are obtained from manufacturers’ literature.

4.3. Large Span Storage Buildings

4.3.1. SPANS OF 45--70 METRES

When buildings of over 45m clear span are required for such purposes ascontainer storage, etc., consideration should be given to the use ofportal--truss systems for economy. Spans of 45 to 70m are economicallysatisfied with such systems (Fig 4.18).

Fig 4.18 Three--pinned portal truss

The factors affecting the economy of the fabricated structure in such trusssystems are those common to truss--work in general and these are

44

discussed in Clause 8.4. Other considerations such as bracing, sheeting,etc., are as discussed in Clause 4.2.

4.3.2. SPANS IN EXCESS OF 70 METRES

Spans greater than 70m are required for structures such as aircrafthangars, large stadia or storage buildings. Several buildings have beenbuilt in recent years using a space frame system of the ’flat double layer’type (Fig 4.19), although other types are also available.

The success of space structures, as in all structures, greatly depends onthe use of an efficient jointingmethod (or connection). In Australia there areseveral proprietary joints readily available and a full discussion of spaceframe systems may be found in Refs 4.5 and 4.6.

The inherent economy of space structures lies in the fact that the frame ismade up of a large number of similar elements which can be fabricated in amass production operation. The erection of the frame can be oftenaccomplished by assembling the frame on site at ground level and jackingit into position on the column supports.

From an overall economy point of view, however, space frames should beconsidered only for applications where extremely large clear spans arerequired to satisfy building function. They may be selected for otherapplications purely for architectural reasons.

Fig 4.19 The basic square grid double layered space frame

4.4. Heavy Industrial Structures

These structures can be considered as almost entirely custom designed tofulfil the function demanded of the engineering or manufacturing processinvolved. It is therefore most important that the designer adopt arationalised approach to member selection and standardized connectiondetails in order to achieve the most economic frame within the functionalconstraints.In structures such as steel--mill buildings or power stations, the membersare often massive in comparison with normal building structures andcertain considerations assume greater importance.

45

4.4.1. ERECTION

The proposed method and sequence of erection should be considered atthe preliminary design stage.

The columns in such structures are often of very stiff box--sectionwith fixedbases and it is obviously not possible to ’spring’ such a column during theerection of a girder. The girder--to--column connection must be selected topermit easy placing of the girder between columns and ready access tocomplete the connection fastening. End plate connections are usually notpreferred in cases such as these, since the need to fabricate girders shortand subsequently shim on site adds greatly to the final cost of the erectedstructural work. Web side plate or angle cleat connections, on the otherhand, provide flexibility in fabrication and erection tolerances, andgenerally will bemore economic for simple flexible connections in industrialstructures. Connections are discussed in more detail in Clause 8.6.

4.4.2. SITE WELDING

In heavy industrial structures there is usually a great number of largeconnections each involving a considerable amount of site work. In thesecircumstances it may be worthwhile considering field welded connections.This is because the cost of establishing welding equipment on the job, andof moving it around, can readily be spread over the total amount of work togive an economic result (see Section 7).

4.4.3. BOLTED CONNECTIONS

Although the general rule for economy is to design bolted connections withthreads included in the shear plane, this may not apply in projects with a

predominance of large connections -- for example 50 or more bolts perconnection.

For these connections significant savings in the number of bolts (andtherefore in the physical size of the details, the number of holes to be drilledand the time needed for erection) can often be made by designing for’threads excluded’ (see Clause 6.4.4).

4.4.4. FUNCTIONAL CONSTRAINTS

In large process plants and similar structures it is sometimes impractical toadhere to all the guidelines for economy in fabricated steelwork. Forexample the need to accommodate a variety of machinery, equipment andservices can make it difficult to maintain uniform column spacings or torationalise on a single floor beam size. Likewise bracing can often presenta problem, and may have to be fitted in by the designer.

While these departures from optimum practice may be unavoidable, thedesigner should nevertheless maintain an overall philosophy of:

• Simplicity -- keep the number of members down to a minimum tosatisfy the structural and functional requirements.

• Standardization -- use as many beams and columns of the samesize and mass as possible; standardize the connections used.

• Symmetry -- although in these custom designed structures it isoften difficult, it should be remembered that connection selectionand bracing disposition can lead to symmetry in members andlayout. Obvious economy will be gained by providing forrepetition in the fabrication shop.

46

5. Commercial Buildings

5.1. Introduction

In contrast to the industrial structures discussed in Section 4, where thecriterion controlling the framing arrangement was often building function,the commercial or office type building is usually of a more regular layout. Itis this characteristic which allows the greatest economy to be obtainedthrough standardization and repetition of structural elements andconnections.

This category of steel building comprises a grid of steel beams connectedto steel columns (or composite columns or concrete shear walls) usingeither simple or rigid connections. Resistance to lateral loads may beprovided by using some form of bracing with steel elements or other typessuch as in--fill walls or shear walls, or by frame action using rigidconnections.

This type of building can be divided into two categories:

(a) Low--Rise Commercial -- e.g. suburban office blocks of up to 4storeys, schools, shopping centres, etc.

(b) High--Rise Commercial -- e.g. city office buildings, hospitals.

5.2. Low--Rise Commercial Buildings

This category can be further sub--divided into:

1. Fully steel--framed structures.

2. Composite frames (steel frames connected to concrete cores orutilising masonry in--fill panels).

5.2.1. FULLY STEEL--FRAMED

Low--rise buildings fully framed in steel offer advantages in building speedand therefore in the overall economyof the final building. Because low--risebuildings do not require large stabilising elements, a steel frame using onlysimple connections can be used, offering economy in both fabrication anderection. The stabilising element is usually provided in the form of a steelcross--bracing system in one or two directionswhich canbe incorporated ina facade treatment so as not to intrude into window openings.

Another framing system which has been used successfully for low--risebuildings is the one--way--rigid, one--way--braced system (see Fig 5.1).

Fig 5.1 Framing system for low--rise commercial building

This is essentially an extension of the industrial portal frame structure andresults in an economic solution for small commercial buildings wherefreedom of layout and planning can be provided across the building widthsince no internal columns or bracing elements are necessary.

In the design of such a building, it should be recognised that bays of equalsize will assist in gainingmaximum economy by allowing the repetitive useof similar sized beam and column sections. The economic detailing of

47

beams and columns is most important in achieving overall economy andaspects of this are contained in Section 8.

Undoubtedly the greatest advantage of a fully steel framed structure lies inthe ability to erect the entire structural framework on prepared footings, asa self sustaining system before any other building trades are required onsite. With proper planning, this feature can lead to faster building speedand the elimination of many of the problems associatedwith diverse tradeson site simultaneously.

5.2.2. COMPOSITE FRAMES

Currently a favoured type of construction for steel low--rise commercialbuildings is the provision of a stabilising element comprising a masonry orreinforced concrete core, with the steel floor beams connected with simpleconnections between periphery steel columns and the concrete core. Forthe low--rise commercial building, it is also common to use in--fill masonrypanels to provide lateral stability. Examples of these systems are shown inFig 3.7.

Typical details of such a framing arrangement are shown in Fig 5.2 for thecasewheremasonry panels are used to provide the stabilising element in abuilding frame.

Fig 5.2 Stability by masonry

For the case shown in Fig 5.2 it should be remembered that the steel framemust be effectively temporarily braced during erection and properlyplumbed before the brickwork or blockwork can be laid. If the temporarybracing has to be removed after stability is provided by the infill panels itcould be placed on the inner flange of the columns in order to facilitate laterremoval and in order not to interfere unduly with the masonry work.

Fig 5.3 Stability by concrete panels

Fig 5.3 shows an alternative method of providing a stabilising element inthe form of a concrete panel cast between two adjacent steel columns andtied into each. In this case, the wall thus produced would normally beconsidered as load--bearing and would support stair--landings etc.,throughout the height of the building.

In addition to the concept of composite frames, the use of compositebeam--slab systems will provide best economy in these buildings. This isdiscussed in Clause 5.5.

5.3. High--Rise Commercial Buildings

5.3.1. GENERAL

In Australia at present a high--rise commercial building will usually be a cityoffice block of up to 50 floors. In these buildings, a regular column grid canbe established resulting in repetitive bays in one or both directions. Aspreviously mentioned, regularity of bays is important since it leads tomaximum economy due to repetition.

The architectural and aesthetic requirements usually control the exteriorcolumn spacing and therefore the bay sizes. A panel wall design withcolumns contained within the wall thickness allows maximum freedom inbay size selection, whereas when columns are exposed externally as an

48

architectural feature this results in the least flexibility in bay size selection.Bay sizes should be selected to produce minimum storey height. It isnoteworthy that a saving of 75mmper floor in a 20 storey building will save1500mmof exterior and interiorwall, partitioning, columns, lifts, etc.On theother hand, columns cannot be spaced so closely as to detract from theusefulness of the space through which they pass. Selection of bay sizes isalways a compromise between these two considerations.

In a way similar to low--rise commercial buildings, high--rise commercialbuildings can be sub--divided into:

(a) Fully steel--framed structures, and

(b) Steel frames connected to reinforced concrete cores.

In the selection of the best framing system the most importantconsideration is to find a structural form which is highly efficient underlateral loadings and which does not require an unreasonable premium inframe cost to resist those forces.

A vast number of alternative steel framing systems have been successfullyused in the past, but not all of these are economic under today’s conditions.Fig 5.4 shows some of the frame types suitable for buildings of variousheights.

49

Fig 5.4 -- Optimum steel framing systems for buildings of various heights

50

5.3.2. FULLY RIGID FRAME

From a planning and layout point of view this system obviously createsmaximum freedomsince no stabilising elements are required in the verticalplanes of the building framework.

The system is suitable for buildings up to 30 storeys in height but should beconsidered only when constraints of planning and layout are unavoidable.

It has the advantage of allowing efficient use of material because of theconsiderable interaction between beams and columns due to the use ofrigid connections with resultant continuity in beams. However, in today’ssituation, rigid connections are more costly to fabricate and this will oftenoffset any savings in material. In addition columns will generally be moreexpensive because equal stiffness about both axes is required.

In the USA where frames of this type have been in use for many years, thebasic method was to erect columns and field--weld beams at floor levels --see Fig 5.5.

Fig 5.5. Field welded connection details

However, since this method required the field welding of the most criticaljoints in the structure, where both high quality welds and high constructionspeed was required (both being subject to weather and operator skill), thismethod has been refined by transferring the welding operation from thefield back into the shop. This is accomplished by using the ’Christmas Tree’concept as shown in Figs 5.6 and 7.9.

In view of the relative costs of shop and field welding, the stub girder shopwelded to the column will generally prove a more economic solution forrigid framework.

5.3.3. FULLY BRACED FRAMES

Fully braced frames of the typementioned below are ’braced tubes’ wherestability against lateral forces is provided by the braced action of theexternal building wall framing.

Fig 5.6 Shop welded connection details

Bracing across full building width

If the total facadewidth of the building can be considered as a vertical truss,the resulting frame offers maximum stability against lateral forces and thissystem can be used for almost unlimited storey height.

The advantages of braced frames lie in the use of simple flexibleconnections throughout and these are themost economical to fabricate. Inaddition, smaller columns can be used, often merely rolled sections. Thefloor beams on the other hand will tend to be heavier because no beamcontinuity is available but this mass addition will almost always be morethan compensated by the less costly fabrication required.

51

Bracing by shear truss in external walls

For buildings up to 50 storeys a shear truss in theplane of the externalwallsprovides good stability characteristics and has the advantage of notintruding into facade treatment as much as the full width bracingmentioned.

Architecturally, cross bracing has never been readily accepted. Someexceptions to this do exist overseas and in Australia, but in generalengineers are expected by their architect to conceal bracing in buildingfacades, and this can often be done by accepting a compromise betweenthe space possible in a bay opening and the bending induced in floorbeams-- see Fig 5.7.

Fig 5.7 Forms of bracing

Fig 5.8 Bracing should connect to column

5.3.4. STABILITY BY MEANS OF SERVICE CORES

Since building structures of the type under discussion invariably require a’Core’ in which are contained lifts, stairs, service ducts etc., it is convenientto consider the core as a major stabilising element to resist lateral forces.The floor beams are ’simply’ connected between steel periphery columnsand the core structure, with resultant economies in fabrication anderection.

Steel framed service core

A fully braced core structure using steel elements can be erected veryquickly as a free standing structure and provides convenient access to alllevels of the building throughout the construction phase.

Bracing can normally be placed to accommodate the necessary openingsand provide adequate stabilising function for buildings up to 50 storeys.

Slip--formed concrete core

Development of efficient slip--forming techniques has resulted in theconstruction of concrete cores becoming a fast, economic buildingprocess. Because such a central core is essential to house buildingservices such as lifts, stairs, ducting, etc., it is logical to consider using thestrong core as themajor stabilising element for amulti--storey building (seeFig 5.9). This system has been successfully used in many recent buildingsconstructed in Australia and overseas.

Using thismethod of stabilising the frame, the lateral forces on the externalwalls of the building are transmitted to the core through the floors. The floor,which usually consists of a concrete slab acting compositely with its steelsupporting beams (see Clause 5.4), is considered as a deep horizontaldiaphragm, and is extremely effective in transmitting lateral forces to thecentral core.

