Concentrated load introduction in CLT elements perpendicular to plane

Concentrated load introduction in CLT elements perpendicular to plane EUROMECH Colloquium 556 May 27-29, 2015 Dresden, Germany Theoretical, Numerical and Experimental Analyses in Wood Mechanics http://556.euromech.org Thomas Bogensperger holz.bau forschungs gmbh Inffeldgasse 24/I, A-8010 Graz Robert August Jöbstl Haas Fertigbau Holzbauwerk GmbH & Co KG Haas Holzprodukte GmbH, Radersdorf 62, A-8263 Großwilfersdorf/Industriestraße 8, D-84326 Falkenberg

content Introduction Cross Laminated Timber (CLT) Short introduction/motivation Verifications (ULS) of CLT plates under bending/shear (engineering level) Research of Mestek (PhD at Munich University of Technology, 2011) Test configuration bending configuration geometry, typ of CLT plate, chosen cross sections linear elastic analysis ULS verifications (bending, shear), expected failure modes Outlook: shear configuration Test results overview of test programm, load displacement curves & statistical evaluation of results observed failure mechanisms Mechanical model Material parameters/length-effect of brittle material (timber tension members) Numerical model and comparison to test results Summary and outlook 2

Overview & Cross Laminated Timber (CLT) Cross Laminated Timber (CLT) classification / grading trimming finger jointing surface bonding/ edge bonding (optional) storage, charging and transport. London (UK) - 8F examples of implementation pressure Pirmin Jung with / without transverse pressure in and / or across the direction of production 3

Overview & Cross Laminated Timber (CLT) Cross Laminated Timber (CLT) classification / grading trimming finger jointing surface bonding/ edge bonding (optional) storage, charging and transport. London (UK) - 8F examples of implementation Milano (IT) - 9F pressure Pirmin Jung with / without transverse pressure in and / or across the direction of production TEKNE 4

Overview & Cross Laminated Timber (CLT) Verifications (ULS) of CLT plates under bending/shear (engineering level) critical bending stress (tension) f m,c,05 = 3,5 14,00,8 = 28,9 E0 E90» 0 From: A Contribution to the Design and System Effect of Cross Laminated Timber (CLT) E0 E90» 0 E0 R.A. Jöbstl, et al, CIB meeting 39 Florence, Italy (2006) M G90 ¹ 0 G90 ¹ 0 tension critical shear stress (rolling shear stress) G0 G0 N/mm² fv,90,05 = 1,25 V N/mm² From: CLT Handbook (in german language) Schickhofer, Bogensperger, Moosbrugger (2010) G0 5

Overview & Cross Laminated Timber (CLT) Research of P. Mestek (PhD at Munich University of Technology, 2011) Rolling shear stress in [N/mm²] PhD Mestek: (punching) shear behaviour of (vacuum-pressed) CLT with reinforcement (45 screws) only few tests without reinforcement Our focus: (punching) shear behaviour of CLT without reinforcement 6

Overview & Cross Laminated Timber (CLT) Research of P. Mestek (PhD at Munich University of Technology, 2011) estimated mean value: Fmean=381 kn F0,05=3 3 kn (estimated) Based on COV of 5 8% Control perimeter for verification of punching through CLT plate Based on test results in Mestek an elastic rolling shear strength of 2,06 2,25 N/mm² can be concluded. fv,90,05» 2,25 N/mm² (+80% for our investigations) 7

Test configuration & linear elastic verifications Parameters of test configuration Test configuration Phd Mestek Failure mechanism (bending failure, shear (punching) failure) Size and thickness of test elements Number of layers (CLT 5s, CLT 7s) Support of CLT (two-four side boundary condition, ) Load introduction (local reinforcement) Numerical study (FEM) for investigation of these parameters. Test configuration in Graz 9

