 

 

.,.,, :

lin‘ .

...o
.2 x. .
.

.(w‘
.5333. an. 3

ﬂagﬁn ..s
3... 11th... .

:5. a...
. .. v2.2
Re: L 0““...
a? be ~
é:

. .3
an“. v

.ﬁ

3:: .
x

6.? .
:5...

r. .u
9*... J.

rt... v u 3/.
.3... 1

...« .

 

 

 

 

 

 

 

 

may

83%?

start

‘LI‘UV E...) w
or u
3...": £3.95
aubmmdm..hx. .
: Mummy“?!
......
5%.:

«no...

u“

$5: --~‘“

Er , . k
knsumlzéx. .
A...“ 7‘ , H” .
.5 x , .

.. a;
.
‘

.
«32m.

”it

rt:

a «.th
n2-

rikwma. xv.
1%... .awnri
r »

..AI.1..«".

5: t
8...
I. .I. J.
. .91....“ a»:
.53.": {A :35.
5.: s... . a my .2.
» ,5...5.3.3 .2.

niﬁma

.. V nsnspunﬂ

Z.

‘.L

 

411:5. . .

1212.518

300$

 

LIBRARY
Michigan State
URIVGI sity

 

 

 

This is to certify that the
dissertation entitled

WOOD-CEMENT PARTICLEBOARD: IMPROVED
MANUFACTURING, MATERIAL CHARACTERIZATION AND
POTENTIAL APPLICATION IN CONCRETE CRASH

 

BARRIERS

presented by

AN GONG

has been accepted towards fulﬁllment
of the requirements for the

 

PhD. degree in Civil ErLgineerinL
7.7 '
J/“i/ng'VW

 

, Major Professor’s Signature

 

%}(LAZ(1 27L 026,05.

Date

MSU is an Afﬁrmative Action/Equal Opportunity Institution

PLACE IN RETURN BOX to remove this checkout from your record.
T O AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE

DATE DUE

DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

2/05 WW“

WOOD-CEMENT PARTICLEBOARD: IMPROVED MANUFACTURING,
MATERIAL CHARACTERIZATION AND POTENTIAL APPLICATION IN
CONCRETE CRASH BARRIERS
By

An Gong

A DISSERTATION

Submitted to
Michigan State University
in partial fulﬁllment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY
Department of Civil and Environmental Engineering

2005

ABSTRACT
WOOD-CEMENT PARTICLEBOARD: IMPROVED
MANUFACTURING MATERIAL CHARACTERIZATION, AND

POTENTIAL APPLICATION IN CONCRETE CRASH BARRIERS

By

An Gong

Recycling waste wood into wood composite products is increasingly attractive to
reduce environmental problems associated with disposal and promote sustainable
development. For use in engineering applications, in addition to developing suitable
manufacturing methods, the material and load bearing characteristics of wood-cement
particle composites must be thoroughly investigated. This research focuses on
wood-cement particleboard (WCPB). An improved manufacturing process that yields
superior wood-cement particleboard is developed; material properties required for
engineering design are experimentally determined; and the potential application of
wood-cement particleboard to enhance the impact characteristics of concrete highway
crash barriers is investigated.

A pressure controlled, fully saturated curing process is developed to enhance the
bond between wood particles and the cement matrix to produce a superior wood-cement

particleboard. The bending, compressive and impact characteristics of wood-cement

particleboard produced using this process are experimentally determined. The optimum
cement/wood ratio also is determined.

The experimental results indicate that WCPB is able to absorb a signiﬁcant amount
of energy by undergoing large nonlinear deformations. With the aim of taking advantage
of its energy absorbing capability, the potential application of WCPB to enhance the
impact characteristics of concrete highway crash barriers is investigated.

Computational simulation is performed using the LS-DYNA ﬁnite element
analysis code. The FE simulation agrees reasonably well with impact test results obtained
in the laboratory. This suggests that the LS-DYNA material models can be effectively
applied to model concrete and wood-cement particleboard.

Subsequently, vehicle crash simulations are performed on a concrete crash barrier
and a concrete crash barrier with WCPB facades to determine whether WCPB can
enhance the impact characteristics. It is found that a ﬂat WCPB facade is too stiﬂ‘ relative
to the ﬂexibility of the vehicle bumper system and does not yield a signiﬁcant reduction
in the impact deceleration force on the vehicle. However, a tubular facade made of
WCPB yields reasonable reductions in impact force. A triple tube facade appears to yield
the best impact characteristics, assures controllable deformations and provides optimal

safety to the vehicle occupants.

To my mother

iv

ACKNOWLEGEMENTS

The author would like to thank Professors Ronald Harichandran and Pascal
Kamdem for their academic guidance and ﬁnancial support in seeing the project through
to completion. The author would like to thank Professors Rigoberto Burgueﬁo,
Zhengfang Zhou and Parviz Soroushian for their academic guidance. Thanks also to

Siavosh Ravanbakhsh for his support in laboratory experiments.

TABLE OF CONTENTS

LIST OF TABLES ..................................................................................................... x
LIST OF FIGURES .................................................................................................. xii
CHAPTER] OVERVIEW AND RESEARCH OBJECTIVES ...................... 1
1.1 Wood-Cement Particle CompOSItesl
1.2 Concrete Barners3
1.3 Research Objectives ........................................................................ 4
CHAPTER 2 INRODUCTION AND LITERATURE REVIEW ....................... 5
2.1 Introduction ..................................................................................................... 5
2.2 History and Development of Wood-Cement Composites ................................ 5
2.3 Environmental Risk of CCA Treated Wood ................................................ 7
2.4 Wood and Cement ........................................................................................... 10
2.4.1 Wood .......................................................................................................... 10
2.4.2 Cement .......................................................................................................... 16
2.4.3 Wood cement compatibility ........................................................................ 17
2.5 Bending Strength of Wood-Cement Particleboard ....................................... 19
2.5.1 Elasto plastic behavior .................................................................................. 19
2.5.2 Effect of cement to wood ratio ................................................................... 22
2.5.3 Effect of manufacturing method on strength ........................................... 23
2.6 Compressive Strength of Wood-Cement Particleboard ........................................ 24
2.7 Durability ....................................................................................................... 26
2.8 Insulation Properties .................................................................................. 29
2.8.1 Thermal properties .................................................................................. 29
2.8.2 Acoustic properties .................................................................................. 30
2.8.3 Fire resistance ........................................................................................ 30
2.9 Theoretical Studies ...................................................................................... 31
2.10 Applications of Wood-Cement Composites ................................................ 33
2.11 Highway Barriers .................................................................. 38
2.11.1 Introduction ...................................................................... 38
2.11.2 Barrier safety data ................................................................................. 40
2.11.3 Crash performance of concrete barriers ..................................................... 41
2.12 Concrete Material Model ................................................................................ 44
2.12.1 Biaxial and triaxial behavior .................................................................. 45
2.12.2 Characteristics of yield surface ...................................................................... 49

vi

2.12.3 Failure surfaces of concrete under different stress states .................... 50
2.12.4 Evolution of yield surface of concrete under muti-dimensional

loading ........................................................................................................... 55

2.13 Finite Element Modeling for Vehicle Crash ........................................................ 56
2.13.1 Brief summary of formulation ..................................................................... 57

2. 13.2 Time integration ............................................................................................ 58
2.13.3 Contact algorithm .......................................................................................... 58
2.13.4 Force and displacement ﬁeld quantities ........................................................ 59
2.13.5 Simulations of barrier crash tests using LS-DYNA ..................................... 59

CHAPTER3 IMPROVED MANUFACTURING PROCESS AND

FLEXTURAL PERFORMANCE ...................................................... 62
3.1 Introduction ................................................................................. 62
3.2 Pressure Controlled and Fully Saturated Curing Process .......................... 63
3.2.1 Advantages of pressure controlled and fully saturated curing
process .............................................................................. 63
3.2.2 Materials ............................................................................ 64
3.2.3 Manufacturing particleboard ............................................... 66
3.3 Flexural Properties of Wood-Cement Particleboard ............................... 69
3.3.1 Experimental design ................................................................. 69
3.3.2 Description of test ................................................................ 69
3.3.3 Test results and analysis ........................................................... 71
3.4 Discussion .............................................................................. 84

CHAPTER 4 COMPRESSIVE BEHAVIOR OF WOOD-CEMENT

PARTICLEBOARD .......................................................... 86

4.1 Introduction .......................................................................... 86
4.2 Experimental Design ................................................................. 87
4.3 Compression Test ........................................................................ 88
4.3.1 Test method ......................................................................... 88
4.3.2 Failure modes ................................................................. 90
4.3.3 Compressive strength and stiffness .............................................. 94
4.3.4 Toughness .......................................................................... 98
4.4 Summary ................................................................................ 103

CHAPTER 5 IMPACT TESTS ON WOOD-CEMENT PARTICLEBOARD. . ..105

5.1 Introduction ....................................................................... 105
5.2 Falling Ball Impact Test ......................................................... 106
5.2.1 Impact test design ............................................................. 106
5.2.2 Design of impact test device ...................................................... 107

vii

5.2.3 Impact data recording devices ............................................. 110

5.2.4 Impact test setup .............................................................. 111
5.3 Multi-Stress Test on Wood-Cement Particle Tube ................................. 120
5.3.1 Test results and analysis ..................................................... 121
5.4 Summary ................................................................................ 125

CHAPTER 6 VALIDATION OF MATERIAL MODEL FOR

WOOD-CEMENT PARTICLEBOARD ............................... 126

6.1 Introduction ......................................................................... 126
6.2 Finite Element Modeling of Falling Ball Impact Test ............................ 126
6.2.1 Modeling software ................................................................ 126
6.2.2 Element modeling ................................................................... 127
6.2.3 Material modeling ................................................................. 127
6.2.3.1 Modeling concrete ......................................................... 127
6.2.3.2 Modeling wood-cement particleboard ................................... 131
6.2.3.3 Modeling steel ................................................................. 135

6.2.4 Contact modeling ................................................................. 135
6.2.5 Boundary conditions ............................................................ 136
6.2.6 Database output node .............................................................. 137
6.2.7 Impact simulation and results ................................................. 137
6.3 Finite Element Modeling of Multi-Stress Test ..................................... 142
6.3.1 Element modeling ............................................................... 142
6.3.2 Material modeling ................................................................ 143
6.3.3 Contact modeling ................................................................ 143
6.3.4 Boundary conditions ............................................................. 144
6.3.5 Database output node ....................................................... 145
6.3.6 Multi-stress test simulation ...................................................... 145
6.4 Summary ............................................................................. 147

CHAPTER 7 FINITE ELEMENT SIMULATION OF VEHICLE
CRASHING ON CONCRETE BARRIER ENHANCED

WITH WCPB ................................................................ 148

7.1 Introduction ............................................................................. 148
7.2 Simulation Test Design and Evaluation Criteria ................................. 149
7.2.1 Concrete barrier .................................................................. 149
7.2.2 Enhanced concrete barrier ......................................................... 150
7.2.3 Simulation matrix ............................................................. 151
7.2.4 Evaluation conditions and criteria ............................................. 152
7.3 Finite Element Modeling ........................................................... 152
7.3.1 Element modeling ............................................................ 153
7.3.2 Material modeling ............................................................ 154

viii

7.3.2.1 Modeling concrete ......................................................... 154

7.3.2.2 Modeling wood-cement particleboard .................................... 155
7.3.2.3 Modeling steel ................................................................. 155
7.3.3 Contact modeling .................................................................. 156
7.3.4 Vehicle model ....................................................................... 157
7.3.5 Center of gravity of the vehicle .................................................. 157
7.3.6 Boundary conditions ........................................................ 158
7.4 Finite Element Simulation ............................................................. 159
7.4.1 Frontal crash simulation setup ................................................. 160
7.4.2 Simulation results .................................................................... 163
7.4.3 Comparison of WCPB, concrete and steel triple-tube facades .............. 176
7.5 Summary ................................................................................... 180

CHAPTER 8 CONCLUSIONS AND RECOMMENDATIONS FOR

FUTURE RESEARCH ..................................................... 181
8.1 Introduction ............................................................................... 181
8.2 Development of wood-cement particle composites .................................. 181
8.2.1 Manufacturing process .......................................................... 181
8.2.2 Effect Of cement/wood ratio on bending strength .............................. 182
8.2.3 Compressive properties ............................................................ 182
8.2.4 Impact Characteristics .................................................... 183
8.3 Computational simulation for the application of wood-cement
composites ............................................................................... 184
8.4 Future research ............................................................................ 185
APPENDICES ................................................................................... 186
Appendix A An 820C vehicle impacting a concrete barrier enhanced with
a 3-tube wood-cement composite panel .......................... 187
Appendix B A steel ball impacting wood cement particleboard overlaying
a concrete block ......................................................... 199
BIBLIOGRAPHY ........................................................................... 204

ix

Table 2.1

Table 2.2

Table 2.3

Table 2.4

Table 2.5

Table 2.6

Table 3.1

Table 3.2

Table 3.3

Table 4.1
Table 4.2

Table 4.3

Table 4.4

Table 5.1

Table 6.1

Table 6.2

Table 6.3

LIST OF TABLES

Industrial Production of Inorganic-Bonded Wood Composites ................ 6
Area of Forestland in the World by Geographic Location and Species

Group 1980 ............................................................................ 11
Elemental composition of wood .......................................... 11
Organic Constituents of Wood ................................................... 12
Differences between hardwood and softwood .............................. 12
Rate of crash death on vehicle-barrier crash ..................... 41
Test Matrix for Bending Test ....................................................... 69
Flexural Performance of Wood-Cement Particleboard Subjected to

Seven Different Cement/Wood Ratios ............................................ 75
Flexural Toughness of Wood-Cement Particleboard ........................... 80
Test Matrix for Compression on Specimens .................................... 87

Compressive Strength and Strain at Peaks stress for WCPB and Normal

Concrete ............................................................................... 97
Stiffness of Wood-Cement Particleboard .................................. 98
Compressive Toughness of Wood-Cement Particleboard ................. 99
Test Matrix for Impact Test ..................................................... 107
Equation of State Algorithm for Generic Concrete ........................... 129
Concrete Material Parameters .................................................. 130
WCPB Material Data ............................................................ 134

Table 6.4

Table 7.1

Table 7.2

Table 7.3

Table 7.4

Material Data for Steel Ball ................................................... 135
Simulation Matrixes for Frontal Crash Impact ................................. 151
Concrete Material Parameters .................................................... 154
WCPB Material Data ............................................................... 155
Steel Bar Material Data ............................................................ 156

xi

Figure 2.1
Figure 2.2
Figure 2.3
Figure 2.4
Figure 2.5

Figure 2.6

Figure 2.7

Figure 2.8
Figure 2.9
Figure 2.10
Figure 2.11
Figure 2.12

Figure 2.13

Figure 2.14

Figure 2.15

LIST OF FIGURES

Softwood: Sitka sprue (x 75) ............................................................... 13
Hardwood: Red oak (x 50) .................................................................... 13
Most common forms of wood composites elements ................................. 14
Element of a knife-ring ﬂaker ................................................................... 15
Main component and plates of disk reﬁner ........................................ 16

Typical load-deformation curve of wood-cement particleboard under
bending ...................................................................................................... 20

F lexural stress vs. deﬂection curves under ﬂexure for WCPC with

normal and accelerated curing ................................................................... 21
Variation of MOR with cement-wood ratio ................................ 23
Compressive strength vs. wood/cement ratio ........................................... 25
Compressive load vs. deformation ...................................................... 26
F lexural strength vs. deﬂection under wetting and drying ........................ 28
F lexural strength vs. deﬂection under freezing and thawing .................. 28

Schematic representation of cracking through a ﬁber reinforced

matrix ......................................................................................................... 32
Siding with WCPB in Japan ..................................................................... 33
Siding with WWCB in the Phillipines ............................................ 34

xii

Figure 2.16a

Figure 2.16b

Figure 2.17
Figure 2.18
Figure 2.19
Figure 2.20

Figure 2.21

Figure 2.22
Figure 2.23
Figure 2.24
Figure 2.25
Figure 2.26
Figure 2.27
Figure 2.28
Figure 2.29
Figure 2.30