The position of the concrete core within the building has a significant effecton its structural behaviour under lateral loads. If the core is asymmetrical,rotation in addition to translation will be generated under lateral loads. Thisis an important consideration when the core is situated at the extreme endof a rectangular shaped building -- see Fig 5.10.

52

Fig 5.9 Service core

Fig 5.10 Service core at end of building

In such a case, it is often necessary to employ the use of an auxiliary steelbracing system in the end wall remote from the core. Thus the stability ofthe building in the direction shown is shared by the core and the bracingsystem.

In general, when building structures using concrete cores as stabilisingelements, connections of steel beams to periphery columns andconnections of floor beams to floor beams can be of the flexible type. The

connection of the floor beam to the concrete core must also be executedeconomically and methods of making such connections are discussed inSection 8.

Table 5.1 summarises situations where the use of shear walls or cores areadvantageous, and also lists situations where steel lattice bracing may bemore appropriate.

TABLE 5.1 SHEAR WALL vs LATTICE BRACING

5.4. Floor Support Systems

Supporting members, suitable for use in floor systems for steel--framedcommercial buildings include the following (see Fig 5.11):

• Universal sections (UB)

• Welded Beams (WB) or plate girders

• Hybrid girders

• Castellated girders.

Universal sections are in general use in steel framed construction, exceptwhere long spans and/or heavy loads necessitate the use of largermembers. The universal beamsections cover a reasonable range of spansand loading conditions, and are best suited for use as main or secondarybeams. Cover plates can be welded to the flanges to increase capacity butit is usually more economic to use a standard welded I--Section. Electricalservices and airconditioning ducts can penetrate through the web to avoidadding to overall floor depth. Simple and economic detailing of suchopenings is essential (Section 8 contains suggested details). Universaland standard welded I--Sections require little fabrication except at thebeam--to--column or beam--to--beam connections.

Non--standard welded Beams or plate girders cater for larger spans andheavier loads than universal sections. The flange plates are normally filletwelded to a single web plate. However, box girders can be fabricated usingtwo web plates where very heavy loads are involved. Like universalsections, plate girders can haveweb holes to enable the electrical servicesand airconditioning to pass through. Economic fabrication of thesemembers is possible using automatic submerged arc welding (see Section7).

53

Hybrid girders are plate girders using a stronger grade of steel on thetension flange of the beam and possibly part of the web. One economicalway of fabrication is to cut two universal sections of different gradessymmetrically and reweld them with a central web butt weld. The beamsmay be made castellated or can have a solid web. These girders areparticularly suitedwhere the beam is to bemade compositewith a concretefloor slab, but have been rarely used in Australia.

The profiled cutting and rewelding of a universal section to form acastellated girder containing web openings results in a girder which isdeeper, stronger and stiffer than the original section. The web openingscan be used for ducts and piping. Consequently, castellated girders canpermit a reduction in the overall mass of the floor system, leading tosavings in total building cost. The savings in material must, however, beconsidered against the increased cost of fabrication with this type of girder.Computer numerically controlled (CNC) cutting and welding equipmenthas improved the economic viability of castellated beams.

Further discussion of these beam types is contained in Section 8.

One other type of floor support system deserves some mention -- the stubgirder system. This is a novel system reported from its use in the UnitedStates to increase the economic span limit for steel beams while, at thesame time, providing space for mechanical ducts without any increase infloor height. The system consists of short lengths (stubs) of rolled sectionwelded at intervals to the top flanges of the girder, the stub beams beingspaced at intervals which allow secondary floor beams and ducts to passbetween the floor slab and the girder.

Fig 5.11 Floor support members

54

5.5. Composite Construction

The current trend to steel framing for commercial buildings has been due toa large extent to the development of composite construction techniques.This concept is based on designing a structure to rely on some degree ofinteraction between elements of differentmaterials. The economical use ofmaterials should be the keynote in all modern building design. Compositesteel--concrete construction in slabs, beams and columns, using both steeland concrete tomaximumadvantage, is one of themost effectivemeans ofachieving this objective.

5.5.1. FLOOR SYSTEMS

In composite structural framing the term composite steel beam refers to afloor system comprising a steel beam acting with a concrete slabcomponent on its top flange, interconnected to the slab such that both forman integral unit. The principal advantage of this lies in the fact that theconcrete slab not only spans between and distributes the loads to themainbeams but also forms part of the beams themselves (Fig 5.12).

Fig 5.12 Composite floor beam system

In the types of composite beam--slab systems in this discussion, theconcrete slab can be constructed in several ways. One of the bestnowadays is to cast it on profiled steel sheeting -- the sheeting serving aspermanent formwork when the slab is poured. The method of achievingcomposite beam action involves the provision of some form of mechanical

connection between the beam and slab at the interface. These elementsare known as shear connectors, of which the most economic type is thewelded stud (see Fig 5.13).

Fig 5.13 Welded stud shear connector

A conventionally formed slab system could be used as an alternative, butrising costs of the removable formwork material and the associated labourare making steel decking systems more attractive. In addition, theprovision of extensive propping to the underside of formwork and the timedelay in its removal mean that following trades are hindered in proceeding,thus negating the advantage of steel’s fast construction.

If steel decking is to be used it is probably better to use a typewhichwill alsoact compositely with the slab by becoming the positive reinforcement.Several forms of composite steel decking are currently available inAustralia and are made from high--strength zinc--coated steel. Typicalprofiles are shown in Fig 5.14.

Fig. 5.14 Profiles of composite galvanized steel decking

55

The use of composite steel decking provides for double economy. Firstly, itprovides a low cost and efficient floor slab by eliminating the need for all ormost of the lower reinforcement. Secondly, it has the benefits of permanentformwork such as speedy installation, a weather and safety cover, and animmediate working platform for other trades.

Steel decking is used to its optimum advantage in steel framed buildingsbecause full advantage can be taken of sheet continuity to increase slabload capacity and because the resultant slab can also be made compositewith the steel beams. This means that composite action is achieved in twoways:

(a) Within the slab, and

(b) Between beam and slab.

Design methods for composite floors are readily available (see Refs 5.3,5.4, and 5.5).

5.5.2. COLUMNS

The concrete encased steel column is a further example of compositeaction. Encasing of columns is often required to satisfy the architecturalfeatures of building facades and to provide fire protection to the steelcolumn. The opportunity exists to consider a relatively small steel columnsection, designed to carry construction loadings, which can besubsequently encased and, as a composite section, designed to carry totalvertical loading. The steel column can be used as reinforcement in the finalcomposite column, or, where a larger final section is required additionalreinforcement can be introduced (see Fig 5.15).

Fig 5.15 Composite columns incorporating a steel erection column

By proceeding in this way the erection of the structural frame is notcontrolled by the time taken for the forming, pouring and curing the finalshape of a wholly concrete column. The steel column can be designed tosupport say 6 to 10 floors of structure, and the building program is plannedso that the encasement of the lower columns becomes a relativelynon--critical item in the construction sequence.

The converse of a concrete encased steel column is a steel tubular columnfilled with concrete, which also provides composite action.

Small or medium sized columnsmight be RHS or CHS; larger columns arebox or tubular sections fabricated from steel plates -- see Fig 5.16.

These tubular composite columnsmake for quick and easy erection and ofcourse they eliminate the need for concrete formwork. In the larger sizestheir overall economy depends upon the ability of the fabricator tomanufacture the tubular sections efficiently.

56

Fig 5.16 Composite column comprising a concrete--filled tubularsection

5.6. Summary

From a technological point of view, the design of commercial buildings isrelatively well understood. However, in today’s scene the important point to

remember is that such buildings, in order to be viable business ventures,require to be constructed with maximum economy of time, materials andlabour.

Many city buildings in Australia in recent years have been constructedusing the steel frame to concrete core method and it is apparent that thissystem is proving economic in the current situation. High on--site labourcosts are causing a return to the principle of prefabricating buildingelements off--site and then simply assembling them to form a buildingstructure. As tall buildings, by virtue of their large number of identical floors,require a vast number of repetitive structural members, it is in thesestructures that economy can be achieved by the adoption of rationalisedmember design and standardization of connections. Steel beams whichconnect periphery columns to a central core and carry the slab on steelsheet decking (composite with the steel beams) will usually prove a mosteconomic solution in commercial buildings.

When assessing different structural systems, designers should becognizant of the relative cost components (see Fig 5.17) to enable a morerational approach to the framing system.

Steel Deck Supply& Fix = 21%

Slab = 23%Steel Supply= 31%

Fabrication= 8%Surface Treatment = 13% Steel Erection = 4 %

Fig 5.17 Cost Components for a Multi--Storey Building

57

6. Bolting

6.1. Introduction

The selection of a bolt for use in a structural steelwork connection will needto have regard to a variety of factors including:

(a) Load capacity of available bolt types

(b) Cost of the installed fastener

(c) Amount of joint slippage

(d) Nature of the forces to be resisted

(e) Degree of flexibility/rigidity desired in the joint;

in order to obtain, at least cost, a safe bolted connection.

’Bolting of Steel Structures’ (Ref 6.1) contains a detailed discussion of all ofthe above factors and provides a state--of--the--art summary of mattersrelated to the use of bolts in steel structures.

This section concentrates on aspects which affect the economic use ofbolts. Ref 6.1 should be consulted for more details of all aspects of the useof bolts in steel structures.

The cost of a bolted connection includes:

1. Cost of obtaining, cutting and holing components

2. Cost of the bolts

3. Cost of installing the bolts

4. Cost of inspection.

Every bolt specified should be a bolt that is needed -- bolt numbers shouldbe kept to the minimum needed from strength considerations.

The cost of installing bolts can vary considerably, depending on the boltingcategory.

6.2. Bolt Types

The two basic metric bolt types in use in structural engineering in Australiaare:

(a) the commercial (Strength Grade 4.6) bolt;

(b) the high--strength structural (Strength Grade 8.8) bolt.

The identification of high--strength structural bolt and nut assemblies canbe readily made from the bolt head and nut markings (see Ref 6.1). Inaddition, a distinguishing feature is the larger bolt head and nut of thehigh--strength structural bolt compared to the commercial bolt.

Only a limited range of sizes of these bolts is of interest to structuralengineers.

6.2.1. COMMERCIAL BOLTS

The commercial bolt is commonly used in the following diameters (theprefix M is used to designate ISO metric bolts):

M12 -- purlin and girt applications

M16 -- cleats, brackets (relatively lightly loaded)

M20, M24 -- general structural connections, holding down bolts

M30, M36 -- holding down bolts.

6.2.2. HIGH--STRENGTH STRUCTURAL BOLTS

The high--strength structural bolt is most commonly used in diameters:

M16 -- designed connections in small members:

M20, M24, M30, M36 -- flexible connections, rigid connections.Larger sizes (M30, M36) ofthe high--strength structural boltshould be avoided when full tensioning is required, since

58

on--site tensioning can be difficult and requires specialequipment to achieve the minimum bolt tensions.

6.3. Bolting Categories

In Australia a standard bolting category system has been adopted for useby designers and detailers. This system is summarised in Table 6.1.

Category 4.6/S refers to commercial bolts of Strength Grade 4.6conforming to AS 1111 tightened using a standard wrench to a ’snug--tight’condition.

Category 8.8/S refers to any bolt of Strength Grade 8.8, tightened using astandardwrench to a ’snug--tight’ condition in the samewayas for category

4.6/S. Essentially, these bolts are used as higher grade commercial bolts inorder to increase the capacity of certain connection types. In practice theywill normally be high--strength structural bolts of Grade 8.8 to AS 1252, butany other bolt of Grade 8.8 would be satisfactory.

Category 8.8/TF and 8.8/TB (or 8.8/T when referring generally to bothtypes) refer specifically to high--strength structural bolts of Strength Grade8.8 conforming to AS 1252, fully tensioned in a controlled manner to therequirements of AS 4100.

The system of category designation identifies the bolt being used by usingits strength grade designation (4.6 or 8.8) and identifies the installationprocedure by a supplementary letter (S -- snug; T -- full tensioning).

TABLE 6.1 BOLT TYPES AND BOLTING CATEGORIES

BoltingCategory

Method ofTightening

Nominal Bolt Tensile Strength(MPa)

Nominal Bolt Yield Strength(MPa)

Bolt Name StandardSpecification

4.6/S Snug 400 240 Commercial AS11118.8/S Snug 830 660 High Strength

StructuralAS1252

8.8/TF(Friction typejoint)

Full tensioning to AS4100 830 660 High StrengthStructural

AS1252

8.8/TB(Bearing typejoint)

Full tensioning to AS4100 830 660 High StrengthStructural

AS1252

For 8.8/T categories, the type of joint is identified by an additional letter (F --friction--type joint; B -- bearing--type joint).

As a consequence the high--strength structural bolt may be specified inthree ways:

(a) Snug--tightened-- category 8.8/S

(b) Fully tensioned, friction--type -- category 8.8/TF

(c) Fully tensioned, bearing--type -- category 8.8/TB;

the level of tensioning being, of course, the same for both 8.8/TF and8.8/TB categories.

Two symbols have been added to the bolting category designations 4.6/S,8.8/S, 8.8/TB.

N: bolt in shear with threads iNcluded in the shear plane (e.g.8.8 N/S).

X: bolt in shear with threads eXcluded from the shear plane(e.g. 8.8 X/S).

