1. IIa - T1-1-C1 Presentation(Vancouver) -QYZ-20120502newbuildscanada.ca/wp-content/uploads/2010/11/NEWBuildS...• Design of CLT‐floors as plate • Benefits in deformation and

Evaluation Methodology for Rolling Shear in Cross Laminated Timber (CLT)

Qinyi Zhou

Faculty of Forestry and Environment ManagementUniversity of New Brunswick

Vancouver, Canada, 2ND May 2012

(NEWBuildS T1-1-C1)

Content

Objectives

Methods

Preliminary results

Future work

Objectives

Determine the rolling shear properties of CLT Examine testing methodology Testing specimens Testing methods ASTM D198 Bending test ASTM D1037 Two‐plate shear test

verify experimental results through numerical analysis

Overall, recommend a more appropriate method

Methods

Clear black spruce boards (20oC/ 65%RH, 12% MC)

Steel strips (Cold roll 16‐gauge )

Adhesives

polyurethane for edge gluing and epoxy for face gluing

Materials

Specimens preparation

6 replicates per growth ring orientation

Specimens Testing methods

F

F/2 F/2L/3 L/3 L/3

SWS

WWW

Assumption — The shear strain is fully produced in the wood layer of a SWS specimen.

Methods

Determination of MOE and the shear modulus by variable-span 3 point bending test

Results-to-date

SWS WWW

0

2

4

6

8

10

12

0.000 0.010 0.020 0.030 0.040

1

2

3

4

5

6

(h/L)2

1/E m

,app

0

2

4

6

8

10

12

14

0.000 0.010 0.020 0.030 0.040

1

2

3

4

5

6

(Flat sawn)

(h/L)2

1/E m

,app

Results-to-date

SWS WWW

0

1

2

3

4

5

6

7

8

9

10

0.000 0.005 0.010 0.015 0.020 0.025

1

2

3

4

5

0

2

4

6

8

10

12

0.000 0.010 0.020 0.030 0.040

1

2

3

4

5

6

(h/L)2

1/E m

,app

Determination of MOE and the shear modulus by variable-span 4 point bending test

(Flat sawn)

(h/L)2

1/E m

,app

Results to-date

Apparent shear modulus (MPa)

Specimens SWS WWW

Testing method

3‐pt 4‐pt 3‐pt 4‐pt

G (MPa) 23.73 28.17 27.37 29.26COV(%) 2.4 1.3 4.2 0.5

Calculate the true rolling shear modulus fromSWS data using Shear Analogy method

Results to-date

Effect of ring orientation on apparent shear modulus

0

5

10

15

20

25

30

35

3pt 3pt 4pt 4pt

Flat sawnIn‐betweenQuarter sawn

G (MPa)

SWS WWW SWS WWW

In-between ring orientation leads to higher rolling shear modulus

Future Work

2‐plate shear test assess the influence of different loading patterns on shear properties

Verification analyze experimental results though numerical analysis

Verify the methods recommended in Phase 1 using the full size CLT product

If time permits, the study will be expanded to 5‐ or 7‐ layer CLT.

Acknowledgements

NEWBuildS ‐ NSERC strategic research Network for Engineered Wood‐based Building Systems in Canada

New Brunswick Innovation Foundation under its Research Assistantship Initiative Program

Thank you for your attention!

Questions & Suggestions?

www.NEWBuildSCanada.ca

T1‐2‐C1Influence of Laminate Characteristics on Properties and Two‐Way Bending Performance of Cross Laminated

Timber Panels

Second NEWBuildS Annual Workshop - Vancouver, BCMay 2nd, 2012

Jan Niederwestberg and Dr. Ying‐Hei Chui



Overview• Problem Statement

• Objectives

• Research Method• Phase 1: Scaled Tests• Phase 2: Full‐Scale Tests

• Conclusion & Expected Outcomes


Problem Statement• CLT‐floors usually designed as beam‐type structure• Two‐way capacity of CLT not utilized

• Design of CLT‐floors as plate• Benefits in deformation and vibration behavior

• More complex method required


Objectives• Evaluate the predictive capability of an advanced laminated plate theory

• Characterize the influence of laminate aspect ratio (width to thickness), growth ring orientation and edge‐gluing on layer properties

• Evaluate the applicability of modal testing methods to determine CLT characteristics