Figure 2.31

Figure 2.32

Figure 2.33

WCP block for building walls ................................................................... 35
Exterior and interior walls with WCP block .............................................. 35
Sound barrier of wood-cement particleboard (Czech Republic) ............... 36
Wood wool cement board for rooﬁng and walls in Phillippines ............... 36
Wood strand cement board for all walls and sidings in Japan ................... 37
Roof decking with wood excelsior cement composites ....................... 37

Exterior-wall cladding of the Sandoz AG Laboratory in Basel,

Switzerland ................................................................................................ 38
Flexible cable barrier ................................................................................ 39
Simi-rigid guardrail .................................................................................... 39
Rigid concrete barrier ................................................................................ 40
Major shapes of concrete barriers ........................................................... 42
Geometry of concrete barrier ..................................................................... 43
Demonstration of field test (top view) ....................................................... 43
Biaxial strength envelope of concrete ...................................... 46
Envelope of elastic limit and maximum strength of concrete ................ 47
Behavior of concrete in hydrostatic compression test ............................... 48

Schematic representation of the elastic limits and failure surface of
concrete .................................................................................................... 49

Geometric representation of the stress state in the principal stress
space ........................................................................................................... 50

Meridian and deviatoric sections for two parameter failure models .......... 52

xiii

Figure 2.34
Figure 2.35
Figure 2.36
Figure 2.37
Figure 2.38
Figure 2.39
Figure 3.1
Figure 3.2
Figure 3.3
Figure 3.4
Figure 3.5
Figure 3.6
Figure 3.7
Figure 3.8
Figure 3.9
Figure 3.10
Figure 3.11
Figure 3.12
Figure 3.13
Figure 3.14

Figure 3.15

Hsieh-Ting-Chen criterion ......................................................................... 54
Evolution of concrete yield surface as proposed by Chen and Han .......... 56
Crash ﬁeld test (ﬁont view) ...................................................................... 60
Crash Simulation (front view) ................................................................... 60
Crash ﬁeld test (plan view) ........................................................................ 61
Crash simulation (plan View) .................................................................... 61
CCA-treated wood .................................................................................... 65
Wood particles ........................................................................................... 65
Wood particle cutting machine ................................................................. 65
Wood particle screener .............................................................................. 65
Cement wood particle mixer ...................................................................... 67
Hydraulic press .......................................................................................... 67
Curing in water ......................................................................................... 67
Saw used to cut board ................................................................................ 67
Flowchart of manufacturing process .......................................................... 68
INSTRON machine ................................................................................... 70
Bending test setup ................................................................................... 70
Specimen’s size .......................................................................................... 70
Bending failure ........................................................................................... 72
Cross-section view of failure .................................................................. 72
Load-deﬂection curves for varying cement/wood ratios ...................... 73

xiv

Figure 3.15
Figure 3.16
Figure 3.17
Figure 3.18
Fihure 3.19

Figure 3.20

Figure 3.21

Figure 3.22

Figure 3.23

Figure 3.24
Figure 3.25
Figure 4.1
Figure 4.2
Figure 4.3
Figure 4.4
Figure 4.5
Figure 4.6
Figure 4.7

Figure 4.8

Load-deﬂection curves for varying cement/wood ratios ...................... 73
Correlation between MOR and cement/wood ratio .................................. 76
Correlation between MOE and cement/wood ratio ................................... 76
Correlation between density and cement/wood ratio ................................. 77
Toughness calculation curve (load vs. disp) ............................................ 79

Variation of modulus of rupture vs. cement/wood ratio
(95%conﬁdence) .................................................................................... 8 1

Variation of modulus of elasticity vs. cement/wood ratio
(95% conﬁdence) ...................................................................................... 82

Modulus of elasticity vs. cement/wood ratio ............................................. 82

Variation of density vs. cement/wood ratio

(95% conﬁdence) ....................................................................................... 83
Density vs. cement/wood ratio ................................................................... 83
Toughness index vs. cement/wood ratio (95% conﬁdence) ...................... 84
Short column specimens ............................................................................ 87
Loading directions .................................................................................... 88
Compression speciemens ........................................................................... 89
Compression test set-up on INSTRON test machine ................................. 89
Failure modes (h/d=2.0, cement/wood ratio = 1.5) ................................... 91
Failure modes (h/d =20, cement/wood ratio = 1.0) .................................. 92
Failure modes (h/d =3.0, cement/wood ratio = 1.5) .................................. 93
Load-deformation curves (h/d=2.0, cement/wood ratio = 1.5) .................. 94

XV

Figure 4.9
Figure 4.10

Figure 4.11

Figure 4.12

Figure 4.13

Figure 4.14

Figure 4.15

Figure 4.16
Figure 5.1
Figure 5.2
Figure 5.3
Figure 5.4
Figure 5.5
Figure 5.6
Figure 5.7

Figure 5.8

Figure 5.9

Figure 5.10

Load deformation curves (h/d =2.0, cement/wood ratio = 1.0) ................. 95
Load deformation curves (h/d =3.0, cement/wood ratio = 1.5) ................. 95

Stress-strain relationships of the wood-cement composite short
columns .................................................................................................... 100

Compressive strength vs. cement/wood ratio (parallel loading) ............. 101

Compressive strength vs. cement/wood ratio (perpendicular
loading) .................................................................................................... 101

Compressive stiffness vs. cement/wood ratio (parallel loading) .......... 102

Compressive stiffness vs. cement/wood ratio (perpendicular

loading) .................................................................................................... 102
Compressive toughness index vs. cement/wood ratio ........................... 103
Schematics of impact test ..................................................................... 107
Release mechanism ................................................................................ 108
Schematic of the impact ball .................................................................... 109
Photographs of impact ball assembly .................................................. 109
353312 accelerometer ............................................................................. 110
Oscilloscope ............................................................................................. 110
Impact test setup .................................................................................... 111

Permanent deformations in wood-cement particleboard following

impact ....................................................................................................... 112
Voltage vs. time (concrete) .................................................................. 113
Voltage vs. time (WCPB on concrete) .................................................. 113

xvi

Figure 5.11
Figure 5.12
Figure 5.13
Figure 5.14
Figure 5.15
Figure 5.16
Figure 5.17
Figure 5.18
Figure 5.19
Figure 5.20
Figure 5.21
Figure 5.22
Figure 6.1

Figure 6.2

Figure 6.3
Figure 6.4
Figure 6.5
Figure 6.6
Figure 6.7

Figure 6.8

Voltage vs. time (wood on concrete) .................................................. 114
Acceleration time histories of steel ball (WCPB on concrete) .............. 115
Velocity time histories of steel ball (WCPB on concrete) ....................... 116
Displacement time histories of steel ball (W CPB on concrete) ............... 117
Acceleration time histories of steel ball (concrete, wood, WCPB) ......... 118
Velocity time histories of steel ball (concrete, wood, WCPB) ................ 119
Displacement time histories of steel ball (concrete, wood, WCPB) ........ 120
Geometry of speciemen .......................................................................... 121
Compression test setup ............................................................................ 122
Compressive failure of WCPB ring ......................................................... 123
Load-displacement curve for multi-stress test on WCPB ring ............... 123
Stress state and stress distribution ............................................................ 124
Hydrostatic stress vs. volumetric strain of concrete ................................ 128

Yield stress vs. mean stress for concrete under hydrostatic

compression ............................................................................................. 130
Effective stress vs. effective plastic strain curve ..................................... 132
Engineering stress vs. engineering strain (W CPB) ................................. 133
Effective stress vs.effective plastic strain (WCPB) ................................ 134
FE model and boundary conditions ...................................................... 136
Output node location .............................................................................. 137
Finite element models .............................................................................. 138

xvii

Figure 6.9

Figure 6.10
Figure 6.11
Figure 6.12
Figure 6.13
Figure 6.14
Figure 6.15
Figure 6.16
Figure 6.17
Figure 7.1
Figure 7.2
Figure 7.3
Figure 7.4
Figure 7.5
Figure 7.6
Figure 7.7
Figure 7.8
Figure 7.9
Figure 7.10

Figure 7.11

Acceleration time histories ﬁ'om impact test and simulation
(concrete) ................................................................................................. 140

Acceleration time histories from impact test and simulation (WCPB)....141

Velocity time histories from impact test and simulation ......................... 141
Displacement time histories from impact test and simulation ................. 142
Boundary conditions ............................................................................... 143
Location of output node ........................................................................... 144
Finite element model ................................................................................ 145
Deformed FE mode] ................................................................................. 146
Load—displacement curves from test and simulation ...................... 147
Geometry of the concrete barrier and steel reinforcement ................ 149
Enhanced vertical concrete barrier ........................................... 150
Cross section of particleboard panels ........................................ 151
Concrete and steel bar elements ................................................ 153
WCPB and concrete elements .................................................. 153
Mesh of steel reinforcement with beam element .......................... 154
1975 Honda Civic (820C) ...................................................... 157
CG and accelerometer location ................................................. 158
Boundary conditions of the FE model ...................................... 159
Vehicle crashing on concrete barrier ........................................... 160
Concrete barrier enhanced with ﬂat WCPB facade ........................ 161

xviii

Figure 7.12
Figure 7.13
Figure 7.14
Figure 7.15
Figure 7.16
Figure 7.17
Figure 7.18
Figure 7.19
Figure 7.20
Figure 7.21
Figure 7.22
Figure 7.23

Figure 7.24

Figure 7.25
Figure 7.26
Figure 7.27
Figure 7.28
Figure 7.29
Figure 7.30

Figure 7.31

Concrete barrier enhanced with double-wave WCPB facade .......... 161
Concrete barrier enhanced with triple-wave WCPB facade .............. 162
Concrete barrier enhanced with triple-tube WCPB facade ................. 162
Concrete barrier enhanced with four-tube WCPB facade .................. 163
Crash on concrete barrier ........................................................ 165
Crash on ﬂat facade .......................................................... 165
Crash on double-wave facade .................................................................. 166
Crash on triple-wave facade ................................................................... 166
Crash on triple-tube facade .................................................................. 167
Crash on four-tube facade ........................................................................ 167
Vehicle damage after crash on concrete barrier (at t=70 ms) ................. 168
Vehicle damage after crash on ﬂat WCPB facade (at t=7 0 ms) .............. 168

Vehicle damage after crash on double-wave WCPB facade
(at t=70 ms) .............................................................................................. 169

Vehicle damage after carsh on triple-wave WCPB facade (at t=70 ms)..169
Vehicle damage after carsh on triple-tube WCPB facade (at t=70 ms)...170

Vehicle damage after crash on four-tube WCPB facade (at t=7O ms). . .170

Displacement time histories of vehicle ...................................... 172
Velocity time histories of vehicle ............................................. 172
Acceleration time histories of vehicle ......................................... 173
Deformation of triple-tube facade ............................................. 175

xix

Figure 7.32
Figure 7.33
Figure 7.34
Figure 7.35

Figure 7.36

Vehicle crashing on WCPB triple-tube facade ............................... 177

Vehicle crashing on concrete triple-tube facade ......................... 177
Vehicle crashing on 3mm thick steel triple-tube facade ................ 178
Vehicle crashing on 5 mm thick steel triple-tube facade ................ 178

Acceleration time histories of vehicle for four triple-tube facades...l79

XX

CHAPTER 1

OVERVIEW AND RESEARCH OBJECTIVES

1.1 Wood-Cement Particleboard

Wood-cement particleboard is composed of wood particles, Portland cement (Type
I) and water. Besides having higher material strength, good plasticity, good durability and
better energy dissipation properties, it also has low cost, is lightweight and simple to
manufacture (Kurﬁel, 1995; Wolfe, 1996). It has been shown in practice that the
wood-cement particleboard has advantages over conventional wood composites.

At present, approximately 10,000 cubic meters of treated wood are removed from
service in the United States and Canada every year. This ﬁgure is expected to increase
in the next decade (Copper, 1993). However, only a small part of this large amount of
waste wood is recycled into products, and most of the waste wood is discarded producing
a negative effect on the environment. Therefore, recycling waste into wood composite
products is highly beneﬁcial.

The tensile strength of wood-cement particleboard (WCPB) is about 3 times higher
than normal concrete (Won, 1995, Moslemi and Pﬁster, 1987) and the compressive
strength of WCPB is comparable to that of normal concrete (Blankenhor, 1994; Sorfa,
1984)

Because WCPB has high porosity due to the presence of soft wood particles, it is

capable of dissipating much more energy than concrete. The ﬁrst toughness index of

1

WCPB for both bending and compression is six times higher than normal concrete (Wolfe,
1996)

WCPB has advantages over conventional resin-bonded wood particle composites
with respect to durability and ﬁre resistance. WCPB meets the requirements for ﬁre
resistance and durability for outdoor applications (Wolfe, 1999). However, conventional
resin-bonded wood particle composites do not meet the requirements of outdoor
applications for either durability or ﬁre resistance.

Because WCPB is mainly made from waste wood removed from service, there is
no cost for acquiring the raw material, and the manufacturing process for WCPB is quite
simple. Use of waste wood to produce WCPB is environmentally ﬁiendly and beneﬁcial
to society.

Even though WCPB technology was introduced about 40 years ago, it has not
been rigorously investigated, especially in North America (Moslemi, 1989). As the use of
WCPB material increases in The United States, more effort is being directed to research.

Recent studies (Won, 1995; Wolfe, 1999; Zhou and Karndem, 2000; Horne, 1997)
on wood cement particle composites have focused on manufacturing methods, material
properties, durability and potential applications in engineering. The manufacturing
method can signiﬁcantly impact the material properties of wood-cement particleboards.
The development of improved methods that yield WCPB with higher strength and
stiﬂiiess and reduce variability in material properties would be beneﬁcial.

With the increasing applications of wood-cement particleboard in construction

2

purposes, more and more research has focused on the study of the strength of the
composite material and the effect of the cement/wood ratio on its material strength (Zhou
and Kamdem, 2000; Eltomation, 2003).

Perhaps of most relevance is the development of new high volume applications for
the large amount of waste wood that is generated in the US. Successful applications of
using wood-cement particle composites as exterior walls in houses (Home, 1997) and as
sound barriers on highways (Wolfe, 1999) have begun a potential bright future for the
application of this material in engineering.

At present, WCPB products have been widely used in construction engineering as
both structural and nonstructural members (Kurpiel, 1995; Home, 1997). Because WCPB
has energy dissipating properties, it could also have special applications in structures

where impact and dynamic load are a design consideration (Wolfe, 1999).

1.2 Concrete Crash Barriers

Highway engineers are facing a signiﬁcant challenge to improve the impact
characteristics of concrete crash barriers to make the roads safer. Concrete crash barriers
are used in the longitudinal direction of roads. Because the deﬂection of concrete barriers
during impact is very small, the impact deceleration forces on vehicles impacting them
are much higher than those on vehicles impacting other kinds of barriers such as
guardrails. Therefore, enhancements to concrete barriers that reduce deceleration forces

on impacting vehicles are highly desirable to improve safety for vehicle occupants.

3

1.3

Research Objectives

The objectives of this research are to:

Develop a new manufacturing method to improve the engineering properties of
wood-cement particleboard.

Characterize the response of WCPB under bending, compressive and impact loads.

Assess the applicability of material models available in the LS-DYNA ﬁnite
element package to simulate impact on WCPB.

Use ﬁnite element simulations to investigate the feasibility of using
wood-cement particleboard facades to enhance the impact characteristics of

concrete crash barriers.