59

In practice 8.8/S category would mainly be used in flexible joints where theextra capacity of the stronger bolt (compared to 4.6/S category) makes iteconomical. It is recommended that 8.8/TF category be used only in rigidjoints where a no--slip joint is essential. Note also that 8.8/TF is the onlycategory requiring attention to the contact surfaces.

A summary of the usage of Grade 4.6 and 8.8 bolts is contained in Clauses6.5.2 and 6.5.3.

6.4. Factors Affecting Bolting Economy

6.4.1. BOLT GRADE

For a given diameter, and assuming snug--tight category, Grade 8.8 boltsoffer far better structural economy thanGrade 4.6. This is because aGrade8.8 bolt costs only around 30%more thanGrade 4.6, but has over twice theshear capacity; moreover the installation labour cost is the same for both.

TABLE 6.2 INDICATIVE COST RATIOS OF DIFFERENT BOLT DI-AMETERS

Bolt Diameter High--strength structural bolt (Grade 8.8) x100 mm long, with nut & hardened washer

Cost Index (supplyonly)

Cost Index per kNofshear capacity

M16

M20

M24

M30

M36

90

100

170

360

670

1.4

1.0

1.2

1.6

2.1

Notes: 1. The indicative cost ratios quoted are valid only within thistable

2. Shear capacity calculations are based on strength limit statedesign.

For this reason Grade 8.8 bolts are rapidly taking over as the standardgrade for structural engineering.

Of coursewhere fully tensioned categories are used;Grade 8.8 bolts to AS1252 are mandatory -- see Clause 6.4.3.

One application for Grade 4.6 is in foundation bolts, especially wherewelded cages are used.

6.4.2. BOLT DIAMETER

Bolts of M20 and M24 diameter represent an optimum in many respectssuch as: purchase price (see Table 6.2), hole drilling, and site installation.They should be preferred in all applications wherever possible.

Where special circumstances demand the choice of larger diameters (M30or M36) they should be specified with the knowledge that a cost premiumwill be involved.

M30 and M36 bolts are not recommended for applications requiring fulltensioning (8.8/TF or 8.8/TB) because it is difficult to obtain suitableportable equipment capable of inducing the high shank tensions requiredby AS 4100.

6.4.3. BOLTING CATEGORY

Table 6.3 shows that snug--tightened bolts of Grade 8.8 (i.e. 8.8/Scategory) offer the best value in terms of cost per kN of shear capacity. Thisis therefore the preferred bolting method.

Category 8.8/TB provides no greater structural capacity and wouldtherefore be used only where some other consideration warrants it. Aninstance iswhere connection behaviour depends on the rigidity afforded bytensioned bolts as in rigid portal frame construction. 8.8/TB category hasalso been used on bolted bridges where the tensioning is merely asafeguard against nuts working loose in service.

Category 8.8/TF (friction--type joint) offers the poorest economy of all theoptions on a cost per kilonewton basis (see Table 6.3). It should be usedonly in applicationswhere joint slippage cannot be tolerated. An example isa structure supporting vibrating machinery such as a coal washery.

60

6.4.4. THREADS IN OR OUT OF SHEAR PLANE

The plain shank area of a bolt is approximately 30% greater than the corearea at the threads. Thus an apparent gain of 30% in shear capacity isavailable if the threaded part of the bolt can be kept out of the joint shearplane.

However, this benefit can often be illusory, especially on averageconnections with up to only 10 or so bolts. Any savings in bolts must bemeasured against the cost of longer bolts required, possible installationproblems and the higher cost of supervision needed to ensure ’threadsout’.

On the other hand on major structures with joints of around 50 bolts ormore, a good case canbemade for basing the design on threads excluded.Savings accrue from fewer bolts, fewer holes, smaller gusset plates andreduced installation time, while there is usually already a high level ofsupervision on these large projects to ensure correct installation.

One final point to be borne in mind is that there is never a case forconsidering 4.6/S category with threads excluded, i.e. Category 4.6X/S. Itwill always be more economic to use Category 8.8N/S.

The topic of threads in vs. threads out is discussed inmore detail in Ref 6.1.

6.4.5. BOLT FINISH

The available choices are ’plain uncoated’, ’zinc coated’ or ’hot--dipgalvanized’. Galvanized bolts do not cost very much more than plain boltsand are now supplied as standard finish for Grade 8.8 bolts.

In general the bolt finish should be matched to that of the structure itself.Uncoated bolts are satisfactory in low corrosion environments; galvanizedbolts are needed where corrosion may be a consideration. They performbetter and are much less costly than site--painted bolts.

Care is needed when galvanized bolts are to be fully tensioned, althoughproper procedures and good housekeeping on site will obviate problems --see Ref 6.1.

Table 6.3 INDICATIVE COST RATIOS OF DIFFERENT BOLTINGCATEGORIES

(One M20 galvanized bolt installed in a group, ”threads included”)

BoltingCategory

ShearCapacity (kN)

Cost Index(installed)

Cost Index perkN of ShearCapacity

4.6/S

8.8/S

8.8/TB

44.6

92.6

92.6

80

100

200

1.66

1.00

2.00

Notes: 1. The indicative cost ratios quoted are valid only within thistable.

2. The above comparison is based on strength limit state. Sinceserviceability generally governs for 8.8/TF bolts, they havebeenexcluded from this table.

6.4.6. INSPECTION

Part of the cost of bolt installation is the necessary inspection. With 4.6/Sand 8.8/S categories such inspection is minimal, and requires only a visualcheck that the correct type and number of bolts have been installed. Sincethe level of tightening is only ’snug’, and this is achieved in the normalcourse of erection, no further checking is required.

In contrast fully tensioned bolts (8.8/TF and 8.8/TB categories) requiredetailed inspection in accordance with AS 4100 to confirm that thetensioning procedure has been carried out. The inspection cost is a bigcomponent of the total in--place cost of a bolt. Inspection procedures areoutlined in AS 4100 and are discussed in Ref 6.1.

61

6.5. Summary for Economic Bolting

6.5.1. CHECKLIST

The essential points to be considered in the economical design of boltedconnections are:

(a) Standardize as much as possible for a project.

(b) Adopt simple detailing.

(c) Only one bolt diameter and one bolting category should be usedin smaller structures; more variety may be justified on a largerstructure, but different diameters or categories should be used inaccordance with a predetermined philosophy.

(d) Only one nominal size of bolt should be used in any single con-nection to facilitate the operation of punching or drilling holes,regardless of the size of the structure.

(e) Arrange for a minimum number of field connections by makinglarge sub--assemblies in the shop.

(f) Bolts in double shear are markedly more efficient and thoughtshould always be given to arranging the connection details ac-cordingly if practicable. In some instances (e.g. flange splices)such an arrangement can be negated by increased erection diffi-culty.

(g) If possible, avoid bolted connections with more than 5 bolts inline parallel to the force, otherwise reduction in bolt efficiency willresult (see Ref 6.1).

(h) Try not to mix 8.8/S and 8.8/T bolting categories on the one job.

(i) For economy, it may appear desirable to exclude threads fromthe shear plane. However, practical reasons dictate that usuallythreads are considered included in the shear plane, unless detail-ing of the bolts indicates exclusion is possible (see Ref 6.1).

(j) Corrosion protection of the bolts should be matched to the enduse of the structure.

62

The following flow chart is for bolt usage with flexible joints:

6.5.2. BOLT USAGE -- FLEXIBLE JOINTS

SIMPLE (FLEXIBLE) JOINTS

Not calculated orvery low stress levels

Structural Joints

Commercial BoltsProperty Class 4.6 to AS1111 -- Snug tightened

Commercial BoltsProperty Class 4.6 to AS1111 -- Snug tightened

High Strength Structural BoltsProperty Class 8.8 to AS1252 -- Snug Tightened.

Category 4.6/S

Category 4.6/S Category 8.8/S

Threads in shear orbearing plane is mostcommon situation

Low capacity Twice capacity of 4.6/S

Threads included in shear planeThreads excluded from shear plane

No “stick through” problem Possible “stick--through” problem

Most realistic from erection viewpoint Difficult to inspect

Lower capacity (30% less)than threads excluded

Greater capacity than threads includedGENERALLY PREFERRED(See Clauses 6.4.4)

63

The following flow chart is for bolt usage with rigid joints:

6.5.3. BOLT USAGE--RIGID JOINTS

RIGID JOINTS

Friction Type

High Strength Structural Bolts Property Class 8.8 to AS 1252Fully tensioned to AS 4100 (Category 8.8/T)

Category 8.8 / TF

Threads permitted in shear plane

GENERALLY PREFERRED(See Clauses 6.4.4)

Category 8.8 / TB

Bearing Type

No slip

Lower Capacity than 8.8 / TB

Design for no slip in theserviceability limit state but alsocheck for strength limit state.

Slip occurs

Higher Capacity than 8.8 / TF

Threads includedin shear plane

Threads excludedfrom shear plane

No “stick--through” problem Possible “stick--through”problem

Most realistic from erection viewpoint

Difficult to inspect

Lower capacity (30% less)than threads excluded Maximum Capacity

64

7. Welding

7.1. Introduction

7.1.1. PRINCIPLES FOR ECONOMY

The aim of weld design should be to provide the necessary structuralperformance throughout the lifetime of the structure for the lowestcompleted cost. To achieve this attention must be given to:

(a) Eonomical design and detailing

(b) Good welding procedure and correct process selection, and

(c) Responsible inspection.

The design and detailing will greatly dictate whether or not an economicalwelded connection can be produced and consequently is one area wheregreat attention should be paid. Whereas the selection of the weldingprocedure and process to be used is the province of the fabricator, thedetailing of the welded connection can often influence or limit the range ofoptions available. Consequently, the design and detailing of the weldedconnection must have some regard to the processes and proceduresavailable if an economical welded connection is to result. Responsibleinspection is also a vital item in keeping the final cost to a minimum.

The design engineer can best approach the objective of obtaining, at leastcost, a safe welded steel structure or connection by considering thefollowing influences during the design:

1. Available welding processes that might be used

2. Welding consumable selection

3. Code requirements (AS 4100, AS 1554)

4. Joint details and type of weld

5. Size of weld

6. Whether to use shop or field welds

7. Accessibility

8. Responsible specification

9. Inspection.

7.1.2. COST COMPONENTS

The cost of welding can be considered as follows:

Cost of actualwelding =

Time to weldper unit length×

Length ofweld × Cost per

hour

OperatingFactor

where:

Time to weldper unit length=

Weld VolumeDeposition Rate

Operating Factor= Actual Arc TimeTotal Time

Cost per Hour= labour rate plus oncost

These relationships indicate that a designer or detailer can minimise thecost of welding by attention to the following items:

(a) Minimising weld volume

(b) Allowing for the use of high deposition rate processes; in someconnections, the detailing can restrict the use of a particular pro-cess thus forcing the fabricator to use a less efficient process

(c) Considering other factors which influence the deposition rate. Forexample, downhand welding is far more productive than over-

65

head or vertical welding, so that details should be oriented fordownhand welding wherever practicable

(d) Using clean and simple detailing to assist in maintaining as highan operating factor as possible

(e) Aiming to permit as much welding in the shop as possible, be-cause the cost per hour and the operating factor are both morefavourable in the shop than in the field

(f) Selecting the material grade to assist in eliminating or minimisingthe costs of preheating or post weld treatment.

7.2. Types of Welds

7.2.1. FILLET WELDS (see Fig 7.1)

The features of fillet welds are:

(a) Economically attractive up to 12--16 mm leg size

(b) Minimum edge preparation

(c) Easy fit--up without tight tolerances

(d) Poorer load carrying capacity than equivalent complete penetra-tion butt weld and poorer fatigue characteristics

(e) Intermittent fillet welds are permitted but these are usually onlyeconomical for limited applications involving the use of manual orsemi--automatic processes; in many applications, a full lengthfillet weld of one size may be placed more economically using afully or semi--automatic process

(f) In the horizontal--vee (HV) fillet position, up to 8 mm fillet sizesmay be placed in a single pass using manual metal arc pro-cesses; with other processes (semi--automatic or automatic) alarger single pass fillet weld is possible. Such processes are nowcommonly used.

(g) If more than a single pass fillet weld is used, the cost of the weldcan increase significantly.

TABLE 7.1 FILLET WELD COMPARISON

Fillet size (mm) Weld strengthrelative to 4mmsize

Weld area rela-tive to 4mmsize

Increase inweld strengthfor next size(%)

Increase inweld area fornext size (%)

4 1.00 1.00 25 565 1.25 1.56 20 446 1.50 2.25 33 788 2.00 4.00 25 5610 2.50 6.25 20 4412 3.00 9.00 33 7816 4.00 16.00

The cross--sectional area of a fillet weld varies as the square of the leg sizewhile the strength of a fillet weld (which is based on the effective throat)varies only linearly with the leg size. As indicated in Table 7.1, there is aheavy cost penalty in over--welding.

Automatic processes can reduce the cost of a fillet weld since, in addition toimproving productivity, the increased penetration allows a reduced leg sizefor the same throat thickness.

66

Cruciform

Fig 7.1 Types of fillet welds

(a) Complete penetration butt welds

(b) Partial penetration butt welds

Fig 7.2 Types of butt welds

7.2.2. BUTT WELDS (see Fig 7.2)

Two forms of butt weld are permitted in AS 1554 and AS 4100:

(a) Complete penetration -- used where the full strength of the con-nected parts is required. Such a joint is given the full strength ofthe joined components.