Research MethodPhase 1: Laminate Conditioning and Testing

• Conditioning of randomly selected material (mainly spruce) to a constant moisture content

• Modal testing of laminates beam in free‐free vibration conditions in order to determine elastic properties


Research MethodPhase 1: Testing of Laminates


Research MethodPhase 1: Laminate Grouping , Sizing & Gluing

• Grouping of laminates by characteristics (growth ring orientation, stiffness properties)

• Sizing grouped laminates to final width (120mm, 76mm & 32mm) with consideration of defects (avoiding major laminate defects)

• Glue layers from grouped laminates in order to gain ‘homogeneous’ layers (1 component PUR glue)


Research MethodPhase 1: Laminate Grouping , Sizing & Gluing


Research MethodPhase 1: Layer Sizing and Re‐Labeling

0 5000 10000 15000

A32

A34

A11

A29

A31

A30

G [N/mm2] E [N/mm2]

0 5000 10000 15000A28B166B258D35B166B178A15B178B43

B167B178A12A33D35D35B166B270B270B270B119

G [N/mm2] E [N/mm2]


Research MethodPhase 1: Testing of Layer Plates

• Modal testing of layers by methods by Sobue& Katoh and Guelzow et. al.

• Determination of modulus of elasticity & shear modulus

• Static bending tests to verify results of modal testing


Research MethodPhase 1: Testing of Layer Plates ‐ Sobue & Katoh

• Modal testing of layer, one long edge simply supported, other edges in free conditions

• Measuring of three natural frequencies and mode shapes (e.g. f11, f12, f31)

• Determination of Ex, Ey and Gxy by simple equations with mode shape related factors


Research MethodPhase 1: Testing of Layer Plates ‐ Sobue & Katoh


Research MethodPhase 1: Testing of Layer Plates ‐ Guelzow et.al.

• Modal testing of layers in free‐free boundary conditions (vertically suspended by wires)

• Measurement of natural frequencies and mode shapes

• Calculation of natural frequencies and mode shapes with plate theory (Reddy)

• Modification of layer properties in plate model to match experimental frequencies and mode shapes


Research MethodPhase 1: Testing of Layer Plates ‐ Guelzow et.al.


Research MethodPhase 1: Testing of Layer Plates


Research MethodPhase 1: Testing of Layer Plates ‐ Results

• No problems to determine natural frequencies and mode shapes

• Comparison between determined natural frequencies and natural frequencies from FEM, based on evaluated stiffness properties (Sobue & Katoh) shows promising results



‐20

0

20

40

Increase [%

]

Increase of E11 to Eaverage for 45°Laminates

32mm 76mm 120mm

‐20

‐10

0

10

20

Increase [%

]

Increase of E11 to Eaverage for Quarter‐Sawn Laminates

32mm 76mm 120mm

‐20

0

20

40Increase [%

]Increase of E11 to Eaverage for Laminate Width 32mm

flat‐sawn 45° quarter‐sawn



‐20

0

20

Increase [%

]

Increase of G12 to Gaverage for Flat‐Sawn Laminates

32mm 76mm 120mm

‐50

0

50

100

Increase [%

]

Increase of G12 to Gaverage for Quarter‐Sawn Laminates

32mm 76mm 120mm

0

50

100Increase [%

]Increase of G12 to Gaverage for Laminate Width 120mm

flat‐sawn 45° quarter‐sawn


Phase 1: CLT Production and Testing

Research Method


Research MethodPhase 1: CLT Production and Testing

• Lab‐prepared CLT elements provide the opportunity for use of strain gauges within CLT elements

• Strain gauge data provides information about internal behavior within CLT panels

Resistance strain gauge


Research MethodPhase 1: Analysis of Test Data

• Compare results of modal and static tests • Analyze and evaluate influence of laminate material and manufacturing characteristics on layer and CLT overall characteristics

• Analyze data from strain gauges to assess advanced plate model predictions



• Analysis of FEM and plate theory models with layer characteristics from laboratory tests

• Comparison between laboratory test results and results from model analysis

• Modify layer characteristics in models to match laboratory test results

• Evaluate relationship between real characteristics and input values


Research MethodPhase 2: Full‐scale Tests

• Full‐scale tests to evaluate applicability of results of Phase 1 for full‐scale CLT