CHAPTER 2

INTRODUCTION AND LITERATURE REVIEW

2.1 Introduction

Wood cement particle composites have been used in the manufacturing of
building materials for more than 60 years (Wolfe, 1997). As a construction material, the
composite products have primarily been used as nonstructural members, and applications
have mainly been based on the non-structural advantages of the composite: cost
effectiveness, resistance to decay and insects, ﬁre resistance, and thermal insulation. In
the last decade, the composite has been successfully used for structural members in
construction, and research on its material properties and its potential in engineering
applications has increased (Kurﬁel, 1995; RBAC Newsletter, 2002).

The literature review presented in this chapter focuses mainly on the description
of the development of the technology of wood-cement composites, and is followed by
some concrete crash barriers and ﬁnite element simulation in highway crash barrier

engineering.

2.2 History and Development of Wood-Cement Composites

The concept of combining inorganic binders with wood or agricultural biomass
can be traced back several thousands of year to ancient ancient time. Incorporating wheat
or rice straw with mud to produce bricks has it roots in prehistory. Such inorganic-

bonded composites are still used in many parts of the world. Mixing straw with mud

improves workability, impact resistance and reduces fracture development by providing
reinforcing ﬁbers (Won, 1995).

The timeline for the development of inorganic-bonded wood composites in
modern times is shown in Table 2.1. The term “inorganic-bonded wood composites”
generally refers to mineral-bonded wood products where the minerals mainly include

magnesite, gypsum and cement.

Table 2.1 Industrial Production of Inorganic-Bonded Wood Composites (Moslemi, 1988)

 

 

Year Production

1900 Magnesite-bonded boards

1905 Gypsum-bonded excelsior board

1915 Magnesite-bonded excelsior board
1915 Cement-bonded excelsior board

1927 Molded wood-cement products

1937 Resin-bonded particleboards

1942 Cement-bonded wood composite panels
1965 Gypsum ﬁberboards (shredded paper)
1972 Magnesite-bonded wood composite panel
1982 Gypsum particleboard

 

Wood-cement particle composites are a special type of wood-cement composites.
There are three main types of wood-cement composites: (1) wood-wool cement
composites (WW CC), (2) wood-ﬁber cement composites (WFCC) and (3) wood-cement
particleboard (W CPB).

The WWCC was ﬁrst developed in Europe in the 19208 (Moslerni and Hamel,
1989) and it is manufacturated in many countries around the world. The WFCC were
developed primarily as a substitute for asbestos-cement and have a history of about 25 to

30 years (Wolfe, 1997). Research in Europe in the 19203 led to the common practice of

using wood chips in cement to make building blocks (Wolfe, 1997). This application is
perhaps the origin of WCPB. However, as a product in commercial industry, WCPB
was developed in the 1960s. The industry was stimulated in Europe by lumber shortages
and in the United States by large quantities of waste softwood from wood manufacturing
industries. The technology of making wood cement composites has many similarities to

that used to manufacture resin-bonded wood particleboard (Moslerni and Hamel, 1989).

2.3 Environmental Risk of CCA-Treated Wood

Chromated copper arsenate (CCA) is a water-borne preservative containing
arsenic, chromium and copper. It is used for the long—term protection of wood against
attack by fungi, insects and marine borers. Applying wood preservatives extends the
productive life of wood and helps reduce demand for forest resources. Prior to the wood
being treated, small cuts are made into the wood surface to help the penetration of the
CCA preservative solution. The wood is stacked and loaded into a sealed chamber where
the wood’s natural moisture is removed by vacuum. Then the chamber is pressurized,
forcing the CCA solution into the wood. A vacuum is then used to extract any excess
solution. The wood is removed ﬁom the chamber and dried (Health Canada, 2003).

Waterborne preservatives today account for the vast majority of chemicals used in
building applications. These waterborne chemicals are used in pressure treating only
include chromated copper arsenate (CCA), ammoniacal copper zinc arsenate (ACZA),
and several others. CCA accounted for 94% of the waterborne chemicals used in 1995.
While the primary chemicals in CCA are highly toxic and are regulated as pesticides by

the Environmental Protection Agency (EPA), they are bonded quite tightly to the wood

during ﬁxing, so leaching from the wood is minimal under normal usage conditions
(EBN, 1997).

There are some important beneﬁts of treating wood with CCA, the most
signiﬁcant of which is increasing the life of wood for outdoor applications. Older
preservatives, such as pentachlorophenal and creosote, pose considerable health risk to
users of the wood, but new waterborne preservatives, such as CCA, are safer. They are
safer in part because the toxins have a strong chemical bond with wood and are ﬁxed in
the wood. However, disposing of treated wood at the end of its life has become a serious
problem (ENB, 1997).

CCA-treated wood is used primarily for structures intended for outdoor settings
and is commonly used in residential construction such as playground structures, fences,
gazebos and decks. It is also used for industrial uses such as utility and construction
poles, marine timbers and pilings.

In 1993, an estimated 5 billion board feet, or 6.5 million tons, of wood were CCA
treated and will be removed from service after approximately 27 years in 2020. This
ﬁgure does not include the 72 million utility poles or 15 million ties treated annually with
creosote or pentachlorophenol or other oilbome preservatives (Stalker, 1993;
Mickleright, 1993). A major challenge facing the wood preservative industry today is
how to deal with this massive quantity of waste wood in the coming decade.

Based on a 30-year service-life, researchers at the Forest Products Laboratory
(FPL) in Madison, Wisconsin estimate that 2.5 billion board feet per year (6 million

m3/year) of treated-wood products (all types of preservative treatments) are currently

entering the solid-waste stream, and that level will increase to 8 billion board feet per
year (19 million m3/yr) by the year 2020 (EBN, 1997).

The environmental hazards of CCA-treated wood are related to the high toxicity
of copper, chromium and arsenic to non-target organisms in freshwater and marine
environments. The potential risk depends on the exposure, i.e., the amount of CCA
being leached from treated wood into the aquatic ecosystem, directly or indirectly. In
general, small ground-based structures made from CCA-treated wood that are properly
treated and ﬁxed are unlikely to cause any environmental hazard. There is some data that
indicates that some sealants can reduce surface dislodgeability and leachability of the
preservative chemicals. This data is being considered as part of the CCA reevaluation.
The environmental risk of CCA-treated wood coming into contact with freshwater and
marine environments (including wetlands) is greater than in terrestrial environments and
possible restrictions on aquatic use are currently being assessed.

Recently, the US. Consumer Product Safety Commission (CPSC) released a risk
assessment of CCA-treated wood on playground structures. This assessment concluded
that children playing on playground equipment built with CCA-treated wood might have
a slightly increased risk of developing cancer. However, at this time, no further action has
been recommended by the CPSC (Health Canada, 2003).

For the general public, on the one hand, disposal in landﬁlls may produce
environmental problems. It is increasingly difﬁcult to dispose solid wood in landﬁlls
because it is a high volume material and deterioration is questionable when necessary
requirements for biodegradation are not met. On the other hand, Stalker (1993) reported

a quoted cost by landﬁll operators of $150 US per utility pole. The cost of disposal and

the current level of environmental awareness will reduce land ﬁlling as an attractive
alternative for disposal of treated wood (Cooper, 1994; Marer and Grimes, 1992).

In addition, CCA-treated wood should never be burned because the heavy metals
in CCA are not destroyed; the chromium and copper become concentrated in the ash,
while the arsenic becomes a vapor that will escapes into air (ENB, 1997). NBC10.com
(2003) reported that a single 12-foot, 2 (in.) x 6 (in.) piece of lumber contains about 27
grams of arsenic — enough to kill 250 adults. Burning CCA releases the strong chemical
bond that holds the arsenic bound to the wood.

As a result, the disposal of preservative-treated wood products removed from
service is problematic. Therefore an attractive solution for the after service disposal of
treated wood is reconstituted wood products. Recycled treated wood or urban wood, if
properly managed, can be a good source of lumber or of ﬁber for engineered wood
products such as hardboard, ﬁberboard, particleboard and cement-bonded particleboard.
Recycling waste into wood composite products becomes increasingly attractive because it

not only beneﬁts the environment but also has economic value.

2.4 Wood and Cement

2.4.1 Wood

Wood is divided into two broad classes, usually refered to as “hardwood” and
“softwood”. Needle-like leaves characterize softwood trees. Such trees are known
commonly as evergreens. In contrast, hardwood trees have broad leaves, which generally
change color and drop in the fall in cold climatic zones.

Hardwood species grow in every major region of the United States. On the

worldwide basis, hardwood is dominant. They are found in most areas of the world

10

(Table 2.2). In total, hardwood growing stock has been estimated to exist in volumes

almost double that of softwoods. However, North America and the former USSR

softwood are dominant in total volume.

Table 2.2 Area of Forestland in the World by Geographic Location and Species Group—

1980 (Haygreen, 1989)

 

 

Area Total Softwoods Hardwoods
(thousand ha)
North America 807,092 533,075 274,017
Latin America 915,019 20,474 894,545
Europe 158,902 97,989 60,913
Aﬁica 743,713 13,891 729,822
Asia 468,230 102,238 365,992
USSR 928,600 679,900 248,700
Paciﬁc area 298,947 15,995 282,952
World 4,320,503 1,463,562 2,856,941

 

Source: FAO (1985).

Wood is composed principally of carbon, hydrogen, and oxygen. Table 2.3 shows

the chemical composition of a typical North American wood and shows carbon to be the

dominant element on a weight basis.

Table 2.3 Elemental Composition of Wood (Haygreen, 1989)

 

 

Element Percent of Dry Weight
Carbon 49
Hydrogen 6

Oxygen 44

Nitrogen slight amounts
Ash 0.1

 

The elemental constituents of wood are combined into a number of organic

polymers: cellulose, hemicellulose, and ligin. Table 2.4 shows the aproximate percent

of dry weight of each in hardwood and softwood.

ll

Table 2.4 Organic Constituents of Wood (Haygreen, 1989)

 

 

Wood Type Cellulose Hemicellulose Lignin
(% dry weight)

Hardwood 40-44 1 5-35 18-25

Softwood 40-44 20-32 25-35

 

Softwoods have a uniform arrangement of a few cell types and therefore usally do
not have a distinctive appearance. Hardwoods, on the other hand, are composed of
widely varying proportions of markedly different kinds of cells and thus often have a
unique appearance. The physical differences between hardwood and softwood are listed

in Table 2.5. Figure 2.1 and Figure 2.2 show the cell types of softwood and hardwood.

Table 2.5 Physical Differences Between Hardwood and Softwood (Haygreen, 1989)

 

 

Hardwood Softwood
Many types of cells few types of cells
Contain vessels no vessels
Wide rays (30% of wood volume) narrow rays ( 5-7% of wood volume)
Not aligned in straight radial rows Straight radial rowsof cells

 

12

Figure 2.1 Softwood: Sitka sprue (x 75), (Haygreen, 198

 

 

Latewood vessels

Earlywood vessels

Figure 2.2 Hardwood: Red oak (x 50), (Haygreen, 1989)

(Scaning electron micrograph by Joseﬁna Gonzalez, Forintek Canada Corp.
Vancouver, B.C.)

l3

Currently, the wood used in wood-cement composites are generally from
softwood. Figure 2.3 (Marta, 1972) shows the various sizes and shapes of wood elements
included in wood composites. For wood-cement composites, the wood materials in the

second and third rows in Figure 2.3 are generally used.

 

 

 

 

 

Figure 2.3 Most common forms of wood composites elements (Marra, 1972)

The wood particles are produced by a special cutting machine. A variety of
machines including reﬁners, hammers, ﬂakers, and waferizers can be used to produce the
type of furnish (particles used to make the board) desired. These machines grind, cut, or

tear the wood into a range of particle sizes. Figure 2.4 illustrate the parts of a ring ﬂaker.

This machine has knives that cut thin particles ﬁ'om whatever kind of wood is fed to it. A

general disk reﬁner is shown in Figure 2.5 and is used to prepare small ﬁberlike particles.

Front view

 

   

' Flakes Impeller

Knife ring

Figure 2.4 Element of a knife-ring ﬂaker (Haygreen, 1989)

  
  

- , Disk

Main components and
plate

plates of a disk reﬁner

Revolving
disk

:6

Gap
(adjustable)

 

Reﬁned
particles or
ﬁber

 

Typical disk plates

Figure 2.5 Main Component and plates of a disk reﬁner (Haygreen, 1989)

2.4.2 Cement

An inorganic binder which has found commercial use in the ﬁeld of certain wood
particle panels is Portland cement. This binder is used in the manufacture of products
intended for service in structural applications which may or may not be load bearing.
Portland cement contains four basic chemical elements, silicon, aluminum, iron, and
calcium, usually in oxidized forms (Moslemi, 1974).

Portland cement is a gray powder that consists of angular particles typically in the
size range of l to 50 pm. It is produced by pulverizing a clinker with calcium sulfate.
The clinker is a heterogeneous mixture of several minerals produced by high-temperature

reactions between calcium oxide and silica, alumina, and iron oxide. When Portland

cement is dispersed in water, the calcium sulfate and the high-temperature compounds of
calcium go into solution, and the liquid phase becomes rapidly saturated with various
ionic species. As a result of combinations between calcium, sulfate, aluminate, and
hydroxyl ions within a certain time of cement hydration, different crystals of calcium
chemical compounds are created and ﬁnally solid cement paste is formed.

Portland cement was formally used in manufacturing wood cement composites in
1928 (Maloney, 1977). Nowadays, in wood cement composites, Portland cement (Type I)

is generally used as a binder in the wood cement composites.

2.4.3 Wood cement compatibility

In wood-cement composites, the compatibility between wood and cement is a
major concern. Wood species are not compatible with cement because soluble
carbohydrates and certain extracts in the wood inhibit the setting and hardening of
cement. In fact, in some instances, the inﬂuence of such substances is so severe that the
cement binder is inhibited from setting at all regardless of the amount of time allowed.
This inhibitory effect appears to be more pronounced in hardwood compared to
softwoods due to the existence of the type of constituents that retard or prohibit element
setting (Moslemi, 1974).

To reduce the inhibitory effects of wood on cement setting, various measures
have been taken. One of these consists of extracting sugars and other water soluble
materials with hot water. The wood particles are soaked in hot water for a period of time.
The soak duration depends on a number of factors such as the temperature of the hot

water and the amount of inhibitory extractives present. The addition of various chemical

l7

additives also have been found useful in reducing setting time. A major compound used
in this connection is calcium chloride. Clare (1956) states that calcium chloride solutions
of 1 - 3 percent are necessary to neutralize the effect of 0.1 percent sugar in the soil
cement mixtures. Further increases in the level of calcium chloride, however, have been
found to reduce strength in such mixtures.

Miller (1987) studied the effects of carbohydates on the tensile strength and
hydration characteristics of Portland cement. He concluded that sugars and other
compounds in wood are detrimental to cement hydration and strength. When wood is
mixed with cement, the setting time of cement increases and the cement strength
decreases. However, if wood particles are treated with special chemicals, the
compatibility between wood and cement may be dramatically improved.

Several studies have suggested that mechanical properties of wood based
composites made of CCA treated ﬁber or particles are lower than those of untreated
ﬁrmish (Vicks, 1990; Gertjejansen et al., 1988; Suzuki, 1993). The reduction in
mechanical and physical properties was attributed to the surface modiﬁcation of the
woody materials by preservative treatment or the interaction between the preservatives
and the resin. However, for cement-bonded particleboard, Schmidt et a1. (1994) reported
that CCA treated red pine particles produced strong wood-cement composites compared
to untreated wood. Generally the CCA treated wood has a better compatibility with
cement, while, untreated wood has a poor compatibility with cement during cement
hydration.

Most wood waste generated by manufacturing industries is treated with CCA.

Consequently, in present-day wood-cement particleboard, the wood particles are

18

generally made ﬁom wood treated with CCA. Previous studies indicate that CCA-treated

wood has improved compatibility with cement (Schmidt, 1994; Hsu, 1994).

2.5 Bending Strength of Wood-Cement Particleboard

Most previous research on the mechanical characteristics of wood-cement
composites has focused on bending strength. Through bending tests, the modulus of
rupture (MOR) and modulus of elasticity (MOE) are investigated. In general, the MOR is
the primary measure of the quality of wood-cement particleboard.