(b) Partial penetration -- used where less than full strength is accept-able, such as in low stress areas. These welds are less costly

than complete penetration, although attention is needed to en-sure that the specified depth of penetration is achieved in prac-tice. These welds are permitted to carry only shear and compres-sion loads and have low ratings for fatigue conditions.

Typical details of both types are shown in Fig 7.2.Butt welds usually require special edge preparation which (depending onthe preparation type and the cutting practice) can add to the cost. Types ofedge preparation normally in use are:

1. Square (no special preparation)2. Single or double bevel3. Single or double V4. Single or double J5. Single or double U.

When selecting joint preparations for butt welds, prequalified preparationsshould be used wherever possible to obviate the need for qualificationtesting of the weld geometry.In selecting the included angle in a butt weld preparation, it has beendemonstrated that, in general terms, the smaller the included angle in thepreparation the less is theweld volume (Ref 7.2). There is a need to temperthis provision with a consideration for leaving sufficient angle for electrodeaccess -- the requirements will vary between processes.It is therefore probably better for the design engineer to specify therequirements (e.g. ’complete penetration butt weld’ or ’partial penetrationbutt weld, depth of penetration 12mm’) and allow the fabricator to select thebest weld geometry/welding process combination to achieve the desiredresult. All such proposals can be submitted to the designer for approval ifnecessary.

7.2.3. BUTT WELDS vs. FILLET WELDS

It is important to note that the volume of weld metal in a butt weld (partialpenetration or complete penetration) depends on the type of preparationused as well as the depth of penetration. In contrast, the fillet weldincreases in weld volume as the square of the leg size.In comparing the relative costs of butt welds and fillet welds, these differingrelationships should be borne in mind, in addition to the fact that the buttweld usually requires edge preparation while the fillet weld does not.

67

The relative economics of the twowill depend on the application and on thefabricator’s equipment and methods, and it is quite feasible for individualfabricators to cost various sizes of both types and plot a graph which willlook something like Fig 7.3. The crossover point of weld size belowwhich afillet weld is the cheaper solution lies generally in the range 12--16 mm formany applications.

Fig 7.3 Weld cost graph

7.3. Welding Processes

The welding processes of interest in the welding of structural steel are:

(a) Manual metal arc (MMAW)

(b) Flux cored arc (FCAW)

(c) Gas shielded metal arc (GMAW)

(d) Submerged arc (SAW)

(e) Electroslag (ESW)

(f) Stud welding.

For efficient design, it is necessary to understand thebasic features of eachwelding process, to know its advantages and disadvantages and tounderstand the implication that the design can have on process selection,

since it is necessary that a design is realistic in terms of both weld cost andweld quality.

Manual metal arc welding (’stick electrode’ welding) is the simplest andmost flexible of all the processes and is suitable for welding in all positionsboth in the shop and in the field. However, it is capable of only lowdeposition rates and has an intrinsically poor productivity because of thestop--start nature of the process. It is gradually being superseded by moreefficient and economic continuous wire processes.

Flux cored arc welding employs a continuous hollow electrode whichcontains the flux. It is capable of relatively high deposition rates, is suitablefor all positions and in its gasless form is ideal for field welding.

Gas metal arc welding uses a continuous solid wire electrode shielded byinert gas. It too is a high productivity flexible process, and is replacingmanual metal arc welding in many fabrication shops.

Submerged arc welding is another continuous wire process, where the arcis submerged under a layer of flux. It is essentially a very high depositionmethod intended for automatic or semi--automatic set--ups in the shop;automatic machines for welding plate girders use this process. Somespecialised field applications have also been developed.

Electroslag welding is a special automatic process normally used by thelarger fabricators to butt weld plates. It is a single pass vertical process andis economic for plates 25 mm thick and above.

Stud welding uses special equipment for the attachment of shear studs tosteel members in composite construction. It is a portable process suitablefor field use, but can be readily adapted to an automatic or semi--automaticset--up in the shop.

These welding processes are described in greater detail in Ref 7.1.

There can be startling savings in the cost of welds produced by the moremodern processes. For example, considering a 6mmdownhand fillet weldmade bymanual welding using traditional rutile electrodes, the cost can behalved if iron powder electrodes are employed. This cost in turn can behalved again by adopting a suitable continuous wire process.

Thus the designer should take great care to avoid introducing unnecessarycosts in a job by restricting, through the details or the specification, the useof the optimum welding process.

68

7.4. Other Cost Factors

7.4.1. WELD CATEGORIES

The Structural Steel Welding Code, AS 1554 specifies two categories ofweld, these being:

GP -- General Purpose

SP -- Structural Purpose

The difference between the two arises from the more stringent quality andinspection requirements of the SP category over the GP category.

The Steel Structures Design Code AS 4100 has been used as thereference standard from which the permissible levels of imperfections forGP and SPwelds have been set. In other words, AS 1554 andAS 4100 arecompatible.

Category GP

The GP weld is the less stringent of the categories. It is intended for use injoints which are statically loaded, and where the design load on the weld issignificantly below its full design capacity. It should be noted that for GPCategory, the capacity factor is 0.6 as compared to a range of 0.70 to 0.90for the SP Category (see Table 3.4 of AS 4100 -- 1990).

Category SP

The SP category is the full--strength structural weld for use in staticapplications where the higher range of capacity factors is used. SPcategory is also mandatory for dynamic (fatigue) applications -- see AS4100 and AS 1554.

Choice of Weld Category

GP category welds will occur quite frequently in certain types ofapplication. The designer should always endeavour to specify GP weldcategory where appropriate in order that advantage may be taken of thelower production costs associatedwith it. Only under circumstanceswhereweld failure could cause a complete collapse of the structure or lead tosevere risk or loss of life, should a designer contemplate specifying as SP

category those welds which could otherwise, according to the guidelinesgiven in the Standard, be categorised as GP.

Mixing Weld Categories

Weld categories can bemixed on a project but should not bemixed along aweld. In Fig 7.4, for example, it would be quite in order in a weldedbeam--to--column moment connection to have SP weld category for theflange butt welds but either SP orGP for the fillet welds along theweb or forthe fillet welds along the column stiffeners.

Fig 7.4 Welded beam--to--column moment connection

The web--flange fillet welds in a three--plate girder (Fig 7.5) may havestress levels which vary along the beam such that an SP category weldmay be required at the ends of the beam, while GP category welds aresufficient elsewhere.Obviously, in this case anSP categoryweld should bespecified for the full length, but weld inspection should be concentrated atthe ends of the beam. If a length of weld which does not comply with the SPcategory was found in the central portion, it could still be accepted if itcomplied with GP category.

It would, however, be quite in order to specify GP category welds forintermediate web stiffeners or stiffening around a web penetration.

Fig 7.5 Stiffened web plate girder with web penetration

69

7.4.2. WELDING SPECIFICATIONS

It is essential that the drawings and specifications detail the functionalrequirements of the design clearly and concisely but avoid needless overdetailing or over specification of items which are better left to the fabricatoror erector. It is advisable to avoid generalising with such items as ’nounder--cut permitted’ or ’all welds to be smooth and free from defects’ or’weld all round’ as these too often lead to confusion and extra cost.

Flexibility in the approach to design is important particularly in consideringproposals for alternative welding details or procedures. The fabricator orerector may have alternative methods to improve productivity and reducecosts and these should not necessarily be excluded by a rigid specification.If tendering is involved, prices for the tender specification and for viablealternatives could be useful.

It is generally quite sufficient to nominate only the functional requirementsplus compliance with an appropriate welding code, such as AS 1554, forsatisfactory results. Standards are prepared for use as referencedocuments and it is not usually necessary to depart from them unless verygood reasons exist.

Where welding is specified in accordance with an Australian Standard, itshould be the one relevant to the service conditions, e.g. specifyingpressure vessel standards for a multi--storey office building is poor design.Fitness for service should be the sole criterion for the quality level specifiedand for the specification of the appropriate levels of inspection. Anydeparture from normal levels is likely to increase costs and should becalled for only when really required.

7.4.3. WELDING INSPECTION

Fabrication costs are very sensitive to the required weld quality and thetype and standard of inspection. Modern equipment and techniques forwelding and testing of welds make it possible to provide near perfectweldments if so required. However, this also adds considerably to the cost.If such standards are not necessary, the benefits previously gained bycareful economic design are frequently negated. It rests with the designengineer to determine the critical areas of a structure requiring closeinspection and then to set a realistic standard for the inspector to follow.

In setting guidelines for the inspector, the best results are achieved bynominating the use of the Structural Steel Welding Code, AS 1554. ThisStandard is well understood by both fabricators and inspectionorganisations and usually results in a good job being achieved at areasonable cost. A confusing and often expensive practice sometimesadopted is to rewrite some existing Standard clauses into the specificationin an attempt to achieve a higher standard than that provided by theStandard. This should be avoided because it usually leads to anomaliesand contractual problems.

Fitness for purpose should be the rule in setting inspection standards andAS 1554 provides realistic levels of both workmanship and inspectionsuited specifically for various weld quality levels required in structuralfabrication .

70

7.5. Economical Design and Detailing

The essential requirement of weld design is that adequate structuralperformance be provided. Usually a variety of alternative methods ofachieving this aim are available and the cost aspects of the alternativesneed to be looked at.

The principal considerations in economical detailing of weldments are:

(a) Simplicity -- details of welded attachments and details of end con-nections should be simple and consist of the fewest possiblenumber of component parts.

(b) Weld volume -- only the minimum required weld volume, as de-termined by structural calculations, should be specified.

(c) Accessibility -- welding electrodes must be able to be positionedin such a way that good quality welding can be achieved withoutdifficulty and without undue strain on the operator.

(d) Erection -- proper detailing should allow for reasonable fit--up tol-erances and weld preparations.

(e) Inspection -- all welds should be located in positions so that visu-al examination and/or nondestructive testing can be carried outeasily.

The following rules are suggested as basic to economical weld design anddetailing (see also Refs 7.2 and 7.3):

(1) Design with welding in mind.

This requires an appreciation of the cost components in welding, the typesof weld available, the types of processes and procedures available andtheir limitations.

(2) Do not specify oversize welds.

Themost cost effectiveweld is the smallest weld that provides the requiredstrength. It is good weld design practice to provide only that amount ofweldingwhich ensures that thewelded fabrication can perform its intendedfunction.

Specifying oversize welds can be harmful in two ways. Firstly, the cost isunnecessarily increased and secondly, oversize welds cause increasedshrinkage forces which may lead to distortion.

As an example, an 8mm fillet is only 33%stronger than a 6mm fillet, yet thevolume of weld metal is 78% higher (Table 7.1). Thus, the cost ofproduction of a joint can be significantly increased, not only due to theincreased volume of weld metal required but more importantly due to theincreased time in welding the joint.

The only qualifying point that should be raised is that the minimum weldsizes required by AS 1554 have to be observed and hence some oversizewelds may be unavoidable.

The ’weld all round’ philosophy should be avoided as it can lead tounnecessary additional cost.

(3) Use welding judiciously when using it to reduce material mass.

If welding is used to reduce the amount of material (e.g. by splicing tochange flange plate thicknesses or to provide stiffeners to a thin web in athree--plate girder), then be sure the cost of thewelding is less than the costsaving in material cost. Weld metal costs many times more than parentmaterial (somewhere from 50--100 times), and it is often cheaper toincrease component mass so as to reduce weld metal volume.

(4) Keep the number of pieces to be welded to the minimum practica-ble.

Asimple designwith the fewest number of pieces is themost economic andoften results in a better product.

(5) Remember the special effects of welding such as distortion (Ref7.2).

(6) Allow welding to be used to maximum advantage.

This particularly applies to allowing the fabricator to take advantage of highproduction processes, and in many cases may be best achieved byconsultation with the fabricator. The detailing of a weldment can oftenrestrict the fabricator to only the one process, and this may not always bethe most suitable.

(7) Aim for as much shop fabrication as possible.

(8) Keep in mind the economics of fillet welding (Clause 7.2.1).

71

Fillet welds are usually limited to 6mm leg size formost processes (notablymanual metal arc), although with other processes, under certainconditions, a 10 mm or larger single pass fillet weld is possible; (forexample a 20 mm single pass fillet weld is possible using tandemsubmerged arc welding but such processes are not commonly used whenwelding short runs on most simple connections). Before specifying largefillet welds, the situation should be checked with the fabricator. Largersingle pass fillet welds can be placed in the flat natural vee position. If morethan a single pass is required, the cost of the weld increases significantly.

Single run continuous fillet welds are usually more economic thanintermittent fillet welds of a larger size.

(9) Keep in mind the economics of butt welding (Clause 7.2.2).

Complete penetration welds need only be specified when they are reallyrequired, and the use of partial penetration welds can reduce weld metaland give other gains which add up to an improvement in productivity. Ifcomplete penetration welds are demanded, the use of backing bars withwelds from one side which do not need back gouging or turning of the workpiece, may lead to improvement.

If selecting joint preparations, use prequalified preparations (AS 1554) toavoid qualification testing.

Select the smallest included angle consistent with achieving the desiredpenetration. Better still, specify only, say, ’complete penetration butt weld’(or specify acceptable alternative details) on the drawing and allow thefabricator to select the method he can do best and most economically.