• Modal tests of full‐scale single‐layer panels for comparison of results from scaled and full‐scale tests

• Estimation of laminate properties (based on modal spot tests) within CLT production

• Evaluation of properties of full‐scale CLT panels by modal test methods



• Compare scaled and full‐scale results of modal test methods

• Analyze influence of size on evaluated relationships

• Evaluate influence of laminate material and manufacturing characteristics on full‐scale layer and CLT overall characteristics


Conclusion & Expected OutcomesConducted Work

• Method by Sobue & Katoh appears suitable for single‐layer stiffness evaluation

• Laminate width of 32mm and a growth ring orientation of 45°lead to an increase of E11, quarter‐sawn laminates lead to a reduction of E11

• Laminate width of 120mm, quarter‐sawn and flat‐sawn patterns lead to an increase of G12


Conclusion & Expected OutcomesUpcoming Research ‐ Expected Outcomes

• Establish relationships between manufacturing characteristics and layer characteristics

• Evaluation of modal testing application• Evaluation of advanced plate model applicability for design use


Thanks you for your attention!

Influence of Manufacturing Parameters on Rolling Shear Behavior of Cross Laminated Timber

Minghao Li, Ph.D.Department of Wood ScienceThe University of British Columbia

NEWBuildS Annual Workshop, Richmond, BC, May 3rd, 2012 Project ID: T1-6-C1

1

1. Introduction

2. Experimental studies

• Short-term bending tests

• Fatigue bending tests

3. Continuing studies

4. Conclusions

Outlines

2

• Rolling shear (RS) strength of wood is fairly low. And rolling shear in CLT may be a concern in some loading scenarios (short-span, concentrated loads, etc.);

• The objective is to study RS behavior of CLT under short-term, fatigue and long-term loads considering important manufacturing parameters (wood species, CLT lay-ups, and clamping pressure for adhesive).

1. Introduction

3

Manufacturing parameters

• 2 Wood species (Hem-fir & S.P.F)

• 3 CLT layups

• 2 Clamping pressure levels for polyurethane adhesive (0.1 MPa & 0.4 MPa)

Species No. of layers

Clamping pressure

(MPa)Laminate Grade

Layer thickness

(mm)

Panel dimension L×W×H (mm)

Hem-fir 50.1

L1/L2/L2/L2/L1 27.5 4000×1219×137.50.4

S.P.F.

50.1

No.2/Stud/No.2/Stud/No.234/19/

34/19/343658×1219×140

0.4

30.1

No.2/Stud/No.2 34 3658×1219×1020.4

4

Species Grade Modulus of elasticityMean (GPa) Stdev (GPa)

Sample size n

Hem-firL1 13.83 2.34 329

L2 12.01 2.43 276

S.P.F.Stud 11.43 1.88 256

No. 2 or better 10.66 1.97 280

Vibration MOE tests of wood laminates

5

CLT panel specimens

6

Short-term bending tests – Group 1 (ASTM D 198-05a)

CLT group Span/depth (mm)

Peak load (kN)Mean stdev Sample size n

HF5-0.1MPa 838/137.5 19.06 2.16 30HF5-0.4MPa 838/137.5 19.70 2.42 30

SPF5-0.1MPa 840/140 20.98 3.41 30

SPF5-0.4MPa 840/140 21.78 1.56 30

SPF3-0.1MPa 612/102 15.13 1.64 30

SPF3-0.4MPa 612/102 16.51 3.15 30

2. Experimental Studies

7• ASTM D 198-05a. Standard test method of static tests of lumber in structural sizes.

Cumulative distributions of short-term capacity

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

10 15 20 25 30

Cumulative Prob

ability

Peak Load (kN)

HF5‐0.1MPa

HF5‐0.4MPa

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

10 15 20 25 30

Cumulative Prob

ability

Peak Load (kN)

SPF5‐0.1MPa

SPF5‐0.4MPa

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

10 15 20 25 30

Cumulative Prob

ability

Peak Load (kN)

SPF3‐0.1MPa

SPF3‐0.4MPa

8

Methods to estimate rolling shear strength

• Multi-layer composite beam theory (Bodig & Jayne, 1982)

τVb

• FE Modeling (ANSYS, 2011)