Wolfe (1997) reported that the maximtun MOR of wood-cement particleboard is

typically less than 10 MPa.

2.5.1 Elasto-degrading plastic behavior

In general, wood-cement particleboard displays an elasto-degrading plastic
behavior under ﬂexural loadings due to the bridging action of wood particles (Wolfe,
1999). Figure 2.6 shows the characteristics of the load-deformation curve of wood-
cement particleboard under a three-point bending test. The response is elastic until the
ﬁrst crack forms, then displays a reduction in stiffness until the peak load. The load
carrying capability degrades after the peak load but large deformations are sustained

before failure.

19

ﬁ Load
Pmax
Maximum load

First Crack

 

 

Deformation

 

>

Figure 2.6. Typical load-deformation curve of wood-cement particleboard under bending

(Wolfe, 1999)

Won (1995) reported that the load-deﬂection curves of WCPB under bending

loads show elasto-plastic behavior both for specimens cured in air and those with

accelerated curing using C02. Figure 2.7 shows typical ﬂexural stress-deﬂection curves.

20

10.

 

 

,3 8 _ C02 - Aspen
9..
E .
V 6 _ C02 — Southern pine
fl)
8 ,
a 4 Control
'3 .-
'5
5
E. 2 _ .
o I I l l l
0 2 4 6 8 10
Deﬂection (mm)
(a) Wood/cement Ratio = 0.28 (Cement/wood ratio = 3.57)
10
C02 - Aspen
,3 C02 — Southern pine
o.
E
f
a
i
E:

 

 

 

Deﬂection (mm)

(b) Wood/cement Ratio = 0.35 (Cement/wood ratio = 2.86)

Figure 2.7 F lexural stress vs. deﬂection curves for WCPB with normal and
accelerated curing (Won, 1995)

21

2.5.2 Effect of cement/wood ratio

The cement/wood ratio is an important parameter in the manufacturing of wood-
cement particleboard because it impacts the weight to strength ratio and the cost. At
present, commercially manufactured boards are produced in the range of 2.75 to 3.0 parts
of cement to 1.0 part of wood (Bison and Creten, 1977).

Presternon (1976) indicated that WCPB had higher bending strength properties
when the cement/wood ratio was increased from 0.75 to 1.5.

Moslemi and Pﬁster (1987) reported that the MOR of wood-cement ﬂakeboard
increased as the cement/wood ratio was lowered from 3.0 to 1.5. In their study, the wood
particles were wood ﬂakes with a thickness of 0.56 mm and length and width ﬁom 12 to
25 mm. They used Portland cement types III and 1. Their test results are summarized in
Figure 2.8.

Lee (1985) reported that an increase of the cement/wood ratios above 2.0 had an
adverse effect on bending stiffness and strength of wood excelsior board.

Oyagade (1990) conducted bending tests with cement/wood ratios of 1.55, 2.33
and 3.10. This study indicated that at a constant board density, a decrease in

cement/wood ratio generally improved the bending properties.

22

MOR (psi)

__ TypeI- 14daycure
----—- TypeIII—l4daycure

   
  
   

2500 '

‘ ................. Type I _ 28 day cure
,,.....:“ ‘ — — — - Type III — 28 day cure

\“ \
\\
\‘ x

\ \ x“.-.---“-"--.”--_.

2400

2300

2200

2100

 

1.5 2.0 2.5 3.0

Cement/wood ratio
Figure 2.8. Variation of MOR with cement-wood ratio

(Moslemi and Pﬁster, 1987)

2.5.3 Effect of manufacturing method on strength

The manufacturing method for WCPB is being continually improved. Different
manufacturing methods can lead to signiﬁcant changes in the material properties of
WCBP, such as strength and durability. The following two major of manufacturing

methods are general used for WCPB:

a. Thickness control and laboratory curing at 65 % relative humidity.

b. Thickness control and C02 accelerated curing.

Thickness control means that the specimen is manufactured to a predetermined
thickness by pressing specimens in a hydraulic press.
Laboratory curing is the most common curing method. In this method, specimens
are cured at a temperature of 20 °C and 65% relative humidity for 28 days after they are

removed from the press.

23

In the accelerated C02 curing method, after a few hours of setting under pressure,
the specimens are steamed with CO2 for 5 to 8 hours for accelerated curing. This method
dramatically shortens the manufacturing period.

Won (1995) reported that the bending strength of wood cement particleboard
cured with C02 is much higher than those cured in air. Figure 2.7 shows the differences
in ﬂexural stress between C02-cured and air-cured (control) specimens.

Zhou and Kamdem (2001) reported that air curing at 20C° and 100% relative
humidity for 28 days produced specimens with higher bending strength than those cured

at 65 % relative humidity.

2.6 Compressive Strength of Wood-Cement Particle Composites

Studies on the compressive strength of WCPB have been much fewer than those
on ﬂexural strength. With the application of wood-cement particle bricks in loading
bearing walls, the compressive strength has recently received more attention.

Wolfe (1997 and 1999) reported that the characteristics of the compressive load-
deformation curve are similar to those for the bending test shown in Figure 2.6. In his
study, the load displacement plots exhibited linear behavior up to 60 to 75% of the
maximum load, at which point the rigid cement matrix began to crack. As the material

began to fail, the material exhibited plastic behavior.

24

Research on hardwood-cement particle composites (Blankenhor, 1994) indicated
that as the amount of red maple and hardwood pulp increased, the compressive strength

decreased. Figure 2.9 shows the compressive test results.

 

 

 

 

I a
.9 .
*5 I
S
3"
5
C
E
a
E
i 8 .
i 0 0.1 0.2 0.3 0.4 0.5
i Wood/Wrath

 

Figure 2.9. Compressive strength vs. wood/cement ratio (Blankenhor, 1994)

Sorfa (1984) reported that WCPB bricks developed for use as supports in mines
exhibited compression properties similar to those of wood loaded perpendicular to the

grain. The load-deformation curves were initially linear until matrix cracking, after which

they softened (Figure 2.10).

25

600 .

 

 

 

I. Wood : Cement
500 *-
..
A 400 I'
:9; I.
300 ,.
'8
o I-
.—I
200 L-
P
100 -
0
0

 

Deformation (mm)
Figure 2.10. Compressive load vs. deformation for WCP bricks (Sorfa, 1984)

2.7 Durability

The major advantage of WCPB over resin bonded wood particleboard is that
WCPB can be used for either indoor or outdoor applications, while resin bonded wood
particleboard can only be used as interior material because it is sensitive to moisture and

susceptible to damage by ﬁre and insects.

Won (1995) conducted ﬂexural tests on WCPB subjected to 25 cycles of
accelerated wetting-drying. He found that in general, repeated wetting-drying cycles led
to increased stifﬁress and reduced toughness of WCPB, and the effects on ﬂexural

strength were mixed. Figure 2.1 1 shows the wetting-drying test results.

26

Won (1995) also reported that repeated freezing-thawing cycles led to increased
stiffness and reduced toughness values, and the effects on ﬂexural strength were mixed.
Figure 2.12 shows the freezing-thawing test results.

Wolfe (1999) reported that freezing-thawing tests on the wood-cement
particleboard support the premise that these composites can be designed to meet the
durability requirements for highway sound barriers.

Sirnatupang (1990) studied the durability of conventional (air curing) and rapidly
curing (using C02) cement-bonded wood particleboard. He concluded that
conventionally made cement-bonded wood particleboards have higher bending strength
and are also more durable. However, the properties of the rapidly setting boards fulﬁll
the requirements of existing standards.

Dinwoodies (1991) and Karam (1992) demonstrated that wood-cement

particleboard could be fabricated to resist cyclic moisture and temperature effects.

27

 

 

 

cor-ASP!!! (W0)
E 8 ” cox-Aspen
é (XL-SP. (VD)
a 6 oo,-s. P.
8
5 4 ’ cmmp)
'31 Control
a ‘
i5 2
LL.
0 I I l J
0 2 4 6 8 10

Deﬂection (mm)

Figure 2.11 Flexural strength vs. deﬂection under wetting and drying for

wood/cement ratio of 0.28 (Won, 1995)

 

Flexural strength (MPa)

 

Deﬂection (mm)

Figure 2.12 Flexural strength vs. deﬂection under freezing and thawing

for wood/cement = 0.28 (Won, 1995)

28

2.8 Insulation Properties

Wood cement composites provide very good insulation, both thermally and
acoustically. The advantageous insulation properties make these composites attractive for

building applications.

2.8.1 Thermal properties

The thermal property of composite materials is generally related to the porosity of
the composites. Composites with higher porosity usually have higher thermal resistance.
Because of its relatively low density, wood-cement particleboard has good thermal
properties. As reported by Eltomation Corporation (2003), the thermal resistance R-value
for 2.5 cm thick wood-wool cement boards is 5. The R-value is the resistance of one
square meter of the material subject to a one degree temperature difference. The equation
of thermal resistance R-value can be expressed as:

_ AAT
W

R (2.1)

where R is the thermal resistance value in units of mzk/W , A is the surface area Of the
heat ﬂows in units of square meter, A T is temperature difference between the warm and
cold sides of the material, and W is the resultant heat ﬂow in units of Watt.

Southern pine excelsior cement board with a thickness of 5 cm has an R-value of
4.0 (Concrete Products, 1984). Lee (1985) reported that wood excelsior cement board
with a thickness of 5 cm has an R-value (thermal resistance) of 3.5. Lee (1985)
concluded that the thermal conductivity or thermal resistant value is related to the density
of the board and the density of the wood. For a given density of the board, lower density

wood aggregate yields aboard with higher thermal resistance.

29

2.8.2 Acoustic properties

Wood-cement composites are considered to have good sound absorption ability.
Wolfe (1997) reported that natural ﬁber cement composites have good sound absorption
properties, and can be used to reduce noise. Wolfe (1999) reported that the Wisconsin
DOT has used WCPB for highway sound barriers.

Neithalath (2003) reported an experimental test on the acoustic performance of
wood ﬁber cement composites. The test results indicate that wood ﬁber cement
composites have a strong potential for absorbing sound. He further concluded that the
sound absorption coefficient of the wood ﬁber-cement composites increases with an
increase of wood ﬁber volume due to the generation of an increased number of porous
channels by wood ﬁbers in the cement matrix.

Eltomation, the world’s largest wood-cement composites supplier, reported the
following sound absorption data for wood-wool cement boards (Eltomation, 2003):

0 Resistance of a plastered wall of 5 cm wood wool cement boards is 36-40 decibel.

0 Resistance of a plastered double wall with air space between WWCBs is 53-57
decibel.

0 Noise reduction coefﬁcient for a 2.5 cm thick board is 0.7.
In comparison, the acoustic resistance of a permanent shuttered wall with 12 cm

concrete core is 54-56 decibel.

2.8.3 Fire resistance

According to Eltomation (2003) a 5 cm thick wood-wool cement board has a 1-

hour ﬁre resistance and a 10 cm thick board has a 2-hour ﬁre resistance. Kurﬁel and

30

Moslemi (1995) reported that wood cement particleboard is in a class A ﬁre rating. To

have a class A ﬁre rating, a material should be able to withstand severe exposure to ﬁre.

2.9 Theoretical Studies

Information concerning theoretical treatment of the behavior of cement-bonded
wood composites is very limited. Linear elastic fracture mechanics, the discipline
concerned with failure by crack initiation and propagation, is often used to evaluate the
failure mechanism in ﬁber reinforced concrete material. Fiber bridging is generally
accepted as the enhancing mechanism in wood-cement composites. Figure 2.13 shows a
schematic representation of a cross section through a ﬁber-reinforced matrix. The
diagram shows several possible failure modes occurring before fracture of the
composites. In the high stress region near the crack tip, ﬁbers may debond from the
matrix (ﬁber 1). Also, sufﬁcient stress may be transferred to a ﬁber (ﬁber 2) to
ultimately fracture the ﬁber (ﬁber 4). When total debonding occurs, the debonded ﬁber
can then be pulled out from the matrix and considerable energy is absorbed by the system
in the form of ﬁictional energy (ﬁber 3). It is also possible for a ﬁber to be left intact as

the crack propagates. This process is called bridging action.

31

 

 

 

 

Figure 2.13 Schematic representation of cracking through a ﬁber reinforced matrix

(Coutts, 1986)

Different theoretical approaches have been used to predict the strength of WCPB.
Dinwoodie and Paxton (1988), Sorfa (1984) and Wolfe (1999) qualitatively related
strength to density. Sarja (1988) presented the following empirical formulae for the
modulus of elasticity, modulus of rupture and compressive strength of cement-bonded
particleboard as a function of density:

f, = 4.5(p/1000)
fa: p/500 (2.2)
EC: 1500(p /1000)‘

where p = density at 40% humidity in kg/m3
f, = compressive strength in MPa
fc, = tensile strength in MPa

E = modulus of elasticity in MPa

C

The composite mechanics approach has been used by several researchers (Laws,
1983; Balaguru, 1992; Bodig, 1982) in the form of ‘rule of mixtures’. The rule of

mixture stated in its most general form is:

32

o". = 0'me + 0'fo (2.3)

where a; is the strength of the composite, 0'", is the strength of the matrix, a; is the
strength of the ﬁbers, V". is the volume fraction of the matrix, the Vf is the volume
fraction of the ﬁbers.

Similarly, for the modulus of elasticity in tension, EC, the rule of mixture is
written as:

EC = Em (1 - VOW," + 77E, V, (2.4)

Where EC, Em and Ef are the modulus of elasticity of composite, matrix and ﬁber,
respectively.

The rule of mixtures assumes that there is no effect from Poisson’s ratio and that
ﬁbers are fully bonded to the matrix before the composite cracks, i.e. equal strain in ﬁber

and matrix.

2.10 Applications of Wood-Cement Composites

The development and use of wood-cement particle composites attest to their
suitability as construction materials. Wood-cement composites have been successfully
used as rooﬁng, cladding, interior wall, exterior wall, highway sound barriers, etc.

The photos in Figures 2.14 to 2.25 show applications of wood-cement composites

(including particleboard, strand board and wood-wool board) in building construction.

33

 

Figure 2.14 Siding with WCPB in Japan

(Eltomation Company)

 

Figure 2.15 Siding with WWCB in the Phillipines

(Milligan, 2004)

34

 

Figure 2.16a WCP Block for building walls (Hoene, 1997)

 

Figure 2.16b Exterior and interior walls with WCP blocks (Hoene, 1997)

35

 

Figure 2.17 Sound barrier of wood-cement particleboard in Czech Republic

(CETRIS, 2005)

 

Figure 2.18 Wood-wool cement board for rooﬁng and walls in Phillippines

(Eltomation Company)

36

 

Figure 2.19 Wood strand cement board for all walls and sidings in Japan

(Eltomation Company)

 

Figure 2.20 Roof decking with wood excelsior cement composites (Moslemi, 1974)

37

 

 

Figure 2.21 Wood ﬁber cement board as exterior-wall cladding of the Sandoz
Laboratory in Basel, Switzerland, (Deppe, 1974).

2.11 Highway Barriers

2.11.1 Introduction

At a number of locations on existing highways there are physical conditions that
cannot be readily modiﬁed to eliminate or minimize the hazard they present to out-of-
control vehicles. These include brigde piers, abutrnents, and parapet noses; heavy,
nonbreakway sign supports; and high embankment slopes. At these locations, barriers
are used to reduce the hazard.

The term “barriers” refers to longitudinal barriers and crash cushions.
Longitudinal barriers function primarily by containing and redirecting errant vehicles.
Crash cushions function primarily by decelerating errant vehicles to a stop or by

redirection.

38

Longitudinal barriers, including bridge railings, are generally denoted as one of
three types: (a) ﬂexible, (b) serni-rigid and (c) rigid. Figures 2.22 to 2.24 show three

types of longitudinal barriers.