(10) Use fillets in preference to butt welds wherever possible.

Butt welds usually involve edge preparation, which adds to costs, and as aresult fillet welds are cheaper than butt welds up to about 16 mm thicknessof connected plates. (Other considerations, such as joints which may besubjected to fatigue, may dictate the use of a butt weld in preference to aless costly fillet weld.)

(11) Provide adequate access.

Another way the designer can significantly help productivity is to ensureadequate access for welding. This is vital as it is essential to ensure alwaysthat the appropriate quality of weld can be made.

Examples of badaccessibility -- togetherwith suggested improvements areshown in Fig 7.6.

(a) Gussets too close to flanges

(b) Angle seats too tight against flanges

(c) Correction: Use butt weld in lieu of fillet

(d) Correction: Use larger channel (e)

72

(f) Column stiffener details

Fig 7.6 Some common detailing faults resulting in pooraccessibility for welding

(12) Consider the method of fabrication.

Allow welds to be made in the downhand position wherever practicable.This can often be achieved by the fabricator using special jigs andpositioners.

Always try to aid fabrication by designing to allow the maximum use of jigsand positioners -- certainly try to make designs so that their use is nothampered.

(13) Avoid dictating the manner of making a welded joint.

The fabricator knows the best joint preparation and welding procedure forease, economy and quality of joint using the facilities available. Thedesigner who details the fabrication method must accept responsibility forany fabrication problems and extra cost.

Ensuring the method of fabrication is acceptable can be achieved bycalling for compliance with a recognised Code or Standard (AS 1554) andrequiring the proposed fabrication and welding procedure to be submittedfor concurrence on important jobs.

Fig 7.7 Use of bending to reduce welding and give clean corners

(14) Be receptive to alternative proposals.

Be prepared to accept alternative welded joints/details proposed by thefabricator which have clear advantages.

(15) Recognise the value of consultation with the fabricator.

(16) Use minimum number of joints by:

(a) Using largest size of plate/section available consistent with thefunctional requirements

(b) Bending or forming in place of welding (Fig 7.7)

(c) Considering the use of castings, forgings in lieu of complexwelded joints

(d) Avoiding excessive detail to reduce parent metal mass -- seeitem (3) and Fig 7.8.

73

Fig 7.8 Beam flange with many different plate thicknesses -- avoidwhen steel mass saved is less than 100 times mass of weld metalrequired

(17) Standardise joint details as much as practicable to reduce vari-ety.

Different sizedwelds at a joint will require changes in current and electrodesize by the operator. This causes lost time and a drop in the operatingfactor. Aim to have the minimum variety of weld sizes and types on amember or at a joint.

(18) Use sub--assemblies to give:

(a) Easier handling and positioning for downhand welding

(b) Better access for welding

(c) Less site welding and more shop welding (see Fig 7.9).Fig 7.9 Exterior column/spandrel sub--assemblies for Sears Tower,Chicago

(19) Use non--destructive testing judiciously.The use of non--destructive testing of welds is very disruptive to the flow ofwork and adds considerably to the cost of a structure. Much of this cost willbe avoided if non--destructive testing is restricted to critical joints andcarried out on a random basis only after careful development of weldprocedures. Modern welding Codes encourage this approach.(20) Test only where required.Testing of welders and weld procedures for each job is expensive. Wherepracticable, consideration should be given to accepting welders andprocedures approved by recognised authorities for other similar work.(21) Specify weld quality consistent with service requirements.

74

Fitness for purpose should be the guiding rule in specifying weld quality.Higher quality specified unnecessarily or for its own sake is wasteful andcostly (see Clause 7.4.2).

Specify tolerances to limits consistent with the purpose of the weld.Adequate tolerances are necessary in order to allow for ease of fit--up.

(22) Avoid, as far aspracticable, requiring turning ofmembers toweldon other side.

Examples are:

(a) Avoid putting stiffeners on both sides of a plate girder web

(b) Truss detailing which requires one side welding only (see Clause8.4)

(c) Angle seat to column flange connections -- a narrow seat in lieuof wide seat avoids turning the member (see Fig 7.10).

(a) (b)

Fig 7.10 Angle seat detail -- (a) preferable to (b)

(23) Avoid joints which create difficult welding procedures.

Joints which create difficult welding procedures, such as two round barsside by side, acute angle intersections, etc., should be avoided. Suchwelds prove time-- consuming and are of questionable quality (see Fig7.11).

Such joints also cause difficulties with any post--weld treatments,(deslagging, brushing, grinding and corrosion protection).

Fig 7.11 These joints are difficult to weld and the welds may be ofquestionable quality

(24) Consult ’Economic Design ofWeldments’ (Ref 7.3) for further ad-vice on ways to use welding effectively and economically.

75

8. Detailing for Economy

8.1. Detailing on Design Engineer’s Drawings

It is in the design office that the potential economy of any steel structure iseffectively determined. Judicious decisions on details at this stage canprovide for simple, economic methods to be used at the fabrication stage.

The designer is faced with the problem that a different fabrication anderection technique could be favoured by each individual fabricator likely totender for the project. It is a good idea at the outset for the designer to havesome preliminary discussions with likely fabricators to check on latesttechniques prevailing in the industry. From these discussions the designand detailing approach for the structure can be carried out with factorsinfluencing economics firmly in mind.

In the normal course of events a steel structure passes through severalseparate stages involving design, detailing, fabrication and erection. Withthis in mind, it is important for designers to remember that a minimum ofdesign detailing by them will assist towards economy, since the shopdetailer is then left free to make the most efficient use of the particularfabricator’s capabilities (Ref 2.12). The need for this flexibility is oftenoverlooked by designers in their anxiety to specify their requirements.

Such things asa fabricator’s ability to fabricate large sub--assemblies in theshop and subsequently transport to site and erect themwill obviously havea bearing on the design of connection types and therefore on the economyof the overall project. In this regard it must be stressed that a maximum ofwork done in the shop will almost always produce better quality and moreeconomical structures.

In the presentation of working drawings therefore, the basic key is’communications’ which normally take place through a chain as illustratedin Fig 8.1.

Fig 8.1 Chain of communication

The processes involved in the design can be summarised in the followingsequence:(a) Initial communication(b) Structural concept including consideration of connection types(c) Integrated design(d) Connection detailing(e) Framing plans.

The Engineer’s structural framing plans must contain all the necessaryinformation to enable the fabricator to have shop drawings prepared for theindividual members, as well as the marking plans to identify each memberfor the erection phase.The following discussion is intended to highlight aspects of the detailing ofboth members and connections to achieve economy in the overallfabrication and erection of structural elements.

76

As an additional consideration the use of AISC Standardized StructuralConnections (Ref 1) will enable designers to specify standardizedconnections directly from the publicationwithout detailing, and if necessarypermit alternatives to be offered by the fabricator with the confidence ofassured design capacity and behaviour.

8.2. Beams

8.2.1. GENERAL

The simplest and therefore the most economic beams in structures will beof rolled universal sections. Wherever possible, it will almost always provemore economic in one--off types of steel structures to use a universalsection or welded beam section as a beam, even if a heavier solutionresults. The alternative fabrication of a three--plate girder introduces platepreparation, assembly and welding, the costs of which will generallyexceed the cost of additional material in the rolled universal section orstandard WB section, unless a vast amount of repetition is required.

8.2.2. PLATED SECTIONS

Where headroom limitations apply (distance from ceiling soffit to floorlevel), it may be necessary to consider plating a universal section of alimited depth instead of choosing a deeper beam. Here, the extra cost ofsupplying plates, assembling and welding causes the cost of the memberto rise, and a plated solution should only be used when a net saving in costresults compared to other feasible alternatives.

Attention to the detailing of the member will assist in keeping fabricationcosts down. For example, selecting cover plate widths as shown in Fig 8.2

will allow thewelding of both plates to the beam tobe done in the downhandposition without the need to turn the member during fabrication.

Fig 8.2 Plated sections

8.2.3. WEB PENETRATIONS IN BEAMS

Holes cut in the webs of beams to provide access for service ducts haveproved to be very costly in the past due to uneconomic detailing. This is dueto the fact that, traditionally, these openings have been compensated for bythe provision of extensive stiffening systems around the openings (see Fig8.3(a)).The position of such openings in the beam length obviously has a majoreffect on the degree of stiffening required -- openings near the centre ofuniformly loaded beams will require little or no stiffening, while openingsplaced near the supportsmay require stiffening. Anearly dialoguebetweenthe structural engineer and the building services designer can lead toducting being located in a favourable position structurally without detrimentto service requirements.Plain circular openings as shown in Fig 8.3(d) obviously represent themosteconomic solution. These can be cut by automatic means and result inminimum additional fabrication costs. If additional stiffening is required forround holes, it is most economic to use a pipe piece, fillet welded to thebeam web (see Fig 8.3(c)).

77

Fig 8.3 Web penetrations in beams (in descending order of cost,(d) being least costly)

Where rectangular holes cannot be avoided and stiffening is necessary,this can be economically accomplished by a web hole with half--pipecuttings and make--up plates or, alternatively, simply reinforcing the beamweb using square edge flat bars fillet welded to one side of the beam webas shown in Fig 8.3(b).

By judicious planning, the duct penetrations required in beams should beselected in position, size and shape to gain maximum economy in thefabrication of such beams.

8.2.4. CASTELLATED BEAMS

Castellated beams are fabricated by cutting a profiled line in the web of auniversal beam -- Fig 8.4. Circular profiles in lieu of the hexagonal profilesare also available from fabricators using computer controlled fabricationequipment. The beam halves are then offset longitudinally and the partwebs welded on member centreline.

Dc= overall depth of castellated sectionDn= nominal depth of original sectionD= actual depth of original section

Dc= D+Dn2

Example:Original section = 530UB82

Dn= 530 D= 528Dc= 528+ 265= 793

Fig 8.4 Typical castellated beam geometry

78

Design paramters

Span 7m full restraint, Grade 300 steelW* = 900 kN

Rolled Section Solution

610UB113, Mass = 113 kg/m

Castellated Beam Solution

800CUB82 cut from 530UB82Mass = 82 kg/m

Comparison of cost indices:

Rolled section (610UB113) 1.00Castellated Beam using CNC equipment 1.02Castellated beam w/o CNC equipment 1.39

Conclusion

Rolled section is more economic in this solution in this instance, althoughusing CNC (Computer Numeric Control) Equipment could be just aseconomical.

Each individual situation should be readily assessed based on usingupdated cost information.

Fig 8.5 Evaluation of economics of castellated beam

The use of castellated beams in steel structures is often seen as amethodof increasing beam strength while using the same mass of material. Whilemany instances have been reported where savings have been effected, it

must again be remembered that a fabrication cost has been introducedwhich could be larger than the saving made in material cost -- dependingupon the quantities required and the methods used.

The cost involved for this additional fabrication varies depending on theequipment available within individual fabrication shops. In some cases,problems can be encountered with distortion of the beam during cuttingthus requiring subsequent straightening of the members and addingfurther to the cost. In general, most fabricating shops are nowwell--equipped to undertake the fabrication of castellated beams, butdesigners should carefully investigate the relative cost differences with theindustry before specifying this type of section.

In the example shown in Fig 8.5 the heavier 610UB113 would be moreeconomic than the castellated 530UB82. This example highlights the needto consider each case on its merits by applying up--to--date cost data to theexamination of the alternative solutions.

8.2.5. THREE--PLATE GIRDERS

Where beams are required of greater depth than the largest universalbeam, consideration should be given to three--plate girders or thestandardized range of welded sections. These will most often offer moreeconomic solutions than trusses for such applications as floor supportingbeams. Three--plate girders are fabricated in modern automatic assemblyand welding machines using the submerged arc welding process.

In designing and detailing three--plate girders the following considerationsare important in achieving economy:

(a) Use flat bar or preferred plate widths and thicknesses for theflange and web plates

(b) Use edge trimmed plate of preferred width wherever possible forthe web plate to avoid additional cutting in the fabrication shop.This type of prepared plate can be fillet welded to the flange platewithout further preparation of the edge.

(c) When considering changing the flange width or thickness in orderto reduce mass, take account of the lengths of plate availableand whether continuation of an ’oversize’ plate is a more eco-nomical solution than introducing butt welded splices in the

79

flange plate. As a rule of thumb, it is probably economic tochange the flange thickness when:

Steel mass saved in flange > 100 x mass of weld metal required

Where lengths of girders are such that butt welded splices are necessary,locate the changes of flange plate size to suit the available lengths of plate.

(d) The cost increase for three plate girders with stiffened websagainst unstiffened webs is about 10--25%, depending on the de-tailing adopted. Consequently, when evaluating whether to use astiffened rather than an unstiffened web, the cost saving due tothe reduced mass of the web plate with a stiffened web must ex-ceed this cost differential, for the stiffened web solution to beeconomic.

(e) If using a vertically stiffened web, use one sided stiffeners toavoid having to turn the girder during fabrication (see Fig 8.6).Terminate intermediate stiffeners by the allowable ’4t’ from theflange (see AS 4100) -- this avoids cutting stiffeners accurately tolength (see Fig 8.6).

(f) Avoid the use of horizontal web stiffeners if at all possible.