1) Case 1 – no gaps in cross layers and rigid glue line bonding2) Case 2 – no gaps in cross layers and flexible glue line bonding3) Case 3 – with gaps in cross layers and rigid glue line bonding4) Case 4 – with gaps in cross layers and flexible glue line bonding

9• Bodig, J. and Jayne, B. 1982. “Mechanics of wood and wood composites.” Van Nostrand Reinhold Co. Inc. NY, U.S.• ANSYS V11.0. 2011, Swanson Analysis System, Inc. Houston, PA, US

Species group Grade MOE (GPa)

EL ET ER

Shear Modulus (GPa)GLR GLT GRT

Poisson’s ratiosνLR νLT νRT

Hem-firL1 13.83 0.429 0.802 0.526 0.443 0.041

0.485 0.423 0.442L2 12.01 0.372 0.697 0.456 0.384 0.036

S.P.F.Stud 11.43 0.777 1.166 0.560 0.526 0.057

0.316 0.347 0.469No. 2 or better

10.66 0.725 1.087 0.522 0.490 0.053

CLT group Shear stiffness (kN/m3)

HF5-0.1MPa 18.2x106

HF5-0.4MPa 19.5x106

SPF5-0.1MPa 19.0x106

SPF5-0.4MPa 20.6 x106

SPF3-0.1MPa 19.0x106

SPF3-0.4MPa 20.6x106

• Orthotropic properties of wood laminates (Wood handbook, 2010)

FE model input parameters

Glue line torsional shear test

• Glue line shear stiffness (Schaaf, 2010)

10• Forest products laboratory, USDA. 2010 “Wood handbook - wood as an engineering material (centennial edition)”. Madison, WI. U.S.• Schaaf, A., 2010. “Experimental investigation of strength and stiffness properties for cross laminated timber.” Diplomarbeit, Karlsruhe

Institute of Technology. Germany.

RS stress distributions in ANSYS models (HF5-0.1MPa)

11

CLT groupPeak Load (kN)

Calculated RS strength (MPa)

Beam theory FE Case 1 FE Case 2 FE Case 3 FE Case 4

HF5-0.1MPa 19.061.61

(-2.4%)1.47

(-10.9%)1.44

(-12.7%)1.63

(-1.2%)1.65

HF5-0.4MPa 19.701.67

(-2.3%)1.51

(-11.7%)1.49

(-12.9%)1.69

(-1.2%)1.71

SPF5-0.1MPa 20.981.85

(-10.2%)1.72

(-16.5%)1.68

(-18.4%)2.01

(-2.4%)2.06

SPF5-0.4MPa 21.781.93

(-9.8%)1.79

(-16.4%)1.74

(-18.7%)2.09

(-2.3%)2.14

SPF3-0.1MPa 15.132.04

(20.7%)1.56

(-7.8%)1.50

(-11.2%)1.75

(3.6%)1.69

SPF3-0.4MPa 16.512.22

(20.0%)1.70

(-8.1%)1.64

(-11.4%)1.91

(3.2%)1.85

Summary of short-term RS strength calculated by different methods

12

2. Experimental Studies

Fatigue bending tests (Group 2 specimens)

13

14

Displacement and load history under fatigue loads

After RS failure was initiated in cross layers, crack propagation proceeded slowly

over cycles until cracks reached glue lines and specimens failed completely.

The rolling shear failure under fatigue loads seems to be more progressive and

ductile.

Test observations

15

Panel No.Average short-term capacity

(N)

25th%-tile Short-term Capacity (N)

Sample size n

No. of Cycles to failureMean Stdev Not fail(>300)

HF5-0.1MPa-1 18337.1 17383.6 10 6 7 0

HF5-0.1MPa-2 21092.0 20353.8 10 74 89 0

HF5-0.1MPa-3 17753.3 15658.4 10 142 128 2

HF5-0.4MPa-1 20256.8 19000.9 10 22 25 0

HF5-0.4MPa-2 19871.6 17407.5 10 36 40 0

HF5-0.4MPa-3 18955.7 17932.1 10 23 35 0

SPF5-0.1MPa-1 23970.0 22843.4 10 25 23 0

SPF5-0.1MPa-2 21917.2 20536.4 10 84 94 1

SPF5-0.4MPa-1 22715.8 21421.0 10 105 89 1

SPF5-0.4MPa-2 20824.7 20116.7 10 101 95 0

SPF5-0.4MPa-3 21785.7 20543.9 10 117 113 1

Fatigue test summary

16

Relationship between stress levels and no. of cycles to failure

17

Complete fatigue bending tests of all Group 2 specimens;