Test configuration & linear elastic verifications bending configuration geometry CLT 7s type of CLT plate chosen cross sections chosen load introduction (25/25 cm and reinforcement with screws) solution of support (tension rods) CLT 5s 7-layered build up boundaries at all sides Cross section in main direction 162 mm main direction 2,5 m 4,0 m 166 mm main direction 4,0 m 2,5 m Cross section in main direction 5-layered build up boundaries only at short sides 10

Test configuration & linear elastic verifications ULS verifications (bending, shear), expected failure modes CLT-7s my qx qy [kncm/cm] [kncm/cm] [kn/cm] [kn/cm] 188,9 med 102,53 29,16 1,57 1,05 kn mrd 102,53 40,41 2,98 1,55 utilisation 100,0% 72,2% 52,7% 68,2% 4, med 61,06 67,87 1,55 1,59 KN mrd 79,94 67,87 2,78 2,31 utilisation 76,4% 100,0% 55,6% 69,1% conclusions: mx CLT 7s CLT-5s calulated max load CLT 5s type CLT-5s support only along two short sides bending failure in main span direction can be predicted CLT-7s support along all four sides bending failure perpendicular to main span direction can be predicted 11

Test configuration & linear elastic verifications Outlook: shear configuration 162 mm CLT-5s 1,0 m 1,0 m 166 mm CLT-7s Properties of chosen CLT sections element E0 E90 G0 G90 # N/mm² N/mm² N/mm² N/mm² t1 t2 t3 t4 t5 t6 t7 tges mm mm mm mm mm mm mm mm CLT-5s 11.600 0 690 5 34 34 34 - - 162 CLT-7s 11.600 0 690 7 19 19 19 19 166 12

overview of test programm Executed test program 5-layered, uniaxial loaded CLT (CLT 5s) 7-layered, biaxial loaded CLT (CLT 7s) cross sections of CLT elements CLT producer quantity #layers tges t1 t2 t3 t4 t5 t6 t7 [-] [-] [mm] [mm] [mm] [mm] [mm] [mm] [mm] [mm] B_B_5s producer 1 3 5 162 34 34 34 B_B_7s producer 1 3 7 166 19 19 19 19 C_B_7s producer 2 3 7 170 20 20 20 20 Sum 9 pieces 14

load displacement curves & statistical evaluation of results Experimental load displacement curves for CLT 5s 15

load displacement curves & statistical evaluation of results Experimental load displacement curves for CLT 7s 16

load displacement curves & statistical evaluation of results Test results versus linear elastic FEM predictions Ffailure [kn] 5s-1 (producer 1) 5s-2 (producer 1) 5s-3 (producer 1) Mean value COV [%] 7s-1 (producer 1) 7s-2 (producer 1) 7s-3 (producer 1) Mean value COV [%] 229,5 242,7 241,0 237,7 3,0 281,8 317,1 338,6 312,5 9,2 7s-1 (producer 2) 7s-2 (producer 2) 7s-3 (producer 2) 336,4 353,0 393,1 Mean value 360,8 COV [%] Observed failure in labor Predicted failure based on FEM F0,05 [kn] (Prediction) Bending (longitudinal, finger joint) Longitudinal bending Bending (longitudinal, finger joint) Longitudinal bending Bending (longitudinal, knot failure) Longitudinal bending ~ 26% Bending (longitudinal, finger joint) Bending (transverse) Bending (transverse) transverse bending transverse bending Transverse bending 38% Bending (longitudinal) Bending (longitudinal) Bending (longitudinal) 188,9 227,0 transverse bending transverse bending transverse bending 8,1 17

observed failure mechanisms Observed primary failures of CLT 5s uniaxial load behaviour Bending failure in top layer (primary failure) Failure due to rolling shear (secondary failure) failure of CLT 5s 18

observed failure mechanisms Observed primary failures of CLT 7s Bending failure in top layer (primary failure, 4x) Bending failure in second layer (primary failure, 2x) Failure due to rolling shear (secondary failure) failure of CLT 7s 19

observed failure mechanisms Observed secondary rolling shear failures (e.g. CLT 7s) cut cut cross section perpendicular to top/bottom lamella cross section parallel to top/bottom lamella 20