The example shown in Fig 8.7 illustrates an evaluation of the relativeeconomics of stiffened vs. unstiffened webs in a typical three--plate girderapplication.

Fig 8.6 One--sided intermediate web stiffener

UNSTIFFENED WEB

Load bearing stiffeners in both cases 8mm & 12mm webs

Load bear-ingstiffeners inboth cases8mm & 12mm webs

Total Mass = 5.5 tonnesCost Ratio = 1.0Mass x Cost Ratio = 5.5

[ Cheaper Solution in this case ]

INTERMEDIATE STIFFENED WEB

Either:Stiffeners: 90x6 square edge flat bars, both sides, at 1500 mm centres (18off)Or:Stiffeners: 90x6 square edge flat bars, one side, at 1500mmcentres (9 off)

80

Total Mass = 5.0 tonnesCost Ratio = 1.25 for two sided (average)

= 1.15 for one sided (average)Mass x Cost Ratio = 6.3 two sided; 5.7 one sided

The unstiffened web solution is most often the most economic solution butit is not intended to suggest that this is always so.

Each individual situation can be readily assessed by the above processusing updated values of the cost ratio for the stiffened web solution.

Fig 8.7 Stiffened and unstiffened webs in three plate girders

8.3. Columns

8.3.1. GENERAL

The most economical columns in most building frames will usually beuniversal beam or column sections. These sections are available in arange of sizes which suit most applications. For applications where goodappearance is important, square hollow sections could be considered.In high--rise buildings it is often economical to consider compositecolumns, where a relatively small universal column is sufficient to carrydead and construction loads and which, when encased in concrete,becomes a composite column able to carry additional live loads (seeClause 5.5.2).

8.3.2. COLUMN BASE PLATES

In the design of column base plates, it is advisable once again to questionthe wisdom of minimising the mass of material and so introduce extensivefabrication, compared to a heavier base plate simply welded to the columnshaft.Fig 8.8 shows three alternative details for moment resisting base plates.

(a) Slab base plate (b) Extended flangeslab base

(c) Gusseted base plate -- avoid, too expensive

81

(d) A pipe sleeve allows easy entry of anchor bolts in a double baseplate

Fig 8.8 Column base plate details (moment resisting or fixed)

Slab base plate (a) is used widely. It calls for a thicker base plate than thegusseted base plate (c) but requires far less labour for fabrication andtherefore it is more economical. Column flanges can be extended asshown in (b) to present a larger bearing surface.

Fillet welds should always be preferred for welding the column shaft to thebase plate. Only in very rare instanceswill complete penetration butt weldsbe required -- these should be avoided if possible for maximum economy.

Typical details for pinned base plate connections are shown in Fig 8.9. Forthe nominally pinned base, there is no need to provide true pin or rockerconnections as these are unnecessarily expensive to fabricate.

Standardised dimensions for ’pinned base’ plates are available in AISCStandardized Structural Connections (Ref 1).

Notes: Weld: 6E41 continuous;Bolts: 4.6/S;Column shafts with cold sawn endsprovide full bearing contact;All dimensions in millimetres

Fig 8.9 Typical pinned base plates (full dimensional details can befound in Ref l)

82

8.3.3. HOLDING--DOWN BOLTS

One of the greatest problems facing the fabricator/erector of structuralsteelwork is inaccuracies in the placing of holding--down bolts. Thisoperation is beyond the fabricator’s control and if corrective measures arerequired on site they usually lead to cost extras and subsequentcontractual difficulties.Several methods have been adopted to overcome this problem and it isessential that the designer presents to the builder very explicit instructionson the method to be used in fixing the bolts. Fig 8.10 shows two typicalholding--down bolt details. More detailed information may be found in Ref1.In addition to providing flexibility in individual bolt location to ensurematching with base plate drilling, it is good practice to cage bolt groups asshown in Fig 8.11.

Fig 8.10 Holding--down bolt details

8.3.4. COLUMN SPLICES

In high--rise buildings economies can be achieved by running columnshafts through three or four floors rather than providing splices at say everysecond floor (Fig 8.12). Since lengths up to 18m (but seeClause 2.2.3) arenow available in most column sections, the greatest economy will begained inmaintaining the same sectionmass for 3 or 4 floors thus reducingthe number of splices required.

Column splices can be welded or bolted. The relative economics of fieldwelding should be checked with the fabricator before deciding on adoptingthis method. Bolted splices will almost always be an economical detail. Fig8.38 shows typical economic welded splices in columns; Fig 8.39 showstypical economic bolted splices. Further information on column splices canbe obtained in Ref 1.

It is essential to locate column splices at a convenient level above the floorbeams in order to provide comfortable access for the erection personnel tofield weld or install the bolts (Fig 8.13).

83

Fig 8.11 Typical holding--down bolt cage

Fig 8.12 Minimise number of column splices -- 1 is preferable to 3

Fig 8.13 Preferred column splice locations

84

8.3.5. COLUMN STIFFENERS

In rigid framed structures, the connections between the beams andcolumns very often require special stiffening of the column section in orderto provide for the satisfactory transfer of forces. These stiffeners addconsiderably to the fabricated cost of the columns and considerationshould be given at the design stage to investigating the alternative use of aheavier column section which requires no stiffening.

The example shows how such an evaluation can be carried out. For thecase investigated, it is seen that to increase the size of the column sectionfrom a 250UC89 to a 310UC137 is a more economical solution than usingthe smaller UC with stiffening.

EXAMPLE TAKEN FROM AISC ”DESIGN OF STRUCTURALCONNECTIONS” (Ref 2)

DESIGN PROBLEM:

SOLUTION 1: Stiffen 250UC89

SOLUTION 2: Increase Column Sizeto Avoid Stiffening

Requires 310UC137 to avoid any column stiffening at all.Note:

250UC89 = $101 /m310UC137 = $160 /m

Cost difference = $59 /m

COMPARISON OF SOLUTIONS

Consider 3m column lift:Solution 1:

Requires 4 stiffeners at $61 = $244

85

Solution 2:Requires 3m x $59 /m = $177

Solution 2 is the more economic

The use of a heavier columnwith a thicker web and flangemay provemoreeconomic in situations such as that illustrated, especially for short columnlifts. Each individual situation can be readily assessed by the aboveprocess using updated cost information.

Evaluation of economics of the use of column stiffeners at rigidbeam--to--column connection

8.3.6. BUILT--UP COLUMNS

Where universal column sections have insufficient capacity for a particularapplication, the use of built--up columns has to be considered. Suchcolumns can be fabricated in a variety of shapes. Fig 8.14 shows economicdetails for built--up columns in ascending order of fabrication cost.

In box columns the detail at the corner can heavily influence fabricationcosts. Where possible the use of fillet welds will afford the best economy --Fig 8.15 (a) and (b). Where fillet weld sizes required are greater than12--16mm, partial penetration welds should be considered (Fig 8.15(c)) asamore economic solution. Complete penetration butt welds at corner jointswill be rarely required and should only be considered in the vicinity of veryheavily loaded rigid beam--to--column connections.

Fig 8.14 Economic details for built--up columns in ascending orderof fabrication cost

Fig 8.15 Welded corner details for box columns

Splices in box columns can be either welded or bolted, but more often thannot the welded alternative is selected because a bolted splice is onlypracticable in large box columns where access can be provided to theinside of the box. A partial penetration welded box column splice can becarried out using the detail shown in Fig 8.16 (a). Fig 8.16 (b) shows agirder connection to box column -- site welded. This connection requiresaccurate fabrication in the overall length of the girder and may presentproblems if a considerable run of beams in a line are delivered to site withtolerances in length cumulative. In addition, allowance must be made in

86

column erection for weld shrinkage, since the relatively large weld volumerequired in heavy girder flanges will cause significant shrinkage in length.Columnsmust be spread by the shrinkage dimension, as shown in Fig 8.17and for heavy box columns this can lead to erection difficulty.

Fig 8.16(c) shows a girder--to--column connection which avoids theproblems encountered with the direct welded connection shown in Fig8.16(b). In the case of a girder stub welded to column in the shop, thecontrol of welding procedures and fabrication tolerances generally will leadto a more economic weld and better quality assurance. The subsequentsite splicing of the girder to the stub can be either welded or bolted, but thebolted alternativewill normally be less costly. In the caseof heavy industrialstructures using grid flooring however, the bolted flange splice will interferewith this type of flooring, and consideration should be given to welding thesplice for such applications.

Fig 8.16(d) shows a bolted girder--to--box column connection. Whereflexible connections are used, the angle cleat connection provides goodsite fit--up. Theweb cleats are usually loosely shop--bolted to the girder andallow movement for any out--of--tolerance during erection. For boxcolumns, provisionmust bemade in this connection for access to the insideof the column for bolt installation.

Alternatively, where flexible girder--to--box column connections areemployed, the web side plate connection will provide about equaleconomy. The web side plate can be welded to the column face, thusavoiding the problem of internal access.

Fig 8.16 Connections to box columns

Fig 8.17 Spreading of columns to allow for weld shrinkage

87

8.4. Trusses

Welded trusses have in the past provided very efficient building elementsbecause of the favourable mass/span ratio possible. Although for manyindustrial building applications, such systems as saw--tooth trusses havebeen supersededby theportal framesystem, there are stillmany long spanapplications where truss portals provide an economic solution (seeClause4.3).In general, trusses fabricated by welding should preferably use speciallydeveloped details suitable for economical welded truss fabrication ratherthan details borrowed from the days of riveted construction. For too longthe old riveted details have been used on welded trusses, on the basis ofsimply replacing rivets by equivalent welding (see Fig 8.18). This leads touneconomic fabrication, since it introduces an unnecessary amount ofwelding and, most importantly, since it requires the truss to be turnedduring fabrication to weld the angles to the gussets on each side.Several alternative details offer far more economic welded trussfabrication. Fig 8.19 shows a detail where single angles have been used asboth the truss chords and the web members. This provides for the mosteconomic truss fabrication since all welding can be done from one side,thus avoiding turning of the truss during fabrication. Additionally, thegussets have been eliminated by using a long leg angle as a chordmember. Obviously this detail requires the designer to consider theeccentricities involved in the design, but it appears in most cases that theuse of slightly heavier angles will cater for these eccentricities.

Fig 8.18 Equivalent truss detailing

Alternatively a T--section can be used for truss chord members with singleanglewebmemberswelded to the vertical leg of the tee (seeFig 8.20). TheT--sections would usually be split universal beam or column sections -- anoperation that can be economically carried out by most fabricators.

Fig 8.19 Single angle welded truss

Fig 8.20 Split tee welded truss

88

(a) Coincident intersection points. Double mitred member ends

(b) Preferred. Spread intersection points. Single mitred member ends

Fig 8.21 Use of universal sections in welded trusses

In large heavy trusses, (i.e. those fabricated from universal beam orcolumn sections), care must be taken with detailing to ensure optimumeconomy. In these

cases the detail at the intersection of members can lead to very costlyfabrication and it is suggested that the spreading of intersection points canprovide a better detail where members can be plain mitre cut to lengthrather than having double mitre end preparations. The resultingeccentricity can usually be accommodated by the relatively massive chordmembers in such trusses. Fig 8.21 illustrates the use of universal sectionsin a welded truss while Fig 8.22 illustrates the use of rectangular hollowsections. In both cases, detail (b) is preferable to detail (a).

Although trusses are usually considered as roof framing members thereare other areas where they offer economical light framing members.

Such a case is in multi--storey construction where secondary floormembers at relatively close centres are required. Economy can beachieved by the fact that a large number of thesemembers will be requiredand the use ofmass--produced trussmembers can be considered. In otherparts of theworld the openweb joist lends itself to this application andmanynotable buildings have incorporated such joists as floor members. Fig 8.23shows the traditional open web joists (a), as well as a proprietary lightweight truss (b). These light weight joists are no longermadeas a standarditem and are usually uneconomic for structural applications unless largequantities are required.

89

(a) Coincident intersection points. Double mitred member ends

(b) Preferred. Spread intersection points. Single mitred member ends

Fig 8.22 Use of rectangular hollow sections in welded trusses

Fig 8.23 Types of open web joist

(a) Non preferred (b) Preferred

Fig 8.24 End plate details

TABLE 8.1 WRENCH CLEARANCES

θ degreesRecommended MinimumDimensions (mm)b1 X (M20 and M24 bolts only)

for AirWrench*

for HandWrench

0 60 60 605 60 100 607.5 60 100 6010 60 100 60

* The use of a universal joint does offer some possibility of reducing thisdimension, and while this may be seen as an advantage from a design

90

point of view, it should be noted that an impact wrench with a universal jointand socket is generally difficult to handle for an operator some height fromground level and sitting on only the width of the beam flange. In addition,the use of a universal joint reduces the efficiency of the impact wrench andthis can be a problem in tensioningM24 bolts or larger, especially if locatedsome distance from the source of the compressed air supply.

8.5. Portal Frames

8.5.1. CONNECTIONS

A discussion of various aspects of the economics of portal frame steelbuildings is contained in Clause 4.2. A number of other items of concern tothe economic detailing of these frames is contained in this Section.