 

Figure 2.23 Semi-rigid guardrail

39

 

Figure 2.24 Rigid concrete barrier

Flexible barriers generally consist of cable barriers, which are primarily designed
to contain the crashing vehicle by deforming a large amount rather than redirecting a
vehicle. These are generally used in local roads. Semi-rigid barriers are usually
guardrails, which are required to contain and redirect vehicles in major highways. The
guardrail is allowed to have a certain amount of deﬂection during vehicle impact. Rigid
barriers are concrete barriers required to contain and redirect vehicles in major highways.

There is no signiﬁcant deﬂection allowed for concrete barriers during vehicle impact.

2.11.2 Barrier safety data

Signiﬁcant improvement in highway safety has been achieved through a multitude
of actions over the past three decades, but one area where serious problems still exist is
the highway roadside. Crash data indicate that highway fatalities nationwide have
increased by 7% since 1992 (FHWA, 1998) and more than 40% of highway fatalities

involve vehicles hitting objects on the roadside, including concrete barriers. Data from

40

the US. Department of Transportation’s Fatality Analysis Reporting System, and the
Insurance Institute for Highway Safety (Roadside hazards, 2001) provide the following
facts:
0 12,311 people died in roadside hazard crashes in 2000, 2% more than in 1999 and
about 12% more than in 1975.
o 96% of roadside hazard crash deaths in 2000 occurred in single-vehicle crashes.
0 The percent of single vehicle roadside hazard crash deaths involving rollover,

in 2000 (Roadside hazards, 2001) is tabulated in Table 2.6.

Table 2.6 Rate of crash death on vehicle-barrier crash

 

 

 

 

Barrier Type Rate of Crash Death
Bridge Rail 41%
Guardrail 50%
Concrete /other barrier 41%

 

 

 

 

The accident data shows that a large number of deaths occur as a result of
vehicles impacting concrete barriers. Therefore, enhancing the impact characteristics of

concrete barriers can be beniﬁcial to the motoring public.

2.11.3 Crash performance of concrete barriers

Concrete Median Barriers are widely used structures on the roadways of North
America. The design of concrete median barriers has been evolving since 1950’s
(Marine, M. C., 1999). As a crash safety device, concrete median barriers play an
important role to contain and redirect crashing vehicle. Numerous crash test research

programs on concrete median barriers have been conducted from late 1960’s through

41

early 1990’s (Marine, M. C., 1999). The concrete barrier’s shape plays an important role
in vehicle-barrier crashes. In general, there are three major types of concrete barriers
with different shapes (Figure 2.25): 1) New Jersey (NJ) shape concrete barriers, 2)

constant slope concrete barriers, and 3) vertical concrete parapets.

 

 

 

 

 

a. NJ Shape Barrier b. Constant slope barrier c. Vertical concrete parapet

Figure 2.25 Major shapes of concrete barriers

For NJ shape concrete barriers, when a vehicle’s front bumper impacts the lower
sloped face and slides upwards, the vehicle is lifted and part of the kinetic energy is
converted to potential energy. This potential energy is converted back into kinetic energy
as the vehicle returns to the ground.

Vertical concrete parapet walls do not have the energy management feature of the
NJ barrier, but crash tests have demonstrated that they can perform adequately
(McDevitt, 2002). All of the energy absorption in an impact with a rigid vertical wall is
due to crushing of the vehicle. Bumpers usually do not slide up vertical concrete walls
and lift the vehicle and all four wheels tend to stay on the ground. This minimizes the

potential for vehicle rollover (McDevitt, 2002).

42

On March 22, 1990, the Texas Transportation Institute conducted a ﬁeld crash test
of a vehicle crashing onto a portable concrete barrier. The vehicle was a Honda Civic
(820C), and the barrier was a longitudinal concrete temporary barrier. Figures 2.26 to
2.27 show the barrier’s geometry, plan view of the ﬁeld test demonstration. This ﬁeld test
indicated that the New Jersey shape barrier successfully contained and redirected the

crashing vehicle.

 

 

Figure 2.26 Geometry of concrete barrier (NCHRP350 Report, 1993)

43

 

 

 

 

 

Figure 2.27 Demonstration of ﬁeld test -plan view (NCHRP350 Report, 1993)

2.12 Concrete Material Model

The response of a reinforced concrete structure, such as concrete crash barrier, is
partly determined by the material response of the plain concrete of which it is composed.
Therefore, analysis of structural behavior to static or dynamic loading requires prediction
of the concrete response to load histories. The fundamental characteristics of concrete
behavior are based on the results of experimental testing of plain concrete specimens
under speciﬁc, relatively simple loading histories. Elasticity and plasticity theories
provide a framework for establishing a material model that describes these fundamental
characteristics. Experimental data provide additional information for validating and

calibrating the material model.

44

2.12.1 Biaxial and Triaxial Behavior

There is a large amount of research that has been done on the behavior of concrete
under biaxial loading. The biaxial strength envelope of concrete obtained by Kupfer et a1.
is quite famous. Kupfer et a1. (1969) conducted a series of tests to investigate the
response of plain concrete subjected to two-dimensional loading. In the investigation,
concrete plates (approximately 200 mm by 200 mm by 50 mm in dimension) were loaded
to failure at prescribed ratios of 61:62 with 63 equal to zero (where 0'; is the ith principal
stress). Two directional loads were applied using steel brushes to minimize stresses
induced through ﬁiction at the specimen boundaries. The failure surfaces obtained
through the testing are presented in Figure 2.28.

Tasuji et a1. (1978) also conducted a series of biaxial tests on concrete. They
obtained a strength envelope similar to that reported Kupfer (1969). In addition to the
failure surface, they also obtained the concrete’s elastic limit surface. Both surfaces are
shown in Figure 2.29. The elastic limit surface has a shape similar to but smaller than the

failure surface.

45

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

0.2
0 '1
-0.2 '
-0.4 I 1
as I ff (PSi) l
--- 2700 g
‘ ' — 4450 ‘ i I
'0.8 \ —-— 8350
-l.0
I
\
\
-1.2 q \‘
s~ _'
-l.4

 

 

 

 

 

 

Figure 2.28 Biaxial strength envelope of concrete (Kupfer et al., 1969)

As shown in Figure 2.29, the maximum strength increases under a biaxial

compression state. A maximum strength increase of approximately 25% is obtained at a

stress ratio of 12. = 0,5 and of 16% at an equal biaxial compression stress state
01

(£2. =10 ). Under biaxial compression-tension, the compressive strength decreases
0'

almost linearly as the applied tensile stress is increased. Under biaxial tension, the

strength is nearly the same as that under uniaxial tension.

46

 

 

 

 

'4
92 O I». $
f. ’
0.2 ’
0.4 "

 

f

Failure strength

1.2-

 

 

0.6

0.8

1.0”
l

 

 

-
_ ~< ‘-
in
l l l I l
2 1.0 0.8 0.6 0.4 0.2

1. 0.0 -0.2

0'2
f.-

Figure 2.29 Elastic limit and maximum strength envelopes of concrete under biaxial

loading (Tasuji, 1976).

Green and Swanson (1973) conducted hydrostatic compression tests on concrete.
The tests indicated that concrete exhibits nonlinear behavior under hydrostatic
compression, as shown in Figure 2.30. The stress-strain curve of Figure 2.30 shows a
reversal in curvature on loading. Upon unloading, the slope of the curvature is almost

constant and approximately equal to the slope of the initial tangent for the loading curve.

47

 

Compressive strength:
7000 @si)

Mean normal stress. Goct (ksi)

 

 

 

 

o -2 -4 -s -3

Volume change, 8v (%)

Figure 2.30 Behavior of concrete under hydrostatic compression

(Green and Swanson, 1973).

Under triaxial loading, experiments indicate that concrete has a fairly consistent
failure surface that is a function of the three principal stresses (Gerstle et al., 1978). If it
is assumed that concrete is isotropic, the elastic limit (onset of stable crack propagation),
the onset of unstable crack propagation, and the failure limit all can be represented as
surfaces in three-dimensional principal stress space. The shape of the elastic limit surface

and failure surface are shown schematically in Figure 2.31.

48

0'1

Failure surface

Elastic limit surface

 

01:62:63 02

Figure 2.31 Schematic representation of the elastic limit and failure surface of concrete

2.12.2 Characteristics of yield surface

The general characteristics of the yield surface of concrete have been determined
through experiments. The experimental results are used to develop mathematical failure
models.

The general shape of the failure surface in the principal stress space can be best
described by its cross-sectional shapes in the deviatoric planes and its meridians in the
meridian planes. The cross sections of the failure surface are the intersection curves
between the failure surface and a deviatoric plane which is perpendicular to the
hydrostatic axis with {f = constant. The meridians of the failure surface are the
intersection curves between the failure surface and a plane (the meridian plane)
containing the hydrostatic axis with 0 = constant. Figure 2.32 shows a geometric
representation of the stress state in the principal stress space. The straight line 0N which

has equal angles with the three coordinate axes is the hydrostatic axis. Deviatoric planes

49

are the planes that are perpendicular to the hydrostatic axis. Meridian planes are the

planes that contain the hydrostatic axis.

0'3
A
P(O' 1,0253)
0 Hydrostatic axis
N 0' 1 = 62 = 0'3
5..
e
0 ’ 0'2

 

 

61

Figure 2.32 Geometric representation of the stress state in the principal stress space.

2.12.3 Failure surface of concrete under different stress states

The actual behavior and strength of concrete is complicated. This behavior
depends on the geometric and material properties of the aggregate and cement paste, as
well as the concrete curing condition and the nature of loading. Even though no one
mathematical model can be expected to completely describe the strength properties of
real concrete material, simpler models or criteria must be used in representing only those
characteristics that are essential to the problem at hand.

Under tensile stresses concrete fails by a cleavage type of brittle fracture. Under
high hydrostatic pressure, concrete can yield and ﬂow as a ductile material on the failure

or yield surface (Chen and Saleeb, 1982).

50

Under tensile loading, concrete failure can be described by Rankine’s the

maximum tensile stress criterion:
alzft’ 02=ft’ 03:fr (25)
where 0',- are principal stresses, and f ' is the tensile strength.

1

Under high hydrostatic pressure, shearing stress criteria, such as the Tresca and
Von Mises criteria, are used to predict the ductile ﬁacture.

Tresca
l l l
maxli'Z—lo'l—(yZl’Eial_02|93|01—02i]:k (2'6)

Von Mises

f(J.)=J2-k’=0 (2.7)
where k is the failure (yield) stress in pure shear. For concrete k is determined from a
simple compression test (with f; used as the yield stress). In this case, k is equal to

. 1 ,
% fc or —‘/—3=fc for the Tresca and Von Mises criteria, respectively. J2 is the second

invariant of the stress deviator tensor, and is

.12 = %[(0'1 «(72)2 + ((72 —0'3)2 +(o'3 —a,)"] , p = 2J2 (2.8)

In the intermediate compressive stress range, failure criteria of concrete are
sensitive to the hydrostatic state of stress. The simple one-parameter models described
above are not adequate for describing failure in the fracture-ductile state under this
intermediate level of compressive stresses, and therefore pressure-dependent failure

models must be used. The simplest and most commonly used models of this type are the

51

Mohr-Coulomb and Drucker-Prager failure criteria. These are two-parameter models

whose meridians depend linearly on the hydrostatic stress component, I 1 or 15

(a)
9 0'1
Compression ‘I It

Tension

O\

o

O

/
’ a:

 

 

O3

 

 

 

0'3

0'2
Figure 2.33 Meridian and deviatoric sections for two parameter failure models:

(a) Mohr-Coulomb criterion, and (b) Drucker-Prager criterion.

In the Mohr-Coulomb criteria, failure is assumed to occur when the shear stress,
1:, on any plane at a point in a concrete material reaches a value that depends linearly

upon the normal stress, 0', in the same plane. Mathematically, this is expressed as

|t|=c—0'tan¢ (2.9)

52

where c and (P are material constants that represent the cohesion and angle of internal
friction, respectively.

The Mohr-coulomb failure surface has sharp corners (Figure 2.33). These corners
are not mathematically convenient because the slope discontinuity at these comers can
cause many difﬁculties and complications in numerical solution schemes.

Drucker and Prager (1952) proposed a smooth approximation to the Mohr-

Coulomb surface by the following simple modiﬁcation of the von Mises criterion:

f(1,,J,)=a1,+JZ-k=0 (2.10)
orin term ofﬁandpas

“awn/3a: +p-J5k=0 (2.11)
where k and a are (positive) material constants. I, is the ﬁrst principal stress invariant
given by,

11 20'1“}2 +03 (2.12)

There are two shortcomings of the Drucker-Prager criterion in modeling concrete

material: the linear relation between I1 and E (or 5 and p), and the independence
with respect to the angle of similarity 6. Experiments on concrete show that relation
between I1 andJZ (or 5 and p) is nonlinear or curved, and the trace of the failure

surface on deviatoric sections is not circular, but depends on the angle of similarity 0.
Bresler and Pister (1958) suggested a three-pararneter failure criterion with a
parabolic relation between Tao: and 032., and with circular deviatoric sections. The relation

between 10c, and 03c, can be expressed in the form of

53

T 0' 0'
——°“. = -b(——°€‘)+ I—“f ’ 2-13
f. a f. c f. ) ( ’

Hsieh et a1 (1982) developed a four-parameter criterion in terms of 11, J2 and

maximum principal stress 07:

 

,/J
f(1.,12,01)=a J2, +b——.2—+9—‘.—+d—I—‘.——1=0 (2.14)
f f. f. f.

t
C

where a, b, c and d are material constants. The nonlinear meridian trace and deviatoric

sections of this failure model are shown in Figure 2.34.

DA A

Compressive meridian

V

\ \ \

Tensile meridian \
\

 

 

 

Figure 2.34 Hsieh-Ting-Chen criterion

A large number of concrete models have been published in the literature. Most of

the modiﬁed models give a better agreement with relevant experimental data. However,

the more accurate models generally involve many parameters that are difﬁcult to obtain

54

experimentally. In practice, the Mohr-Coulomb and Drucker-Prager failure criteria are

the most commonly used models.

2.12.4 Evolution of the yield surface of concrete under multi-dimensional loading
There are few investigations that have considered the evolution of the yield
surface of concrete from the initial inelastic response through complete loss of strength.
The concept of the evolution of the yield surface of concrete is cited from the theory of
plasticity. In plasticity theory, a change of yield surface is always described by a
hardening rule. There are two common hardening rules: ‘isotropic hardening’ in which a
change occurs in the size of the yield surface only (all dimensions being scaled by the
same constant); and ‘kinematic hardening’ in which a change occurs in the location of the

yield surface. ‘Isotropic hardening’ is the more appropriate hardening rule for concrete.

Chen and Han [1988] proposed a more detailed evolution of the yield surface of
concrete under a variable load history (Figure 2.35). This model proposes that at the
maximum load, the yield surface may appropriately be deﬁned by the recommendations

of any of several researchers [0ttosen, 1977; Hsieh et al., 1982; Willarn and Wamke,

1975].

55

Uniaxial compressive loading path

Failure surface

    
  

Loading surface

 

\ -11
- ‘ +

 

Figure 2.35 Evolution of concrete yield surface as proposed by Chen and Han [1988].

2.13 Finite Element Modeling for Vehicle Crash Analysis

In the last four decades, the ﬁnite element method (FEM) has become the
dominant technique for analyzing physical phenomena in the ﬁeld of structural, solid, and
ﬂuid mechanics.

The nonlinear, explicit, three-dimensional ﬁnite element analysis code DYNA3D
developed in the 19705 has made it possible to analyze contact impact problems. The ﬁrst
full vehicle car crash model was built and analyzed in the mid-19808 (Du Dois, 1999).
Nowadays, crash simulation has become a routine task in vehicle design. However,
making a high quality ﬁnite element (FE) model that can match actual vehicle behavior is
still a formidable challenge for engineers. A high quality FE model should not only have
good geometry with high quality elements, but also have good material modeling, contact

modeling and control commands.