Material parameters/length-effect of brittle material Elastic values and mean tension strength (in fibre direction) Elastic constants of CLT Stiffness value E0 12000 N/mm² E90 370 N/mm² G0 690 N/mm² G90 (GR) N/mm² Effects due to poisson numbers are neglected! Various tests results bending tension of CLT client Number Mean value N/mm² 1 4 44,2 1 8 40,6 2 10 56,46 3 41 37,2 3 20 39,4 4 9 37,4 mean value over all tests: 39 N/mm² 22

Material parameters/length-effect of brittle material Mean Tension and compression strength perpendicular to fibre direction in fibre direction σ0 39 N/mm² σ90 Fracture energy: Gc=16 N m/m² char. length of FE-element: 1 cm yield plateau in numerical model 0,5 N/mm² tension compression ε90 ε0 All given values are mean values! compression tension 37 N/mm² 3,0 N/mm² Sources for strength values: in fibre direction: scientific test program (Comet P01, LS0231) at competence centre holz.bau forschung and Diploma thesis RULI 2008 at Institut for timber engineering and wood technology at TU GRAZ Perpendicular to fibre direction: Diploma thesis RULI 2008/Master thesis STUEFER 2011 23

Material parameters/length-effect of brittle material perpendicular to fibre direction (Rolling shear strength) Mean shear strength in fibre direction τ0 yield plateau in numerical model τ90 4,2 N/mm² 1,5 N/mm² g0 4,2 N/mm² yield plateau in numerical model g 90 1,5 N/mm² All given values are mean values! Shear failure happens subsequently softening behaviour not considered in models with bending failure Sources for strength values: in fibre direction: G. Schickhofer: Determination of Shear Strength Values for GLT using Visual and Machine Graded Spruce Laminations, CIB W18 34th, Venice, 2001 Perpendicular to fibre direction: Master thesis Ehrhard 2014 Institut for timber engineering and wood technology at TU GRAZ 24

Material parameters/length-effect of brittle material Length effect Illustration of mean size effect of brittle materials (acc. to Bažant) Experimental exploration of influence of length on strength (project qm_online 2007, holz.bau forschung Graz) Test site/arrangement for experimental investigation of length effect of brittle materials (2005-07) length effect: f t,k,mean ft,k,ref,mean æ l =ç çl è ref ö ø -a With exponent a» 0,15 a» 0, 22 5 % quantile values for mean values 25

Material parameters/length-effect of brittle material relevant area comparison of moments Length effect 1,200 1,000 0,800 70 cm 0,600 1 cm 0,400 lclt=400 cm 0,200 0,000 0,0 20,0 40,0 60,0 80,0 100,0 120,0 140,0 160,0 180,0 200,0 reference system for bending strength lref=18 h 292 cm ft,k, mean f t, k, ref, mean æ l =ç ç lref è ö ø -a æ 70 ö =ç è 1 ø -0,20 = 1,16 Tension strength 39 45 N/mm² 26

Numerical model and comparison to test results FEM Model: shell-solid elements support for CLT5/7s additional support for CLT7s 2,5 m 4,0 m additional support for CLT7s support for CLT5/7s 27

Numerical model and comparison to test results FEM Model: shell-solid elements shell-solid coupling additional support for CLT7s support for CLT5/7s F F/4 plane of symmetrie 28

Numerical model and comparison to test results FEM Model: shell-solid elements CLT 5s (two side support) In tension zone: one row of elements with softening material behaviour 29

Numerical model and comparison to test results FEM Model: shell-solid elements CLT 5s (two side support) In tension zone: Small area with softening elements bottom layer 16, 32 and 80 cm