In portal frames using bolted end plate connections for the knee and apexjoints (see Fig 4.2), close attention must be paid to the detailing of theseconnections, especially where tensioned bolts (8.8/TB category) areemployed --the most common practice. Any cost savings obtained bysimplifying connection details to make fabrication simpler can be lostduring site erection if clearance problems are encountered during siteassembly. Recommended dimensions for such connections, extractedfrom Ref 1, are given in Table 8.1. These dimensions are sufficient toensure that the bolts can be installed and tensioned, since sufficientclearance is provided to accommodate either hand or air wrenches.

In the design of the end plates, designers can approach the proportioningof the end plate to resist the bending moment developed due to thebehaviour of the plate under loading in two ways; viz:

(a) Use a thick unstiffened end plate

(b) Use a thin stiffened end plate.

Fig 8.24(a) shows an excessively stiffened thin end plate which would bean extremely expensive detail compared to the thicker end plate detail ofFig 8.24 (b). For this reason, (b) is much preferred. Another problem with

excessively stiffened end--plates is that insufficient clearance may thenexist to allow the bolts to be installed. Design guidanceon the design of endplates without stiffening may be found in Ref 2.

At a bolted apex joint, care must also be taken to allow sufficient clearancebetween the adjacent purlin cleat and the end plate to enable the end platebolts to be installed and tensioned. The dimension ’Z’ (see Fig 8.25) mustbe larger than the bolt length to be installed plus a clearance dimension,and also be large enough to permit the wrench socket to be placed on thenut.

Where split universal sections are used to hauncha portal frame rafter (seeFig 4.2), stopping short the fillet weld joining the split haunch to the flangeofthe rafter is suggested as an economical and structurally sound device.Any fillet weld placed in the tight confines of the junction is likely to be ofdoubtful quality due to the difficult access involved -- see Fig 8.26.

The recommended method of attaching purlins and girts in portal framebuildings is illustrated in Fig 8.27.

Fig 8.25 Clearance at apex joint

91

Fig 8.26 Termination of haunch

NOTES:

1. Place girts and purlins to most effectively shed water and debriswith due consideration to ease of erection.

2. Ensure adequate clearance to avoid interference with cleat weld-ing.

3. Design cleats to accommodate standard punching -- refer tomanufacturers’ brochures.

4. Ensure adequate capacity in top girt to carry load from sag rods.

Fig 8.27 Attachment of purlins and girts

8.5.2. PORTAL FRAME PRE--SET

In order to ensure that the columns of a portal frame will be within the basicerection tolerances in the final erected position, it is necessary to provide a’pre--set’ of the frame during fabrication.

This is done by determining the deflection at the frame ridge under deadloads and calculating the resultant horizontal deflection at the knee joints.This latter dimension is then used in the set--out for fabrication to pre--setthe geometry of the frame -- see Fig 8.28.

Fig 8.28 Precambering details of a rigid frame

92

8.6. Connection Detailing

8.6.1. GENERAL

In general, the greatest economy in detailing of beam--to--column andbeam--to--beam connections is achieved by selecting combinations ofconnections to require only one type of operation to be executed on eachmember in the fabrication shop. Preferred ways in which this can beachieved are suggested in Fig 8.29.

Such a method of selecting connections enables the fabricator to reducethe handling operations required to fabricate the member and lends itselfreadily to a ”flow--through” system in the shop.

Preferred -- Holed only

Preferred -- Welded Fitments Only

Fig 8.29 Typical beam details for fabrication economy

The designer and detailer should look at rationalising the selection ofdetails and connections in this way. Naturally, holing operations on anygroup of similar members would use the same set--out parameters (gaugelines, pitch, hole diameter, etc).

An example of this type of selection process can be illustrated using thebeammarking plan shown in Fig 8.30. In this instance, the frame is bracedin both planes and flexible connections only are to be used.

In this frame the critical connections are those to the two box columns. Ifthese columns are small they cannot accept connections requiring boltingthrough their walls. If they are large, bolting through may be possible (withsome difficulty and expense) but the connections must be of a type wherethe beams can be entered without the need to ’spring’ the very rigidcolumns.

On both grounds the logical choice is Fig 8.34, web side plate (WP), forevery connection to the box columns.By the rule of symmetry (Clause 4.4.4.) use theWP connection at the otherend of the beams in question, B1, B4, B8 and B9. By the rule ofstandardization use the WP connection on both ends of the otherlongitudinal beams B7 and B10, checking that there will be adequateclearance at those ends of B7, B8, B9 andB10which frame into thewebsofthe l--section columns. Standardise further by using the WP connectionalso at both ends of B3 and at the column end of B6 (see Summary below).For the connections selected so far, the beams require only to be cut tolength and drilled. Therefore the connections for the transverse membersframing into them should be chosen so that the beams require only furtherdrilling (as in Fig 8.29 upper).Choosing Fig 8.33, angle cleat (AC) will achieve this aim. Another option isFig 8.32, flexible end plate.

Fig 8.30 Typical floor beam layout

Summary:

We now have a frame requiring only two different connection types,selected in such a way as to minimise fabrication and erection costs.

93

The columns themselves require welded fitments only. Beams B1, B3, B4,B7, B8, B9 and B10 require only cutting to length and drilling. Beams B2,B5, and B6 again require only cutting to length and drilling (assuming theAC connection).All beamshave the same typeof connection at eachendexceptB6where itis necessary to make aminor compromise of WP at one end and AC at theother.

8.6.2. SPECIFIC CONNECTIONS

This Clause presents notes on the efficient and economic detailing of avariety of individual connection types, as follows:Fig 8.31 Angle seat connection

8.32 Flexible end plate connection

8.33 Angle cleat connection8.34 Web side plate connection8.35 Bearing pad connection8.36 Welded moment connection8.37 Moment end plate connection8.38 Welded splice connection8.39 Bolted splice connection8.40 Stiffener connections8.41 Bracing connections8.42 Connections to concrete cores

The recommendations made in this Clause follow the AISC StandardizedStructural Connections (Ref 1).

94

• Use bolted restraint cleats for maximum economy and to allowmargin for rolling tolerances on rolled section beams.

• For welded seats, it may be necessary to taper the vertical leg of theseat in cases where the seat is welded to an H--section column webbetween flanges to allow access for welding see Fig 7.6 (b).

• Check length of seat to ensure satisfactory fit onto column. Wherethe seat is wider than the column flange, welded angle seats requirewelding from behind the column flange. This involves turning thecolumn and may prove costly -- see Fig 7.10.

• Observe recommendations on economical aspects of the use ofbolting (Section 6) and welding (Section 7).

Fig 8.31 Angle seat connection

• Select gauge ’g’ to ensure bolt clearance, (usually 90 mm).

• Fabrication of this type of connection requires close control in cuttingthe beam to length. Adequate consideration must be given tosquaring the beam ends such that both end plates are parallel andthe effect of any beam camber does not result in out--of--square endplates which makes erection and field fit--up difficult. Shims may berequired on runs of beams to compensate for mill and shoptolerances.

• The use of this connection for two sided beam--to--beamconnections should be considered carefully. Installation of bolts inthe end plates can cause difficulties in this case. When unequalsized beams are used, special coping of the bottom flange of thesmaller beam may be required to prevent it fouling the bolts.

• Since the end plate is intended to behave flexibly, damage of theend plate during transport is not normally of concern and may berectified on site.


Fig 8.32 Flexible end plate connection

95

• Cleat holes must allow for variations in beam depth due to standard rolling tolerances and also provide for erection tolerances. Standard holes (2mm larger than nominal bolt diameter) are usually sufficient.

• Check that cleat components will fit between column flanges for connections to column webs.• The use of this connection for two sided beam--to--beam connections should be considered carefully. Installation of bolts in the outstanding legs of

the angle cleats can cause difficulties in this case. When unequal sized beams are used, special coping of the bottom flange of the smaller beammay be required to prevent fouling the bolts.

• For double angle cleats, the nominal gauge required in the supporting member is (2 g3 + t). Standard gauges can hence accommodate onlycertain web thicknesses (t) of the supporting member when using normal holes (2 mm clearance). Drifting widens the range of web thicknessesthat can be accommodated, but may result in some distortion of the cleat. Alternatively, a special gauge may be used in the supporting member.

• In order to obviate both drifting or the use of a special gauge, custom detailed horizontal slotted holes may be used in the outstanding leg of theangle cleat component. Alternatively, oversize (4 mm larger than nominal bolt diameter) holes could be used, but this may complicate levelling thesupported member during erection.

• Observe recommendations on economical aspects of the use of bolting (Section 6).Fig 8.33 Angle cleat connection

96

• Bolt holes must allow for variations in beam depth due to standardrolling tolerances and also provide for erection tolerances. Standardholes (2 mm larger than nominal bolt dia.) are usually sufficient.

• In connections to column webs, a check must be made on the lengthof bolt to ensure sufficient clearance is available between the sideplate and the inside of the column flange to permit the bolt to beinstalled.

• Erection clearances must be especially considered for this detailbecause of the necessity to angle beams into place during erection.This consideration is most important for the case of a series ofbeams in the one row, all connected between the same mainsupporting members.


Fig 8.34 Web side plate connection

• The connection may need to be shimmed to suit during erection.The connection detail consequently includes provision for shims of0--5mm nominal thickness. Shims will need to be holed to the samegauge as the end plate.

• Sawn or machine flame cut edges are recommended at the bearinginterface in order to avoid edges with slopes, such as

• Check width of components when welding to H--section column webto allow access for welding -- see Fig 7.6 (b). Where the bearing padis wider than a column flange, welding is required from behind thecolumn. This involves turning the column and may prove costly.

• Observe recommendations on economical aspects of welding(Section 7).

Fig 8.35 Bearing pad connection

97

(a) Stub Girder Connection, Fully shop welded beam stub, spliced on site.

(b) Field Welded Moment Connection -- including erection cleat.

(c) Field Welded Moment Connection -- using fillet welded web cleat(s)

• The economics of field welding should be checked with the fabricator before it is specified.• Flange weld preparation assumes the use of a backing strip -- which

requires coping of the beam web.• Details (b) and (c) are not considered as economical in Australia.• Observe recommendations on economical aspects of welding (Section 7).• Site welding should be kept to a minimum, and should be used in an integrated manner.

Fig 8.36 Welded moment connection

98

• Holes are normally 2 mm larger than the nominal bolt diameter, although oversize or slotted holes may be used.• Fillet welds or butt welds may be used as the beam flange to end plate weld. A discussion of the use of fillet welds larger than 8 mm as related to

available welding processes is contained in Section 7.• Fillet welds only are recommended for the beam web to end plate weld.• Fabrication of this type of connection requires close control in cutting the beam to length and adequate consideration must be given to squaring

the beam ends such that end plates at each end are parallel and the effect of any beam camber does not result in out--of--square end plateswhich makes erection and field fit--up difficult. Shims may be required to compensate for mill and shop tolerances.

• Select a gauge for the end plate bolts which allows sufficient clearance to install the bolts.• Bolts adjacent to the tension flange should be as close as possible to the flange. Dimensions must be sufficient to ensure that bolts can be

installed and tensioned -- sufficient clearance must be provided, see Table 8.1.• Stiffeners on the end plate should be avoided -- a thicker end plate is recommended instead.• Observe the recommendations on economical aspects of the use of bolting (Section 6).

Fig 8.37 Moment end plate connection

99

• The economics of field welding should be checked with thefabricator before it is specified.

• Flange weld preparation assumes the use of a backing strip -- whichrequires coping of beam web. The backing strip should be requiredto be removed only in special instances.

• Details avoid accurate fitting up of member sections.• A shop splice with complete penetration welding without web plate is

a detail used at the discretion of a fabricator and is not a detail inuse as a site connection.

• Edges required to be prepared for bearing can be obtainedsatisfactorily and economically by cold sawing.

• Column splices should be located in positions where access can beeasily obtained for site welding -- as in Fig 8.13.

Fig 8.38 Welded splice connection

100

• Where flange splice plates are used, assemble joints with nuts tooutside of splice plate as in (a). This arrangement is recommendedfor ease of tensioning, since in universal sections sufficientclearance is not always available between flanges for a standard airwrench.

• Members can be prepared for bearing satisfactorily andeconomically by cold sawing.

• The cap plate detail of (d) is usually reserved for column splicesbetween members with significant differences in member depth.

• In order to accommodate out--of--alignment of member webs at asplice, the use of shims may be necessary. To mitigate the effects ofany out--of--alignment, holes in member flanges should be locatedusing the centre--line of the member web as a reference point.

• In order to accommodate out--of--square of member flanges at asplice, the use of tapered shims may be necessary.

• Column splices should be located in positions where access can beeasily obtained for the installation of the bolts -- as in Fig 8.13.

Fig 8.39 Bolted splice connection

101

NOTES:

1. The use of column stiffeners should be kept to a minimum formaximum economy, commensurate with design requirements.

2. All welding of stiffeners should be shop welding.

3. Only tension stiffeners need be welded to the inside face of thecolumn flange(s). Compression stiffeners may be fitted againstthe inside face of the column flange.

4. Fillet weld sizes on stiffeners should be 6 or 8 mm, to ensuresingle pass welds. Welds to column web may be one--sided.

5. Where tension stiffeners extend across the full column depth(A2), the tension stiffeners should be (fillet) welded to the columnflange and only fillet welded to the column web where flange filletwelds have insufficient capacity to transmit the design force inthe stiffener. Where tension stiffeners extend only part wayacross the column depth (A1), welding to the column web is re-quired.