56

2.13.1 Brief summary of formulations

LS-DYNA (Hallquist, 1999) is a nonlinear, dynamic, three-dirnensional ﬁnite
element analysis code that evolved from DYNA3D. It is being widely used from the
automotive to the biomechanics ﬁelds.

LS-DYNA uses a Lagrangian formulation in which the mesh moves with the
material. Formulation in ﬂuid dynamics is, in contrast, Eulerian, and the FE mesh
remains ﬁxed in space. In LS-DYNA, conditions of equilibrium are derived by the

principle of virtual work and are expressed by the following equation:

M W 562', = F." (2.15)

where
My“ is the diagonal (lumped) mass matrix;
56:, is the acceleration at node k with respect to degree of ﬁ'eedom I;

F i,- is the sum of the internal and external forces; and the superscript n refers to the n-th
integration time step.

The diagonal mass matrix enables equation (2.14) to be solved easily for the
accelerations. The velocities and displacements are then obtained through the explicit
central difference method:

1
"——

l
o -2- - O 2 n can
x0. — xi]. + At xi}.

- n+1 - n n+:li "r”;
xi}. = xi]. ‘1' At xij (2.16)

l
n+—

At 2 = %(At"+' + At")

57

where At is the integration time step.

2.13.2 Time integration

The central difference method is an explicit integration method that is
conditionally stable. If At is too large, the method becomes unstable. If At is much
smaller than necessary, computations become expensive. Therefore, it is necessary to
choose the time step At so that the method is stable and effective. The stability condition
is given by the Courant, Friedrichs, and Lewy condition.

At 3 5 (2.17)
C

where c = E / p is the speed at which information travels in the element, E is the

modulus of elasticity of the material, and p is the density of the material. L is the
characteristic length of elements.
The physical interpretation of this condition is that At must be small enough so

that information does not propagate across more than one element per time step.

2.13.3 Contact algorithm

The general contact approach taken in LS-DYNA is referred to as the “penalty
method.” When a node penetrates a surface, it is opposed by a linear spring force that
acts normal to the surface. This force is calculated using plate theory based on the depth
of the penetration and the stiffness of the surface element. If friction is speciﬁed, then the
friction force is calculated from the normal force. A node trying to penetrate the surface

is usually referred to as the “slave” node, and the surface it is trying to penetrate is the

58

“master segment.” These terms appeared in the ﬁnite difference literature when most

contact algorithms were not symmetric.

2.13.4 Force and displacement field quantities
The LS-DYNA analysis mainly outputs force ﬁeld quantities and displacement
ﬁeld quantities. The force ﬁeld quantities mainly include stresses and strains, and the
displacement ﬁeld quantities mainly include displacements, velocities and accelerations.
In LS-DYNA analysis, force ﬁeld quantities vary over the element. To reduce the
CPU time, it is assumed that the stress and strain are constant over the element and equal
to the values calculated at the center of the element (or at the integration point). However,

the displacement ﬁeld quantities (acceleration, velocity and displacement) are evaluated

at the nodal points.

2.13.5 Simulation of barrier crash tests using LS-DYNA

Designing safer roadside crash barriers in highway systems is a primary goal of
the state DOTS and the Federal Highway Administration (FHWA). The crashworthiness
of some roadside barriers has been evaluated through full-scale ﬁeld-testing. Crash
testing has been carried out for many years but at considerable expense. If accurate
computational models can be developed and then used to predict the effects of vehicle-
barrier collisions, enhanced designs of such hardware systems can be achieved more
efﬁciently.

Finite element analysis based upon the failure mechanics of traditional materials

provides the best computational tool with which to undertake research that will lead to a

59

better understanding of the behavior of roadside safety structures. This will eventually
lead to the formulation of improved designs, possibly utilizing new high-perforrnance
materials in the fabrication of roadside safety structures.

Previous simulation studies of vehicles impacting a guardrail (Plaxico, 1995) and
a concrete barrier (Nilsson, 1998) show that the ﬁnite element simulations of vehicle-
barrier impacts provide good agreement with the results of full-scale ﬁeld tests. Figures
2.36 and Figure 2.37 Show a crash simulation using LS-DYNA computer program

conducted at Texas A&M University in 1999 and a crash ﬁeld test conducted by the

Washington State DOT in 1998 on a new portable concrete barriers.

 

Figure 2.36 Crash ﬁeld test (front view) Figure 2.37 Crash Simulation (front view)

(Roger, 2002) (Roger, 2002)

Comparing Figure 2.36 with Figure 2.37, the vehicle’s frontal right corner
deformation and the shape of the deformed concrete barrier are almost identical. The
computer crash simulation with LS-DYNA provides an excellent prediction of the

vehicles motion during impact.

60

Figures 2.38 and 2.39 show the crash simulation and crash ﬁeld test on a
longitudinal concrete barrier conducted by the National Crash Analysis Center in 1995
(Zaouk, 1995). Again, the simulation predicts the motion of the vehicle quite well. The
simulation also provide insights that are not available from ﬁeld tests about the design

parameters such as detailed material responses.

 

t =0 ms
t =75 ms
t =120 ms
t =240 ms
Figure 2.38 Crash ﬁeld test Figure 2.39 Crash simulation
(plan view) (plan view)

61

CHAPTER 3
IMPROVED MANUFACTURING PROCESS AND FLEXURAL

PERFORMANCE

3.1 Introduction

The manufacturing of cement-bonded particleboard is similar to that of resin-
bonded particleboard. Making wood-cement particleboard stronger and tougher is the
primary goal of the wood-cement composite industry. Better manufacturing processes
have been developed to create high quality wood-cement particleboards. Up to now, two
major modiﬁcations of the conventional manufacturing process for wood-cement
particleboard are generally employed. One of these is the accelerated process with CO2
injection that greatly shortens the manufacturing time. Another is the addition of
chemical additives to improve the bonding strength between wood particles and the
cement matrix.

Sirnatupang (1990) studied the performances of conventional (air curing) and
rapidly cured (using C02) cement-bonded wood particleboard. He concluded that
conventionally cured cement-bonded wood particleboards have higher bending strength
and are also more durable. However, the properties of the rapidly cured boards fulﬁll the
requirements of existing standards (Hsu, 1995).

Chemical additives in wood cement composites does improve the bending
strength of the composites, but the improvement is limited. And as with the C02

injection process, the chemical additive increases the manufacturing cost.

62

In the research reported herein, a pressure controlled and fully saturated curing
process was investigated. This manufacturing process yields Signiﬁcant advantages over
currently used manufacturing procedures. A total of 70 ﬂexural specimens with different
cement/wood ratios were fabricated using the fully saturated curing process and tested. In
addition to the effect of the curing process the effect of the cement/wood ratio on the

strength of wood-cement particleboard also was investigated.

3.2 Pressure Controlled and Fully Saturated Curing Process

In the pressure controlled and fully saturated curing process, the particleboard is

pressed with a constant pressure and is fully saturated in water for a 28-day curing period.

3.2.1 Advantages of pressure controlled and fully saturated curing process

In the currently used accelerated and air curing methods, wood particles shrink
during the curing period. The shrinkage induces some degree of debonding between the
wood particles and the cement matrix before the cement is fully hardened. Further, the
debonding action weakens the bond strength between wood particles and the cement
matrix resulting in the particleboard having a lower material strength.

The fully saturated curing procedure improves bond strength between wood
particles and the cement matrix. The bond strength between wood particles and cement
is related to the interlock between the irregular surfaces of wood particles and the cement
matrix. Because wood particles do not shrink in water during the 28-day curing period

due to the fully saturated condition, the interface between wood particles and the cement

63

matrix provides considerable bond strength. Pressure controlled manufacturing with fully
saturated curing should therefore produce superior wood-cement particleboard.

In addition, in the pressure controlled manufacturing process specimens with
different cement/wood ratios experience the same pressure during the setting period and
the cement and wood particles in the specimens develop the same internal prestresses.
Therefore, specimens made by this process yield stronger correlations between material

properties and cement/wood ratios.

3.2.2 Materials

The materials used in this research are soft wood (southern yellow pine) particles,
Portland cement (Type I) and water. The particles were made from waste wood treated
with chromated copper arsenate (CCA).

In this study, CCA-treated wood (southern yellow pine) removed from service is
used to produce wood particles. Figure 3.1 shows a typical cross section of CCA-treated
southern yellow pine removed from service. The wood particles were produced through
two steps. First, crushed wood is placed into a cutting machine (Figure 3.3) to produce
raw wood particles. Second, the particles are separated by size by sieving with Tyler
Screens (Figure 3.4). Particles (Figure 3.2) with size ﬁom 2 to 8 mm long x 1.0 mm in
diameter are selected for making the particleboard.

All the wood particles used in this research were made at the Wood Technology

Laboratory in the Department of Forestry at Michigan State University.

64

      

A ‘.~
.‘n

“Figure 3.1. CCA-treated wood Figure 3.2. Wood particles I

   

Figure 3.3. Wood particle outing machine Figure 3.4. Wood particle screener

65

3.2.3 Particleboard manufacturing

All wood-cement composite Specimens were fabricated in the Concrete Material
Laboratory in the Department of Civil and Environmental Engineering at Michigan State
University. In making test specimens, the quantity of wood particles, water and cement
required to produce particleboard measuring 315 mm (wide) by 315 mm (long) by 12~20
mm (thickness) were estimated for a given cement/wood ratio.

The required quantity of wood particles, cement and water were mechanically
mixed until homogenous slurry was formed. The composite slurry was transferred onto a
metal mat of315 mm x 315 mm x 14 mm, and loaded onto a hydraulic press for pressing.
The mat remained under pressure for 24 hours. After 24 hours of cold pressing, the
particleboard was removed and put into a water container for 28 days of curing under the
laboratory room environment.

After 28 days of curing, the particleboard was removed from the water and left to
dry in the laboratory environment for two days. A table saw was then used to out each
particleboard into the test specimens with dimensions of 230 mm (long) by 100 mm
(wide) by 10 to 22 mm (thick).

Figures 3.5 to 3.8 show the mixer, hydraulic press, curing container and saw used
to cut the board, respectively. Figure 3.9 shows the ﬂowchart of the manufacturing

process for the wood-cement particleboard.

66

 

Figure 3.5. Cement wood particle mixer Figure 3.6. Hydraulic press

 

Figure 3.7. Curing in water Figure 3.8. Saw used to cut board

67

Flowchart of Wood-Cement Particleboard Manufacturing Process

 

 

CCA-treated
Wood

 

 

I

 

 

 

Cutting/Screenin

Cement

 

 

 

 

 

I

I

 

 

 

Wood Particles

ﬂu Dry mixing

 

 

 

~————> Mixing

 

 

 

 

Cutting/Sizing

 

 

 

 

Water

I

 

 

 

 

 

I

Forming

I

Pressing

I

 

 

 

 

 

 

 

 

 

 

Setting/Hardening
Under Press

 

I

Demolding

I

Curing

I

 

 

 

 

 

 

 

 

 

 

 

 

I

Storage

 

 

 

 

 

Drying/Conditioning

 

Figure 3.9 Flowchart of manufacturing process

68

 

 

3.3 Flexural Properties of Wood-Cement Particleboard

Wood-cement particleboard is presently used mainly as non-structural members
in building construction. They are often subjected bending load (e. g. particleboard siding
subjected to wind load). The bending strength of the wood-cement particleboard is
generally used to evaluate the quality of the boards in the wood-cement composites

industry and is an important property.

3.3.1 Experimental Design

Wood—cement particleboard using CCA-treated softwood particles, Portland
cement (Type I), and water, were fabricated as described in Section 3.1.3. The test matrix
for the bending test is given in Table 3.1. All specimens had a length of 230 mm, a width

of 100 mm and a thickness that varied ﬁom 12 to 20 m.

Table 3.1 Test Matrix for Bending Test

Cement/wood
ratio 4.0 3.5 3.0 2.0 1.5 1.25 1.0
Number of
specimens 10 10 10 10 10 10 10

 

 

 

 

 

 

 

 

 

 

 

 

3.3.2 Description of Test

Bending tests were conducted to determine material properties of the wood-
cernent particleboard. The American National Standards Institute Speciﬁcations for
particleboard (ANSI A208.1-1993) and The American Society for Testing and Materials
(standard method of evaluating the properties of wood based ﬁber and particle panel

materials, ASTM D 1037-78, 1996) were used.

69

Three point bending tests were conducted on an INSTRON machine (Figure 3.10)
to determine the modulus of rupture (MOR), modulus of elasticity (MOE) and the strain
at peak load. The tests were conducted in the Department of Forestry at Michigan State
University. Figure 3.11 shows the test set-up. The sizes of the test specimens is shown

in Figure 3.12.

   

 

 

 

 

 

 

 

 

Figure 3.10. INSTRON Machine Figure 3.11. Bending test set-up
I—
8 g. (Thickness = 10 ~ 22 mm) a
"‘ o.
:2 :2
__\L_
I] 200 mm
|\ ’I
L 230 mm

 

Figure 3.12 Specimen’s size

70

The loading speed was based on ASTM D 1037-78 (1996). The loading rate of
the moveable crosshead of the test machine calculated as follows:
N = zL2/6d (3.1)
where:
N = rate of motion of moving head, in./min (mm/min),
z = unit rate of ﬁber strain, in./in.(mm/mm) of outer ﬁber length per minute (0.005),
L = span, in. (mm) and

d = depth (thickness) of specimen, in. (mm).

The Specimens to be tested in the soaked condition were submerged in water at a
temperature of 20° 3: 3° C for 24 hours before the test and were tested immediately

upon removal from the water.

3.3.3 Test results and analysis

The typical failure characteristics of the specimens under bending are shown in
Figures 3.13 and 3.14. Observation of the bending test showed that failure does not
generally occurs due to a straight crack but due to a network of independent cracks as
shown in Figure 3.13 that are most likely caused by the wood particle’s bridging action.
The cross sectional view of the particleboard’s failure (Figure 3.14) further indicates that
the wood particles play a bridging action role in resisting bending load.

Figure 3.13 and 3.14 Show that the wood-cement particleboard failure under

bending occurs after the wood particles pull out ﬁ'om the cement matrix. As the wood

71

 

particles were pulled out, the bridging action disappears and the board loses its bending

capacity.

 

Figure 3.14. Cross-section view of failure

The typical load-deﬂection curves (Figure 3.15) show that the WCPB behaves as
an elastic-degrading plastic material with signiﬁcant loss of strength following the

achievement of peak load.

72

 

140 I
120 ‘ Cement/wood = 1.5
100 .
80 i
60 2

(1b.)

Cement/wood = 1.0

Load

40 i /
20 « Cement/wood = 3.0

0 - i .
0 0.02 0.04 0.06 0.08 0.1 0.12 0.14

 

 

 

I I I

Deﬂection (in.)

Figure 3.15. Load-deﬂection curves for varying cement/wood ratios

The test was conducted based on the American National Standard ASTM D 1037-

78, (1996).

For the static bending test, the ASTM D 103 7-78 deﬁnes the modulus of rupture

and the modulus of elasticity as follows:

MOR =3PL/2bd2 (3.2)

MOE = PIL3/4bd3y. (3.3)
where:
b = width of specimen, in. (mm),
d = thickness (depth) of specimen, in. (mm),

L = length of span, in. (mm),

73

P = maximum load, lbf (N),

P1 = load at proportional limit, lbf (N),

MOR = modulus of rupture, psi (kPa),

MOE = stiffness ( aparent modulus of elasticity), psi (kPa),

y] = center deﬂection at proportional limit load, in. (mm).

The measured ﬂextural strength (modulus of rupture), modulus of elasticity and

 

density of the wood-cement particleboard are presented in Table 3.2 and Figure 3.16 to

Figure 3.18.