Numerical model and comparison to test results 400 CLT5s (two side support) 3 Model 1: 0 No length effect Failure may occure over 32 cm Displacement controlled 2 200 1 100 FEM (ABAQUS) FEM (cont. ) BSP_KLH_B_5s-1 BSP_KLH_B_5s-2 BSP_KLH_B_5s-3 0 0 10 20 40 60 70 80 90 100 31

Numerical model and comparison to test results 400 CLT5s (two side support) 3 Model 1: 0 No length effect Failure may occure over 32 cm Displacement controlled 2 200 Stresses in edge: 1 39 N/mm² 45 40 35 100 FEM (ABAQUS) 25 FEM (cont. ) 20 BSP_KLH_B_5s-1 15 BSP_KLH_B_5s-2 10 BSP_KLH_B_5s-3 5 0 0,0 10,0 20,0,0 40,0,0 60,0 70,0 80,0 90,0 100,0 0 0 10 20 40 60 70 80 90 100 32

Numerical model and comparison to test results 400 CLT5s (two side support) 3 Model 2: 0 length effect Failure may occure over 32 cm Displacement controlled 2 200 1 100 FEM (ABAQUS) FEM (cont. ) BSP_KLH_B_5s-1 BSP_KLH_B_5s-2 BSP_KLH_B_5s-3 0 0 10 20 40 60 70 80 90 100 33

Numerical model and comparison to test results 400 CLT5s (two side support) Symmetry: 32 cm 4 lamella 3 Model 2: 0 length effect Failure may occure over 32 cm Displacement controlled 2 200 Failure in lamella in bottom layer 1 100 FEM (ABAQUS) FEM (cont. ) BSP_KLH_B_5s-1 BSP_KLH_B_5s-2 BSP_KLH_B_5s-3 0 0 10 20 40 60 70 80 90 100 34

Numerical model and comparison to test results 400 CLT5s (two side support) Lack of stress interactions in numerical model! (near load introduction) 3 Model 2: 0 length effect Failure may occure over 32 cm Displacement controlled 2 ~15% 200 1 100 FEM (ABAQUS) FEM (cont. ) BSP_KLH_B_5s-1 BSP_KLH_B_5s-2 BSP_KLH_B_5s-3 0 0 10 20 40 60 70 80 90 100 35

Numerical model and comparison to test results FEM Model: shell-solid elements CLT 7s (all side support) In tension zone: Small area with softening elements bottom layer 16, 32 and 80 cm 36

Numerical model and comparison to test results 400 CLT7s (all side support) 3 Model 2: 0 length effect Failure may occure over 32 cm Failure in longitudinal direction Displacement controlled 2 200 1 100 FEM (ABAQUS) FEM Elastic Solution BSP_KLH_B_7s-1 BSP_KLH_B_7s-2 BSP_KLH_B_7s-3 0 0 10 20 40 60 70 80 37

Numerical model and comparison to test results 400 CLT7s (all side support) 3 Model 2: 0 length effect Failure may occure over 32 cm Failure in transverse direction Displacement controlled 2 200 1 FEM (ABAQUS) FEM (cont. ) 100 FEM (failur e longitudinal) BSP_KLH_B_7s-1 BSP_KLH_B_7s-2 BSP_KLH_B_7s-3 0 0 10 20 40 60 70 80 38

Numerical model and comparison to test results 400 CLT7s (all side support) Failure in longitudinal direction ~20% 3 Model 2: Failure in transverse direction 0 length effect Failure may occure over 32 cm Failure in transverse direction Displacement controlled Difference in bending strength in longitudinal and transverse direction STILL OPEN! 2 200 1 FEM (ABAQUS) FEM (cont. ) 100 FEM (failur e longitudinal) BSP_KLH_B_7s-1 BSP_KLH_B_7s-2 BSP_KLH_B_7s-3 0 0 10 20 40 60 70 80 39