Conduct long-term duration of load tests of Group 3 specimens;

Calibrate and verify damage accumulation models based on the test

database; and

Conduct reliability analysis to quantify long-term behavior using verified

models

3. Continuing Studies

18

4. Conclusions

The beam theory might not be accurate enough to calculate the actual

RS strength for non-edge-glued CLT specimens due to the existing gaps

in cross layers;

The thickness of cross layers in SPF specimens seems to affect short-

term RS strength significantly;

Increasing clamping pressure from 0.1 MPa to 0.4 MPa helps increase

short-term RS strength for HF5 and SPF3 CLT configurations;

Fatigue test results indicated large variability in number of cycles to

failure for the specimens; and

The linear relationship between stress levels and the logarithms of

number of cycles to failure has been observed.

19

• NEWBuildS – NSERC Strategic Research Network for Engineered Wood-based Building Systems;

• Prof. Frank Lam, Mr. George Lee and Mr. Alex Yuan Li from the timber engineering research group at University of British Columbia; and

• CST Innovations

Acknowledgement

20



NEWBUILDS WORKSHOPTHEME Ι ‐ T1‐7‐C3

IN‐PLANE STIFFNESS OF CLT DIAPHRAGMS

Sepideh Ashtari (MASc Student, Dept. of Civil Engineering, UBC)

Supervisors:Dr. Terje Haukaas Dr. Frank Lam

May 2012


Objectives

• In‐plane behaviour of connected CLT panels

• Influential parameters

• Distribution of seismic forces to shear walls

• Recommendations for design


In‐plane Stiffness of Wood

Diaphragms/Shearwalls

CLT Panels Mechanical Behavior

Literature on In‐plane Stiffness of CLT

I. Numerical Models

II. Experimental Studies


Numerical Models– In‐plane shear stiffness by FEM– Simplified equivalent shear modulus

Experimental Studies– Cyclic and monotonic tests– Force‐Displacement curves


On‐going Research at UBC

I. Connection Tests at TEAM Laboratory

II. Numerical Modeling in ANSYS


Connection Tests at UBC

In-plane loading – connection layout E


Test Results


Numerical Modeling in ANSYS

• In‐plane stiffness of connected CLT panels

• 2D Local/Globalmodels

• connection test data

• Different Boundary Conditions

• Push‐over curves / Lateral Force Distribution


Influential Parameters

I. Orthotropic Material Constants

II. Diaphragm Configuration

– Dimensions of Panels

– Number of Connected Panels

III. Nonlinear Curve Parameters of Connections

IV. Boundary Conditions


Orthotropic Material Constants

Ex 8000 MPa Pxy 0.35Ey 9000 MPa Pyz 0.3Ez 1000 MPa Pxz 0.47Gxy 450 MPaGyz 500 MPaGxz 400 MPa

Selected Material Constants for Numerical Modeling in ANSYS


Diaphragm Configurations

Configuration 1 Configuration 2


Nonlinear Curve Parameters of Connections

0

10

20

30

40

50

60

70

0 1 2 3 4 5 6 7 8 9 10

Force (KN)

Displacement (mm)

Average Force‐ Displacement Curve for Layout E

Average Curve Piecewise‐Linear Approx

Slope of Segments

Failure Point

Number of Segments


ANSYS Models

1. Local connection model (validation model)

2. Local diaphragm model

3. Global building model


Rigid BCs at the location of shearwalls


Linear springs at the location of shearwalls –rigid links in X direction


Linear springs at the location of shearwalls –non‐linear springs in X direction




Rigid BCs at the location of shearwalls


Linear springs at the location of shearwalls –rigid links in X direction




Results

• Almost Linear Response

• Non‐linearity source in connections

• Force distribution depending on relative stiffness


Future Works

• Sensitivity Analysis

• Classifying push‐over curves

• Design recommendations

• Reasonable failure criteria for diaphragms


Thank You!