6. Compression stiffeners should be fillet welded to the columnweb. When diagonal shear stiffeners are used, it is recom-mended that compression stiffeners be fillet welded to the col-umn flange adjacent to the shear stiffener.

7. Tension and compression stiffeners need to be cropped 30 mmto clear column section radiused fillets.

8. Shear (diagonal) stiffeners are fillet welded at their ends in themanner shown below. Fillet welding along the stiffener lengthmay be introduced either to increase the capacity and/or to re-duce the l/r of the stiffeners.

Fig 8.40 Stiffener connections

102

(a) Bracing gussets should be detailed as rectangular shapes to re-duce marking--off and cutting time.

(b) In braced frames it will generally prove more economic to weldbracing gussets to columns rather than to beams. The eccentrici-ty caused by spreading intersection points can usually be easilyaccommodated by the column section.

(c) For roof bracing, the most economic solution will be to weld gus-sets to the rafter top flange. Where this cannot be done, the gus-set can be welded to the rafter web but sufficient clearance mustbe provided for welding electrode access.

Fig 8.41 Bracing connections

(a) A steel plate of fairly generous proportions is presented flush withthe exterior wall of the core to which is welded a web side plateat the time of erection. Such a connection does not impose stricttolerances on (i) beam overall length (by using slotted holes inthe web side plate) or (ii) beam level and lateral location (cateredfor in the site positioning of the web side plate provided the em-bedded plate is reasonably oversize). If anchor lugs are tack--welded into the general reinforcement cage, little drift of the em-bedded plate will occur during slip forming.

(b) The older method employed for this connection is that of leavinga cored hole in the wall of the slip--formed core. Originally it wasthought necessary to embed a steel seating in this opening inwhich to bolt the bottom flange of the beam. This is not now rec-ommended since the accurate positioning of this cored hole, in-cluding an embedded seating, is almost impossible to achieve onsite. It is now considered better to leave a simple cored openingin the wall, pack the beam to level alignment during the erectionphase, and fully grout up the remaining opening.

From an economy viewpoint the alternative (b) should normally be better.However, in the overall building design it is suggested that designersconsult with the slip--core contractor to check the more economicalmethod. It is possible that in somecases a large number of coredopenings,with resultant complication of reinforcement pattern, would be moreexpensive than the embedded plate shown in alternative (a).

Fig 8.42 Connections to concrete cores

103

9. References & Further Reading

COMPLEMENTARY REFERENCES:

1. ”Standardized Structural Connections’: Australian Institute ofSteel Construction.

2. Hogan, T. J. and Thomas, 1. R., ”Design of StructuralConnections”, Australian Institute of Steel Construction, 4th ed.,1994.

SPECIFIC REFERENCES BY SECTIONS:

Note: References not mentioned specifically in the text are listed for thepurpose of further reading or as additional references.

Section 1. PRELIMINARY CONSIDERATIONS

3. Main T., Watson, K.B., and Dallas S., ”A Rational Approach toCosting Steelwork”, International Cost Engineering Council/TheAustralian Institute of Quantity Surveyors InternationalSymposium, Construction Economics -- The EssentialManagement Tool, Australia, May 1995.

4. Standards Association of Australia/Australian Institute of SteelConstruction, ”Steel Structures, Part 1 -- Planning”, SAA MA1.1--1973.

5. Standards Association of Australia/Australian Institute of SteelConstruction, ”Steel Structures, Part 7 -- Design”, SAA MA1.7--1977.

6. Firkins, A., ”Design for Economy” Third Conference on SteelDevelopments, Australian Institute of Steel Construction, 1985.

Section 2. GENERAL FACTORS AFFECTING ECONOMY

7. Day, G. A. ”Fabrication and its Future”, Steel Fabrication JournalNo. 42, Australian Institute of Steel Construction, February 1982.

8. Potter, P. D. ”Fast Steel Erection”, Steel Fabrication Journal No.46, Australian Institute of Steel Construction, February 1983.

9. Oakes, D. L. T. ”Philosophy for Economical Design, Fabricationand Erection”, Steel Construction Vol. 17 No. 4, AustralianInstitute of Steel Construction, 1983.

10. ”Hot--Dip Galvanizing’ Galvanizers Association of Australia, 13thed., 1993.

11. Macpherson, 1. J. ”Unprotected Steel Framed Open Deck CarParking Structures -- A Case Study”, Metal StructuresConference Adelaide 1976, Institution of Engineers Australia.

12. Resevsky, C. G. ”Economical Fire--Rated Composite Steel Floornow established in Australia”, Steel Construction Vol. 7 No. 3,Australian Institute of Steel Construction 1973.

13. Steel Structures Manual, Part 8 -- Fabrication, SAA MA1.8,Standards Association of Australia, 1982.

14. Steel Structures Manual, Part 9-- Erection, SAA MA1.9,Standards Association of Australia, 1975.

15. Hogan, T. J. and Firkins, A, ”Welding in a Limit State SteelStructures Code”, Tables 1, 2 and 3, Proceedings of 31 st AnnualConference, Australian Welding Institute, October 1983.

16. Quinn, N. ”Specifications: the Fabricator”, Steel FabricationJournal No. 40, Australian Institute of Steel Construction, August1981.

17. ”Handbook of Fire Protection Materials for Structural Steel’Australian Institute of Steel Construction 1990.

18. ”A Guide to the Requirements for Engineering Drawings ofStructural Steelwork”, Steel Construction Journal, Vol. 29, No. 3,September, 1995.

104

Section 3. FRAMING CONCEPTS AND CONNECTION TYPES

19. Steel Structures Manual, Part 3 -- Forms of Construction, SAAMA1.3, Standards Association of Australia, 1971.

Section 4. INDUSTRIAL BUILDINGS

20. Gaylord, E. H. and Gaylord, C. N., ”Structural EngineeringHandbook”, Section 19.2, McGraw Hill Book Co., 2nd ed., 1979.

21. Macdonald, A. J., ”Wind Loading on Buildings”, Applied SciencePublishers Ltd., 1975.

22. Gorenc, B. E., Tinyou, R., and Syam, A. ”Steel DesignersHandbook”, University of New South Wales Press, 6th Edition,1996.

23. Gorenc, B. E. ”Crane Runway Girders”, Australian Institute ofSteel Construction, 1983.

24. ”Wide Span Structures”, Steel Construction Vol. 16 No. 2,Australian Institute of Steel Construction, 1982.

25. ”Australian Conference on Space Structures”, Australian Instituteof Steel Construction, Papers, Melbourne 4/5 May, 1982.

26. Firkins, A., ”Connections for Tubular Bracing Members”, SteelFabrication Journal No. 46, February 1983.

Section 5. COMMERCIAL BUILDINGS

27. Schueller, W., ”High--Rise Building Structures”, John Wiley, 1977.

28. Hart, F., Henn, W. and Sontag, H., ”Multi--Storey Buildings inSteel”, Crosby Lockwood Staples, English Edition edited by G. B.Godfrey, 2nd ed., 1985.

29. Patrick, M. and Poon, S.L., ”Composite Beam Design and SafeLoad Tables’ Australian Institute of Steel Construction, 1989.

30. CONDECK Technical Design Manual, Stramit Metal Industries.

31. BONDEK II Composite Slabs, BD II -- 2A Composite DesignManual, BHP Building Products.

32. Johnson, R. P. and Smith, D. G. E. ”A Simple Design Method forComposite Columns”, Steel Construction Vol. 16 No. 4,Australian Institute of Steel Construction 1982.

33. Firkins, A., ”City Buildings”, Steel Construction Vol. 17 No. 1,Australian Institute of Steel Construction 1983.

34. Firkins, A., ”City Buildings -- The Steel Solution’ Structural SteelConference, Singapore Structural Steel Society, 1984.

35. Hogan, T. J. and Firkins, A., ”Economic Design and Constructionof Medium Rise Commercial Buildings using Structural Steel’Pacific Structural Steel Conference, NZ Heavy EngineeringResearch Association 1986.

Section 6. BOLTING

36. Firkins, A. and Hogan, T. J., ”Bolting of Steel Structures”,Australian Institute of Steel Construction .

37. Fisher, J. W., Kulak, G. and Struik, J. H. A., ”Guide to DesignCriteria for Bolted and Riveted Joints”, John Wiley, 1987.

Section 7. WELDING

38. The Lincoln Electric Company, ”The Procedure Handbook of ArcWelding”, 1 2th Edition, 1973.

39. Blodgett, O. W., ”Twelve Commandments to Design Engineers”,reprinted in Steel Fabrication Journal, No’s. 9, 10 and 11,Australian Institute of Steel Construction, November 1973/May1974.

40. Australian Welding Research Association, ”Economic Design ofWeldments”, AWRA Technical Note 8, March 1979.

41. Magnusson, D. J., ”Using the Structural Welding Code’ SteelFabrication Journal NQ 48, Australian Institute of SteelConstruction, August 1983.

42. Firkins, A., ”Design for Welding’ Australian Welding InstituteConference, 1988.

43. Firkins, A., and McGeachie, 1., ”Fillet Welds -- What Size isNormal?’ Asian Pacific Regional Welding Conference,International Institute of Welding, 1988.

105

10. Standards

This list doesnot purport to beexhaustive, but coversmost of the standardscurrently in print that are likely to concern the structural steel fabricationindustry.

MATERIALS

Steel

AS 1085.1 Railway Permanent Way Material, Part 1: Steel Rails.

AS 1163 Structural Steel Hollow Sections.

AS 1450 Steel Tubes for Mechanical Purposes.

AS 1594 Hot--Rolled Steel Flat Products.

AS 3597 Structural and Pressure Vessel Steel -- Quenched andTempered Plate.

AS 3678 Structural Steel -- Hot--Rolled Plates, Floorplates and Slabs.

AS 3679.1 Structural Steel, Part 1: Hot--Rolled Bars and Sections.

AS 3679.2 Structural Steel, Part 2: Welded Sections.

Bolts

AS 1110 ISO Metric Hexagon Precision Bolts and Screws.

AS 1111 ISO Metric Hexagon Commercial Bolts and Screws.

AS 1112 ISO Metric Hexagon Nuts including Thin Nuts, Slotted Nutsand Castle Nuts.

AS 1214 Hot--dip Galvanized Coatings on Threaded Fasteners.

AS 1237 Flat Metal Washers for General Engineering Purposes.

AS 1252 High--Strength Steel Bolts with Associated Nuts andWashers for Structural Engineering.

AS 1275 Metric Screw Threads for Fasteners.

AS 1559 Fasteners -- Bolts, Nuts and Washers for TowerConstruction.

Electrodes

AS 1167.2 Welding and Brazing -- Filler Metals, Part 2: Filler Metal forWelding.

AS 1553 Covered Electrodes for Welding (Parts 1 to 3).

AS 1858 Electrodes and Fluxes for Submerged--Arc Welding (Parts 1and 2).

AS 2203.1 Cored Electrodes for Arc--welding, Part 1: Ferritic SteelElectrodes.

WORKMANSHIP, DESIGN

AS 1418.1 Cranes (including hoists and winches), Part 1: GeneralRequirements.

AS 1538 Cold--formed Steel Structures Code.

AS 1554 Structural Steel Welding (Parts 1 to 6).

AS 1562.1 Design and Installation of Sheet Roof and Wall Cladding,Part 1: Metal.

AS 1657 Fixed Platforms, Walkways, Stairways and Ladders --Design, Construction and Installation.

AS 1796 Certification of Welders and Welding Supervisors.

AS 2214 Certification of Welding Supervisors -- Structural SteelWelding.

AS 2327.1 Composite Construction in Structural Steel and Concrete,Part 1: Simply Supported Beams.

106

AS 4100 Steel Structures.

SURFACE TREATMENT

AS 1627 Metal Finishing -- Preparation and Pretreatment of Surfaces(Parts 0 to 7, 9, and 10).

AS 1650 Hot--dipped Galvanized Coatings on Ferrous Articles.

AS 2311 The Painting of Buildings.

AS 2312 Guide to the Protection of Iron and Steel against ExteriorAtmospheric Corrosion.

TESTING AND INSPECTION

AS 1391 Methods for Tensile Testing of Metals.

AS 1530.4 Methods for Fire Tests on Building Materials, Componentsand Structures, Part 4: Fire Resistance Test of Elements ofBuilding Construction.

AS 1544.2 Methods for Impact Tests on Metals, Part 2: CharpyV--Notch.

AS 1710 Non--Destructive Testing of Carbon and Low Alloy SteelPlate -- Test Methods and Quality Classification.

AS 1929 Non--destructive Testing -- Glossary of Terms.

AS 2177 Non--destructive Testing -- Radiography of Welded ButtJoints in Metal (Parts 1 and 2).

AS 2205 Methods of Destructive Testing of Welds in Metal (Set ofParts).

AS 2207 Non--destructive Testing -- Ultrasonic Testing of FusionWelded Joints in Carbon and Low Alloy Steel.

WELDING TERMS AND SYMBOLS

AS 1101.3 Graphic Symbols for General Engineering, Part 3: Weldingand Non--Destructive Examination.

AS 2812 Welding, Brazing and Cutting of Metals -- Glossary ofTerms.

Economical Structural Steel Work 1

Documents

inorganic

gusseted base

computer numerically

open web joist

coincident

unequal sized

rise rectangular

standardized