Summary and outlook Summary Developed test configurations and test results of local concentrated load introduction into CLT plates were shown. An observed apparently ducilitiy can be explained with the ductility of compression perpendicular to grain. Expected and observed failures and load capacities were presented. Bending tests were computed with a numerical FE - model. All implemented strength values are deducted from a pool of test results (various research programs, master and diploma thesis). Well known length effect is applied to brittle tension strength. Good congruence between test tesults and numerical results can be achieved. Numerical results are about 15% higher than results from experimental 41

Summary and outlook and outlook This disagreement between numerical model and test results (~15%) can be explained with lack of interaction conditions in strenth model. IMPLEMENTATION of MATERIAL with full coupled stress interactions Transverse bending strength seems to be higher than well known bending strength in main direction stress redistribution possible at both sides of layer can be expected up to 20% OPEN field for scientists: experimental and/or theoretical based proof is still missing 42

Contact DI Dr.techn. Thomas Bogensperger holz.bau forschungs gmbh +43 (0) 316 873-4608 bogensperger@tugraz.at Inffeldgasse 24/I, A-8010 Graz

Overview & Cross Laminated Timber (CLT) Short motivation to investigated problem Punching of CLT elements 44

Test configuration & linear elastic verifications bending moments linear elastic analysis bending moments mx line 1 260.0 240.0 220.0 bending moments my line 2 80.0 mx,ma mx-v2 mx-v3 mx-v4 mx-v5 70.0 60.0 my [kncm] 200.0 mx [kncm] my,max= 77,63 kn cm/cm my-v2 my-v3 my-v4 my-v5 180.0 160.0 140.0.0 40.0.0 120.0 20.0 100.0 10.0 80.0 0 20 40 60 80 100 120 0.0 0 100 x [cm] Reference force: Fref = 400/4 kn symmetry shear forces qy line "q y" 2.6 qx-v2 qx-v3 qx-v4 qx-v5 2.5 2.3 E0 N/mm² CLT-5s 11.600 E90 2.2 3.0 qx,mean =3,09 kn/cm 2.5 2.1 qy,mean =2,08 kn/cm 2.0 1.9 1.8 2.0 1.7 1.5 108 110 112 114 116 118 120 122 1.6 182 124 184 186 188 x [cm] G0 G90 # N/mm² N/mm² N/mm² 0 qy [kn] qx [kn] symmetry elemen t qy-v2 qy-v3 qy-v4 2.4 3.5 Properties of elected CLT sections 200 shear forces shear forces qx line "q x" 4.5 4.0 1 x [cm] 690 5 190 192 194 196 198 200 x [cm] t1 t2 t3 t4 t5 t6 t7 tges mm mm mm mm mm mm mm mm 34 34 34 - - 162 45

load displacement curves & statistical evaluation of results Experimental load displacement curves for CLT 5s 46

load displacement curves & statistical evaluation of results Experimental load displacement curves for CLT 7s 47

Numerical model and comparison to test results 400 CLT5s (two side support) 3 Model 3: 0 length effect Failure may occure over 80 cm Displacement controlled 2 200 1 100 FEM (ABAQUS) FEM (cont. ) BSP_KLH_B_5s-1 BSP_KLH_B_5s-2 BSP_KLH_B_5s-3 0 0 10 20 40 60 70 80 90 100 48

Numerical model and comparison to test results 400 CLT7s (all side support) 3 Model 1: 0 No length effect Failure may occure over 32 cm Failure in longitudinal direction Displacement controlled 2 200 1 FEM (ABAQUS) FEM Elastic Solution BSP_KLH_B_7s-1 100 BSP_KLH_B_7s-2 BSP_KLH_B_7s-3 0 0 10 20 40 60 70 80 49

Numerical model and comparison to test results 400 CLT7s (all side support) 3 Model 1: 0 No length effect Failure may occure over 32 cm Failure in transverse direction Displacement controlled 2 200 1 FEM (ABAQUS) 100 FEM (cont. ) BSP_KLH_B_7s-1 BSP_KLH_B_7s-2 BSP_KLH_B_7s-3 0 0 10 20 40 60 70 80