T1‐11‐C1Connections in CLT Building Systems

Connections for CLT Diaphragms in Steel‐framed Buildings

Tom Joyce, MScFE CandidateDr. Ian Smith, Supervisor


Cross Laminated Timber

Commonly called CLT or X‐Lam Massive timber panel product Layers of boards with alternating orientation

Potential for two‐way bending High in‐plane strength and stiffness

Image courtesy of FPInnovations.


Properties influenced by number, arrangement and thickness of layers

Produced and developed originally in Europe Commercially‐available CLT now produced in Canada

Nordic EWP as of June 2011 Structurlam as of September 2011


Standard 3‐, 5‐, and 7‐layer options are available with uniform layer thicknesses of 35mm

Can design custom orders 7‐Layer modified to be stiffer in one direction

For this project, considering normal 7‐layer

3-Layer: 105mm 5-Layer: 175mm

7-Layer: 245mm 7-Layer (modified)

CLT Construction in Europe

Massive timber buildings Walls (typically 3‐5 layers)

Large parallel‐to‐grain area Cross layers transfer shear

Floors (typically 7 or more layers) Thicker to reduce vibrations, deflections, carry moments

CLT Construction in Europe

Tall modern timber buildings Stadthaus, London (8+1 storeys) Bridport House, Hackney (8+1 storeys) Plans for timber buildings of 10 – 15 storeys

Stadthaus, Murray Grove, London, UKImage courtesy of Waugh Thistleton Architects

Bridpoirt House, Hackney, UKImage courtesy of Stora Enso.

CLT Construction in N.A.

Typical CLT buildings use large volume of CLT Unlikely to find large market in North America

May find better application in hybrid buildings Building cores – Equilibrium Consulting Diaphragms – Asiz and Smith, 2009, 2010

Image courtesy of Smith and Frangi, 2008

CLT Construction in N.A.

Diaphragm feasibility studied by Asiz and Smith Compared CLT vs. RC (reinforced concrete) floors over steel and RC frames Reduced inter‐storey drift Lower building weight, reduced load on framing and foundation

CLT‐Steel buildings appear to be more efficient than CLT‐RC Higher weight of concrete buildings reduces benefits of light‐weight CLT

Loss of benefits from T‐beam design

Floor Functional Requirements

Ideally, floors and connections should perform well in the following areas... Structurally: acceptable strength and stiffness Fire performance Constructability: ease of assembly, placement, installation

Ability to disperse loads into panels Cost(Adapted roughly from Borg Madsen's Behaviour of Timber Connections)

Project Objectives

Develop a connection capable of meeting requirements needed for the application of CLT as diaphragms in steel‐framed buildings.

Three phase approach: Phase 1 ‐ Investigate existing CLT connections Phase 2 ‐ Design diaphragm and connections Phase 3 ‐ Develop design procedure

Phase 1

Phase 1: Investigation of existing CLT connections: Literature review Design of timber and CLT connections Predicting connection strength Modes of CLT failure, particularly connection failures

Experimental tests Determine benefits of different connection systems, screw diameters, angles

Phase 1 – CLT Connections

Design embedment and withdrawal formulae proposed by Uibel and Blass for CLT Two embedment formulae One based on assumption of uniform properties throughout for thin layers

Other for thicker layers Can be used with existing design codes

Strength and stiffness of line of connections between panels varied by adjusting spacing

Cross‐wise layers restrict splitting Likely can reduce end spacing restrictions

Phase 1 – CLT Connections

Typically connections between CLT wall panels use dowel‐type fasteners

Use splines, laps, or butt joints

For tests, used the two in lower row Double spline and angled (inclined) screws Double spline provides out‐of‐plane moment resistance

Angled screws act more as a hinge

Phase 1 – Method

Testing of CLT connections Loads: Shear and tension loads Stiffness in elastic compression characterized

Fasteners: 8 and 10mm Ecofast ASSY for double spline Partially (Ecofast ASSY) and fully (ASSY VG) threaded screws for angled screws

Load Types: Monotonic – 5 repetitions Cyclic – 2 repetitions

Phase 1 ‐ Results

Screw diameter: Thinner diameter showed slightly lower strength and stiffness

Similar behaviour in failure One plastic hinge in CLT

Phase 1 ‐ Results

Screw angle: Axially loaded screws showed higher stiffness than laterally loaded

Potential for brittle splitting failure of outer laminates with angled screws

Phase 1 ‐ Results

Threaded length: Threading on both sides of angled connections reduces risk of splitting failure