74

Table 3.2 Flexural Perforrnence of Wood -Cement Particleboard Manufactured with
Seven Different Cement/Wood Ratios

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Modulus of Rupture Modulus of Elasticity Density
Test Results (MPa) (MPa) (kg/m3)
Mean St. Dev. Mean St. Dev. Mean St. Dev.
1.0 10.1 1 0.72 1670 180 967 52
1.25 10.33 0.63 1980 186 978 49
.2
$3 1.5 11.71 0.44 3027 305 1044 31
'8
‘3’
E 2.0 9.53 0.68 3319 402 1151 34
O 3.0 7.03 0.29 3606 414 1329 46
3.5 7.4 0.65 4376 565 1468 67
4.0 7.31 0.71 5026 637 1551 71

 

Figure 3.16 indicates that a cement/wood ratio of 1.5 yields the highest modulus

of rupture and the largest strain at peak load. When the cement/wood ratio is large, there

are insufﬁcient wood particles to provide effective bridging action and hence the MOR is

low. When the cement/wood ratio is too low, there is insufﬁcient binder to fully bond the

ﬁbers with the matrix. A cement/wood ratio of 1.5 appears to be optimal. Figure 3.17

shows that the modulus of elasticity increases almost linearly with increasing

cement/wood ratio. Figure 3.18 shows that the density of the board increases linearly

75

 

with increasing cement/wood ratio. The cement/wood ratios of 1.5 and 1.0 were used for

all other tests.

 

(MPa)
5

 

 

Modulus of rupture

 

O I I ﬁr I I

0 l 2 3 4 5

Cement/wood ratio

 

6000
5000 7
4000 -
3000 -

2000 ‘

 

1000 .

Modulus of elasticity (MPa)

 

 

 

O T T r I
0 1 2 3 4 5
Cement/wood ratio

Figure 3.17. Correlation between MOE and cement/wood ratio

76

 

2000

 

 

 

”a 15004
to
é
g 1000-
E
8

500 4

0 I I I I
0 1 2 3 4 5

Cement/wood ratio

Figure 3.18. Correlation between density and cement/wood ratio

Toughness is generally used as a measure of energy dissipation for a material.
The most common method to measure toughness is to use the load-deﬂection curve of a
simply supported beam under bending. There are at least three procedures that are
suggested to quantify toughness using the load-deﬂection curve obtained in static
loading. As reported by ACI Committee 544 on Fiber Reinforced Concrete, the ideal

way to deﬁne the toughness index I, is by the following ratio:

Area under the load —deﬂection curve until the

load reaches zero for ﬁber composite

 

= 3.4
' Area under the load —deﬂection curve until the ( )

load reaches zero for plain matrix

77

In equation (3.4), the energy absorbed by a ﬁber-reinforced composite beam is
normalized by dividing it by the energy absorbed by an identical beam made with the
plain matrix. While equation (3 .4) is conceptually sound, it poses a number of practical
difficulties (e.g., plain concrete often suddenly collapses).

In practice, the ASTM C1018 procedure is more suitable for the toughness index
1,. In the ASTM C1018 procedure, the denominator in equation (3.4) is taken as the area
under the load-deﬂection curve up to the ﬁrst crack. The ﬁrst crack is assumed to occur at
the point where the load-deﬂection curve deviates from the initial linear portion. The
numerator is taken as the area under the load-deﬂection curve up to a certain speciﬁed
deﬂection. Three levels of deﬂection, namely, 35, 5.58 and 10.55, are suggested for the
numerator, where 5 is the deﬂection at ﬁrst crack. The three suggested indices, called 15,

1,0, and 120 are deﬁned as:

I __ Area under the load - deﬂection curve up to 3d

 

 

 

3.5
5 Area under the load —deﬂection up to d ( )
Area It to 5.55
1l0 = p (3.6)
Area up to 5
Area u to 10.56
120 = p (3.7)

Area up to 5

78

7 Load
Pmax
Maximum load I -------

    

First Crack

  

 

Deformation

b

 

 

OO

36 5.56

Figure 3.19 Toughness calculation curve (load-deformation curve)

The procedure is shown schematically in Figure 3.19. If a plain concrete beam is
tested under load control, the beam would break into two pieces right after the ﬁrst crack
occurs, yielding a toughness index of 1.0.

In this research, the ASTM C1018 procedure was applied to calculate the
toughness of wood-cement particle composites. Table 3.3 shows the calculated
toughness index 15 for wood cement particleboard with seven different cement/wood
ratios under three-point bending tests. The calculated toughness indices (15) are roughly

around 6.0.

79

 

 

 

 

Table 3.3 Flexural Tougmess of Wood Cement Particleboard
Cement/Wood Ratio
Toughness Results
1.0 1.25 1.5 2.0 3.0 3.5 4.0
Toughness Mean 5.95 6.04 6.22 6.18 6.33 5.81 5.72
15 St. Dev. 0.11 0.08 0.16 0.21 0.25 0.22 0.33

 

 

 

 

 

 

 

 

 

 

 

Statistical analysis of the test results conﬁrmed that the fully saturated curing
process yielded improved ﬂexural performance at the 95 % level of conﬁdence. The

conﬁdence interval level is calculated by the formula:

7 + 196(3) (3.8)
n

where, Y is the mean value, 0’ is the standard deviation and n is the number of samples.

The modulus of rupture, modulus of elasticity, density and toughness index of
wood-cement particleboards with seven different cement/wood ratios are shown in Figure
3.20, Figure 3.21, Figure 3.23 and Figure 3.25, respectively. The 95% conﬁdence interval
is shown using error bars.

Figure 3.22 and Figure 3.24 Show the linear regression lines for modulus of
elasticity (MOE) vs. cement/wood ratio and density vs. cement/woOd ratio, respectively.
The linear regression equations has the following form:

(3.9)

y = a + bx, with an estimate of R2

where x is the explanatory variable, y is the dependent variable, and a and b are constant.

80

The a, b and R2 are calculated from the equations:

I In =20, -;)2

= in. —§)(y. 4;)

lyy =Z(yi -;)2
l
< b=ll
In
_ I”.
= ——-'—x
y 1

LI:

 

 

1 1‘.
R2 =—1,1— ’0
n ”( ll )

t xxyy

where n is the number of samples.

Modulus of Rupture, (MPa)

 

 

I III

(3.10)

 

Cement/wood ratio

Figure 3.20

conﬁdence interval

81

Variation of modulus of rupture vs.

10 ll 12 13

cement/wood ratio and 95%

Modulus of Elasticity, (MPa)

 

 

5000 I
4000 I
3000 i
2000
1000 -
(1m 1 l 1 l I
30 3.5 4.0

1.0 1.25 1.5 2.0 2.5
Cement/wood ratio

Figure 3.21 Variation of modulus of elasticity vs. cement/wood ratio and 95%

conﬁdence interval

6000 y = 1015+978x

5000 R2: 102865

T— _' '1

(MPa)

 
   

 

I
.5 4000 'r
.3 3000 L
‘2 l
2 2000 I A A A Testdata
g 1000 I — Regression line
2 0 I I I I

0 1 2 3 4 5
Cement/wood ratio

Figure 3.22 Modulus of elasticity vs. cement/wood ratio

82

Density, (kg/m3)

Figure 3.23

Density, (kg/m3)

1800
1600
1400
1200
1000
800
600
400
200

1800
1600
1400 I
1200

1000
800 b
600 F-
400 _
200 t
0.00 I
25 30 3.5 4.0

.1

 

1.0 1.25 1.5 2.0
Cement/wood ratio

Variation of density vs. cement/wood ratio and 95% conﬁdence

 

interval

I" y = 744 + 201x

E R2 = 218

I

I A Test data

[ — Regression line

I

I I I I I
0 1 2 3 4 5

Cement/wood ratio

Figure 3.24 Density vs. cement/wood ratio

83

Toughness Index, [5

 

 

.0 3 0 3.5 4.0

1.0 1.25 1.5 2

Cement/wood ratio

Figure 3.25 Variation of toughness index vs. cement/wood ratio and 95% conﬁdence

interval

3.4 Discussion

Flexural test results show that the pressure controlled and fully saturated curing
process improves the bending strength of the wood-cement particleboard to 11.7 MPa
which is higher than the commonly accepted limit of 10 MPa (Wolfe, 1996).

A more reliable relationship between bending strength and cement/wood ratio was
found for wood-cement particleboards manufactured using the pressure controlled and
fully saturated curing process. The test results show that wood-cement particleboard with
a cement/wood ratio of 1.5 yields the maximum bending strength (or modulus of rupture).
Consequently, the cement/wood ratio of 1.5 was considered as an optimum value for
making wood-cement particleboard. The test results also indicate that the bending
strength of the composite increases with the increase of cement/wood ratio from 1.0 to
1.5 and then decreases as the cement/wood ratio increases further from 1.5 to 4.0. The

correlation between bending strength and the cement/wood ratio is consistent with test

84

results reported by Prestemon (1976) that wood-cement particleboard had higher bending
strength properties when the cement/wood ratio was increased from 0.75 to 1.5, and also
with the results reported by Moslemi and Pﬁster (1987) that the modulus of rupture of
wood-cement ﬂakeboard increased as the cement/wood ratio was lowered ﬁ'om 3.0 to
1.5.

The test results also show that the modulus of elasticity increases almost linearly as
the cement/wood ratio is increased from 1.0 to 4.0. And the density of the wood cement
particleboard increases linearly with the increase of cement/wood ratios.

The toughness indices derived from test results under bending test are around 6.0.
And the results indicate that a cement/wood ratio has no signiﬁcant inﬂuence on the
toughness index of the particleboard. The toughness index (15) of 6.0 indicates the wood
cement particleboard has a higher energy dissipation capacity comparing with steel ﬁber
reinforced concrete that has a toughness index of 5.0. Therefore, the wood—cement

particleboard can be used where the energy dissipation property is desired.

85

CHAPTER 4
COMPRESSIVE BEHAVIOR OF WOOD-CEMENT

PARTICLEBOARD

4.1 Introduction

Wood—cement particle composites used as structural members in building
construction must often carry compressive loads. It is therefore necessary to characterize
the properties of the composites under compression. The ﬂexural properties of the wood-
cement composites have been investigated extensively. Very little information is
available on the compressive properties of cement bonded wood particle composites. The
use of wood-cement particle bricks in load bearing walls requires information on
compressive load-deformation and toughness. In most case, wood-cement composites
were assumed to be isotropic, and the inﬂuence of particle orientation on mechanical
properties has not been well characterized.

In this research, the compressive properties of the wood-cement particleboard in
directions parallel and perpendicular to the wood ﬁber including the failure modes,
stress-strain relations, and energy dissipation are investigated. A total of 22 specimens
were manufactured using Portland cement (type I) and wood particles from CCA-treated
southern yellow pine. The specimens were fabricated into rectangular short columns
with different aspect ratios (width/height). The cement/wood ratios by weight of the

specimens were 1.5 and 1.0.

86

4.2 Experimental Design

Wood-cement particleboard specimens with the aspect ratios shown in Table 4.1
were cut from thick particleboards (35 mm thick board with cement/wood ratio of 1.5 and
30 mm thick board with cement/wood ratio of 1.0). A schematic of the short column

specimens is shown in Figure 4.1 and Figure 4.2 shows the loading directions. Table 4.1

shows the test matrix.

Table 4.1 Test Matrix for Determination

      

Aspect

specimens

 

Figure 4.1 Short column specimens

87

Loading perpendicular to ﬁber direction

  
 

Loading parallel to ﬁber direction

 

Figure 4.2 Loading directions

4.3 Compression Test
4.3.1 Test method

The compression tests were conducted at the Wood Technology Laboratory in the
Department of Forestry at Michigan State University using an INSTRON machine
(Figure 4.3). A photograph of typical specimens are shown in Figure.4.4. The specimens
were tested with load parallel and perpendicular to the board thickness. The procedures in
ASTM D1037-78 (1996) were applied to evaluate the material properties of the
laboratory manufactured particle composites. The compressive loading rate was is 0.06

inch/minute.

88

 

Figure 4.3 Compression test set-up on IN STRON test machine

‘-

. ImIvumﬂ

rm 9%.

r

 

Figure 4.3 Compression specimens

89

4.3.2 Failure modes

Figure 4.5 shows the failure modes of specimens with an aspect ratio of 2.0 and a
cement/wood ratio of 1.5 loaded parallel and perpendicular to the ﬁber directions. Figure
4.6 shows the failure modes for parallel and perpendicular loading on specimens with an
aspect ratio of 2.0 and a cement/wood ratio of 1.0. Figure 4.7 shows the failure
characteristics for parallel and perpendicular loading on specimens with an aspect ratio of

3.0 and a cement/wood ratio of 1.5.

90

 

a. Parallel loading

 

b. Perpendicular loading

Figure 4.5 Failure modes (h/d =2.0, cement/wood ratio = 1.5)

 

a. Parallel loading

 

b. Perpendicular loading

Figure 4.6 Failure modes (h/d =2.0, cement/wood ratio = 1.0)

 

a. Parallel loading

 

b. Perpendicular loading

Figure 4.7 Failure modes (h/d =3.0, cement/wood ratio = 1.5)

The failure modes on all samples were independent of the cement to wood ratios
and column aspect ratios used in this study. The compressive failure typically occurred
along diagonal bands similar to compression-shear failure on concrete specimens. It may
be concluded that short column specimens failed predominantly in shear under

compressive loading irrespective of the orientation of the wood particles.

4.3.3 Compressive strength and stiffness
The typical load deformation curves corresponding to the specimens shown in

Figures 4.5, 4.6 and 4.7 are shown in Figures 4.8, 4.9 and 4.10, respectively.

 

 

 

 

MID Parallel
-800 ~
Perpendicular
,3 -600-
E «1004
Dimension: 17.5 x 17.5 x 35 mm
-2m .
o - T T T T
0 -0.05 -0.1 -0.15 -0.2 -025

Deformation (in.)

Figure 4.8 Load-deformation curve (h/d = 2.0, cement/wood = 1.5)

94

Load (lb)

Load (1b.)

 

-1m -I

 

Parallel

Perpendicular

Dimension: 15 x 15 x 30 mm

 

 

OJ

r

02 03 04 05 05

Deformation (in.)

Figure 4.9 Load deformation curves (h/d = 2.0, cement/wood =1.0)

 

Parallel

/

Perpendicular

Dimension: 11.7 x11.7 x 35 mm

 

 

 

-0.05

I I T

-0.1 -0.15 -0.2 -0.25

Deformation (in.)

Figure 4.10 Load deformation curves (h/d = 3.0, cement/wood =1.5)

95

The load-deformation curves obtained in the tests display signiﬁcant nonlinearity,
and indicate that the wood-cement particleboard has the capability to absorb energy.
Further, the mechanical properties are not isotropic and indicate directional dependencies.
The directional properties are due to the alignment of wood particles caused by pressing
during the manufacturing process.

Three signiﬁcant observations are made:

(1) Under compression WCPB is much stiffer in the direction parallel to

the particle ﬁber than in the perpendicular direction;

(2) The deformation of the specimens in the direction perpendicular to

particle orientation is much larger than in the parallel direction;

(3) For an h/d ratio of 2.0, the compressive strength of specimens with

cement to wood ratio of 1.5 was higher than that of specimens with a

cement/wood ratio of 1.0 for both loading directions.

The compressive properties of WCPB and normal concrete are shown in Table
4.2. The compressive strength of WCPC was comparable to that of normal concrete, but
the strains at maximum stress and at failure were much larger than that of normal
concrete. The values in Table 4.2 indicate that the highest strength for WCPB occurred
when the cement/wood ratio was 1.5, the specimen aspect ratio (h/d) was 2.0, and the
loading direction was parallel to the ﬁber direction. For this case the compressive
strength of WCPB is comparable to that of normal concrete and the strain at peak stress is

ten times larger than that for concrete.