Failure in withdrawal shows highest strength and stiffness of tested connections

Phase 1 ‐ Lessons

Observed failures in CLT were likely result of close proximity of connection to edge of panel Delamination and splitting of outer lamellas Splitting did not extend further into panels

Large diameter screws caused splitting of outer lamellas during assembly

Double spline typically failed in plywood splines Withdrawal gives stronger and stiffer connection Longer threaded length can change failure mode, but can also hinder ease of assembly

Phase 2 ‐ Objectives

Goal is to select a connection for further testing and development

Select steel member for beams Design floors and compared based on set of criteria (allowable spans, deflections...)

Design set of connections for given member Perform tests to determine properties of connections

Compare options and select best, to be tested further in Phase 3

Phase 2 ‐ Criteria

Strength

Out‐of‐plane floor loads

Criteria: must not fail under applied loads

Diaphragm forces Diaphragm type Flexible – forces transferred by bending

Rigid – forces transferred to supports according to stiffness

Max Shear

Max Moment


Stiffness

Slab deflections

Criteria: should meet L/360 serviceability criteria for slab deflections, and deflect less than twice the sway of the columns (rigid diaphragm)

Diaphragm deflection

d

d


Deflections and vibrations

Limit spans and allowable loads Influence occupant comfort Potentially damaging to partition walls, windows, and operation of doors

Beam Deflection: Can modify moment of inertia, I, and length

Moment of Inertia: Increases most withincreased depth


Constructability and cost

Criteria: should be easily constructable and not use complicated or expensive systems

Difficult or impossible to position panel

Fastening from below – working overhead


Eccentric loading of beams and instabilities

Criteria: forces and load cases that may induce buckling of beam or diaphragm should be avoided

Phase 2A ‐Method

Design diaphragm and connections Two general design directions: Slab and Beam Connect CLT simply to steel beam

Plate System Create plate system, possibly with new steel section

Top image courtesy of Kuhlmann et al., 2008Image courtesy of Aziz and Smith, 2009.

Phase 2A ‐Method

Use NBCC 2005 Non‐ground floors of office building: DL: 1.2kPa (concrete) + 1.3kPa (CLT) LL: 2.4kPa (occupancy) + 1.0kPa (partitions)

Select or design beams and determine properties What are the limiting properties? How is the CLT diaphragm influenced by different choices for steel members?

What is the maximum allowable span? Goal is for bays to be larger than 6m Have fewer columns and more open space

Phase 2A ‐ Results

For plates, with depth limited to roughly 245mm, spans were limited by deflection Angles and T‐beams have little steel in compression, which lead to low neutral axes and greater deflection

Moment of inertia of shallow beams was low Overdesigned for shear and bending Two‐way action of panels limited by difference in bending stiffness in both directions EIlongitudinal ~ 10 x EIperpendicular

Phase 2A ‐ Results

For beams and slabs, floor thickness was large Panels over beams created large floor thicknesses Less occupiable space for given height

Could use sub‐beams to provide extra support to CLT and create larger spaces

In some cases there is a risk of instability Potential for buckling of diaphragm

Next step is to select best option and develop connection options given type of steel member

Phase 2B ‐Method

Tests of connection options Create three connection options Static tests with loads acting perpendicular in, perpendicular out, and parallel to beam axis 4 repetitions for each option under each load

Estimate properties of each Elastic stiffness Mode of failure Strengths at yielding and failure

Select a final design for development

Phase 3

Development of design and design procedure Perform additional tests required to characterize properties of system

Where applicable, modifications will be made to determine influence of various factors Fastener diameter, angle, type Steel plate or angle thickness or dimensions Panel orientation

Expected Output

Project oriented around design of connection for CLT diaphragms Identify beams / connection combination Develop design procedure based on experimental findings of strength and stiffness

Results will combine with other work done at UNB on the development of CLT diaphragms and the addressing of potential areas of concern


1. IIa - T1-1-C1 Presentation(Vancouver) -QYZ-20120502newbuildscanada.ca/wp-content/uploads/2010/11/NEWBuildS...• Design of CLT‐floors as plate • Benefits in deformation and

Documents