96

Table 4.2

Compressive Strength and Strain at Peak Stress for WCPB and Normal

 

 

 

 

 

 

 

 

 

Concrete
Loading Mean Std. Dev
Material h/d C/W Direction Strength Strain Strength Strain
// 2600 0.023 243 0.002
WCPB 2-0 1-5 _L 2511 0.123 221 0.019
// 713 0.041 67 0.005
WCPB 2.0 1.0 _1_ 1808 0.284 192 0.032
// 1650 0.013 154 0.002
WCPB 3 .0 1.5 I 2060 0.094 189 0.013
Concrete _ _ _ 3000 0.002 __ _

 

 

 

 

 

 

 

 

 

Note: h/d=aspect ratio; C/W=cement/wood ratio; // = parallel to wood ﬁber direction;

J. =Perpendicular to wood ﬁber direction; Strength=compressive strength;
Strain=strain at peak stress; Concrete=normal concrete.

The compressive strength for specimens loaded perpendicular to the ﬁber
direction was smaller than that for specimens loaded parallel to the ﬁber direction in
some cases and larger for other cases. The strain at peak stress was much larger when
specimens were loaded perpendicular to the ﬁber direction than when they were loaded
parallel to the ﬁber direction and 45 to 140 times larger than for normal concrete. The
difference in properties for columns with aspect ratios of 2.0 and 3.0 may be attributed to
the likelihood of eccentric loading for the longer specimens.

The ability of WCPB to sustain large plastic deformations compared to normal

concrete implies that it has a strong potential to dissipate energy.

97

 

Table 4.3 shows the mean and standard deviation of the stiffness of the wood-
cement particleboard under compression load. The stiffnesses of the WCPB are three to
four times higher when loaded parallel to the ﬁber direction than when loaded
perpendicular to the ﬁber direction. Both stiffnesses are three to four times larger for a

cement/wood ratio of 1.5 compared to those for a cement/wood ratio of 1.0.

Table 4.3 Stiffness of Wood-Cement Particleboard

 

 

 

 

 

 

Material h/d C/W Loading Mean St. Dev
Direction

// 1 1 78 96
WCPB 2.0 1.5

_L 400 23

// 402 34
WCPB 2.0 1.0

.1. 1 10 l 3

 

 

 

 

 

 

 

Notes: // = parallel to wood ﬁber direction; .L = perpendicular to wood ﬁber direction,
h/d=aspect ratio; C/W=cement wood ratio

4.3.4 Toughness

Toughness is a measure of the energy absorbed per unit area of material and is
usually deﬁned as the area under the load-deformation curve. ASTM C1018 is often
used to calculate toughness indices. ASTM C1018 deﬁnes a set of toughness indices for
ﬁber reinforced concrete, which are deﬁned as the area under the load-deformation curve
up to the deformations of 3, 5.5 and 10.5 times the deformation at ﬁrst crack divided by
the area under load-deformation curve up to the ﬁrst crack. In the concrete industry, the

toughness index of plain concrete is 1.0 and the toughness index of steel ﬁber reinforced

concrete is about 5.0.

98

In this study the ASTM C1018 toughness index (15) is used to deﬁne the
compressive toughness of wood-cement particleboard. The toughness index (15) is

deﬁned by the following equation:

_ Area under the load — deformatio n carve up to 36

15 (4.1)

 

Area under the load - deformatio n up to (5

where, 5 is the axial deformation up to the ﬁrst crack.

Table 4.4 shows the calculated toughness index 15 for WCPB with two
cement/wood ratios under uniaxial compression. The calculated toughness index (15) of
the WCPB is about 7.0. The toughness index 15 is close to a constant irrespective of the
loading direction because the deﬁned 36 area under the load-deformation curve can not

cover the values of peak stress and strain at peak stress due to peak stress being too large.

Table 4.4 Compressive Toughness of Wood-Cement Particleboard
Toughness 15

 

 

 

 

Material C/W Mean St. Dev
WCPB 1.5 6.99 0.41
WCPB 1.0 7.03 0.37

 

 

 

 

 

 

Figure 4.11 appears to indicate that the energy dissipation ability of the WCPB

perpendicular to the ﬁber direction is higher than that in the direction parallel to the ﬁber

99

direction. However, the toughness index 15 was close to 7.0 irrespective of the loading

direction.

 

 

Stress (psi)

 

I

 

I I

I

0115 0.1 015 02 025 03 «0.
Strain

25 -0.4 -0.45

(1) Parallel, C/w =l.5, h/d = 2.0; (2) Perpendicular, C/w =1.5, h/d = 2.0.
(3) Parallel, C/w =l.5, h/d = 3.0. (4) Perpendicular, C/w =l.5, h/d = 3.0.
(5) Parallel, C/w =l.0, h/d = 2.0. (6) Perpendicular, C/w =1.0, h/d = 2.0.

Figure 4.11 Stress-strain relationships of the wood-cement particleboard

At the 95% level of conﬁdence, the compressive strength, stiffness and toughness

index of wood-cement particleboards with two different cement/wood ratios are shown in

Figures 4.12 to 4.16.

100

 

 

   

 

A
N

0..

2 3000
V I
g 2500 i
8 2000 I
5 I
o 1500
> I
'63 1000
.8. I
g, 500 r
8 0
U

1 .5 1 .0
Cement/wood ratio

Figure 4.12 Compressive strength vs. cement/wood ratio (loading parallel to ﬁber)

 

 

 

 

E
3000
2 I
v I
r I
"a 2500 I
g I
m 2000 I
o |
.2. I
m
1500
13. .
2‘ I
o 1000
U

 

1.0

Cement/wood ratio

Figure 4.13 Compressive strength vs. cement/wood ratio (loading

perpendicular to ﬁber)

101

 

1400
1200
1000
800
600
400
200

,~'-'.
1
A; .
I .
? I.
3;; 12'
r -. ,_
rs' ‘,
5. ,":~
yr, '1‘

‘TI I—

 

   

Compressive stiffness, (MPa)

1 .5 l .0
Cement/wood ratio

Figure 4.14 Compressive stiffness vs. cement/wood ratio (loading parallel to ﬁber)

 

 

   

Compressive stiffness, (MPa)
8
O

1 .5 1.0
Cement/wood ratio

Figure 4.15 Compressive stiffness vs. cement/wood ratio (loading

perpendicular to ﬁber)

102

 

7.5

 

 

 

Compressive toughness index 15

 

Cement/wood ratio

Figure 4.16 Compressive toughness index [S vs. cement/wood ratio

(irrespective of loading direction)

4.4 Summary

Compressive tests conducted on WCPB show signifﬁcant inelastic behavior. the
composite specimen fails in shear under compressive loading parallel to or perpendicular
to the wood ﬁber direction. Composites with a cement/wood ratio of 1.5 exhibited a
compressive strength comparable to that of normal concrete, but the strain at peak load
was ten to forty ﬁve times larger than the strain at peak load for normal concrete
depending on the direction of the compressive loading. The ability of the composites to
sustain such large plastic deformations suggests that WCPB has a strong potential to
dissipate energy. The toughness index I 5 for the composites with either a cement/wood

ratio of 1.5 or 1.0 was about 7.0, which is seven times larger than that of normal concrete.

103

This study suggests that CCA-treated wood removed ﬁom service can be used for
manufacturing wood-cement particleboard for use in applications where compressive

strength and energy dissipation are desirable properties.

104

CHAPTER 5

IMPACT TESTS ON WOOD-CEMENT PARTICLEBOARD

5.1 Introduction

Information about the impact response and energy dissipation of wood cement
particle composites is needed if these composites are employed as structural members
that may sustain impact loads. While the energy dissipation of wood-cement particle
composites have been evaluated by measuring ﬂexural and compressive toughness
(Wolfe, 1996 and Hsu, 1995), no work appears to have been done using direct impact
tests.

In this study, a falling ball impact test on wood-cement particleboard was
conducted to investigate the impact response and energy dissipation behavior of the
composite. In order to compare its response with those of more traditional building
materials, such as wood and concrete, impact tests on concrete and on a wood board
resting on concrete were also conducted.

These impact tests were evaluated by measuring the acceleration of a steel ball
during impact. A hi gh-precision accelerometer was employed in the impact test to record
the acceleration of the steel ball when it impacted different materials.

A total of ﬁfteen tests were conducted in this research for a steel ball impacting
on the following surfaces: (1) wood-cement particleboard resting on concrete block, (2)
wood board resting on concrete block, and (3) concrete block.

In addition to the direct falling ball impact tests, multi-stress tests on wood-

cement composite tubes were conducted as well. One of the purposes of this test was to

105

investigate the material behavior of the wood-cement particleboard under a multi-stress
state. Another was to investigate the effect of assuming the WCPB as an elasto-plastic
isotropic material by comparing test results with the results of ﬁnite element simulation
that use the isotropic material model presented in Chapter 6. The compressive tests were

conducted using an MTS compression machine.

5.2 Falling Ball Impact Test

5.2.1 Impact test design

The falling ball impact test was designed to investigate the impact characteristics
of WCPB. An accelerometer was mounted in the steel ball for recording the impact
response of the ball as it impacted different materials. Three different materials were used:
WCPB resting on a concrete block, wood board (wood grain parallel to horizontal plan)
resting on a concrete block, and a plain concrete block.

The wood cement particleboard had a dimension of 120 x 120 x 35 mm (length x
width x thickness) and a cement/wood ratio of 1.5. The wood board had a dimension of
120 x 120 x 35 mm and was made of CCA-treated southern yellow pine (the same kind
of wood spices for the wood particles in the wood-cement particleboard). The concrete

had a dimension of 150 x 150 x 150 mm and was made of normal concrete with a
compressive strength of f c = 4000 psi. The steel ball was released from rest at a height,

H, of 152 mm (or 6 inches) above the impact surface.
Figure 5.1 shows schematics of the impact tests. The test matrix detailing the

ntunber of specimens and specimen dimensions are given in Table 5.1.

106

  

Concrete
/

a. Impact on concrete

Figure 5.1

Steel ball .

Wood '1

Concrete

Table 5.1 Test Matrix for lrnpact Test

 
   
 

 

Schematics of impact test

 

Steel ball I

 

b. Impact on wood on concrete c. Impact on WCPB on concrete

 

Concrete

Wood on concrete

WCPB on Concrete

 

Numbers of Test
Specimens

5

5

5

 

Dimensions, mm

 

 

150 x 150 x 150
(concrete)

 

120 x 120 x 35
(wood)

 

120 x 120 x 35
(WCPB)

 

In order to conduct the impact test, a test device was designed and manufactured.

5.2.2 Design of impact device

1

ball. Figure 5.2 is the release mechanism for the impact test.

07

The impact test device includes two parts: one is an impact steel ball mounted with an

accelerometer, and the other is an electro-magnetic release mechanism that released the

 

 

Figure 5.2 Release mechanism

The impact ball selected was a solid steel ball with a diameter of 107 mm and a
weight of 5.0 kg. In order to record the acceleration of the steel ball during the impact,
an accelerometer was mounted in the steel ball. A schematic of the steel ball with the
mounted accelerometer is shown in Figure 5.3.

The accelerometer was recessed into the steel ball to protect it ﬁ'om damage

during impact. Figure 5.4 shows photographs of the ball and mount.

108

Mount bed Accelerometer

d=107mm W=5kg

 

 

Figure 5.3 Schematic of the impact ball

 

Mount Impact ball with threaded hole Impact ball with mount

Figure 5.4 Photographs of impact ball assembly

109

5.2.3 Impact data recording devices

The impact data was recorded using an accelerometer and oscilloscope. A model
353812 piezoelectric accelerometer made by PCB Piezotronics, Inc was used. The
accelerometer can measure the shock exposure experienced by space vehicles and cargo
during stage separation. It may also be used to determine maximum impact acceleration
levels experienced by vehicles and crash dummies. The accelerometer is a high
frequency, quartz shear ICP accelerometer with a measurement range of i IOOOg (: 4905
m/sz). In addition, the accelerometer’s size (Hex x height) is 7.1 x 14.7 (mm) and its

weight is 1.5 gm (0.05 oz). Figure 5.5 shows a photo of the accelerometer.

 

I,“
I»

-y
-

l3.

 

Figure 5.5 353312 accelerometer Figure 5.6 Oscilloscope

The oscilloscope used was an Agilent 54620A instrument made by Agilent
Technologies and is shown in Figure 5.6. It is used to display the time history of the
acceleration and record the test data. The oscilloscope had the capability of storing

recorded data on a computer ﬂoppy disk.

110

5.2.4 Impact test set up and results
The impact test on wood-cement particleboard resting on a concrete block was
conducted for the purpose of determining the impact characteristics of the composite.

Figure 5.7 shows the set-up for the impact test.

Release mechanism

 

Figure 5.7 Impact test setup

Figure 5.8 shows the permanent deformation in the particleboard subsequent to
impact by the steel ball. As shown in Figure 5.8, the impact caused an indentation of
about 1.5 mm at the center of the impact area. Because the board’s deformation is

permanent, a plastic deformation occurs in the wood-cement particleboard due to the

111

impact of falling steel ball. The impact did not produce a crack at the bottom of the

board.

Board top surface

  
 

Impacted dent

Figure 5.8. Permanent deformations in wood-cement particleboard following impact

The accelerometer voltage time histories for the steel ball impacting plain

concrete, a wood-cement particleboard panel resting on concrete and a wood board

resting on concrete are shown in Figures 5.9, 5.10 and 5.11, respectively.

112

. O .
O o
o

,::;2-.:f Agilent Technologies

 

 

 

 

Figure 5.9 Voltage time history for impact on concrete

I Agilent Technologies

 

:

l
H=m ii

 

A mm: = im: LLJ
: i

 

 

Figure 5.10 Voltage time history for impact WCPB resting on concrete

113

~" Agilent Technologies

 

 

 

 

Figure 5.11 Voltage time history for impact on wood board resting on concrete
The conversion from voltage to acceleration is,
10.18 mV = 1.00 g

where g = 9.81 m/sz.

For the case of wood-cement particleboard resting on concrete, the accelerometer time

histories are shown in Figure 5.12.

114

 

Acceleration, (g)

 

 

( Time, (ms)

 

Figure 5.12 Acceleration time history of steel ball (W CPB on concrete)

The velocity time histories were computed from acceleration time histories
through integration. The time was assumed to be zero at initial impact when the
acceleration exceeded that due to gravity. Since the ball was released from rest from a
height of six inches (152.4 mm) above the impact target, the velocity of the ball, V, at

initial impact can be computed using the conservation of energy:

1 ,
—mv‘ =m H
2 0 8

Solving for v0 yields

v0 = ,/2gH =17.2 (m/second)

115

where m is the mass of the steel ball, g is the acceleration under gravity, and H =152 .4

mm. The velocity time histories are obtained through

v(t) = v0 + £a(t)dt, t2 0

where v0 =17.2 m/s

The velocity time history corresponding to the acceleration time history in Figure

5.12 is shown in Figure 5.13.

 

_____.._,W___1

 

 

 

 

 

E
E
E . f . ﬂ
s: -2 2 4 6 8
.é‘
.8.

l D

l >

1

Time, (ms)

 

 

Figure 5.13 Velocity time history of the steel ball (W CPB on concrete)

116

Similarly, the displacement time histories of the steel ball were obtained by
integration of the velocity time histories. Deﬁning the displacement at the top of the
impact surface to be zero, the displacement time histories can be obtained by the

following equation.
do): Emma, with d(0)=0

The displacement time history corresponding to the velocity time history in

Figure 5.13 is shown in Figure 5.14.

 

‘7

 

 

8
-6 -4 2 4 6 8

 

Time, (ms)

AMT “BEMJC am”) i w T

 

 

Figure 5.14 Displacement time history of the steel ball (W CPB on concrete)

117

Figure 5.15 compares the acceleration time histories of the steel ball impacting on

wood—cement particleboard resting on concrete, concrete and wood board resting on

 

   
   

 

 

 

 

 

concrete.
1
l 900 _
l 800 ﬁ /Concrete
1
. 700 «
| 600 “
l ’59 500 ~
1 g 400 -
g g 300 _ WCPB on concrete
1 .2
l 33 Wood on concrete
‘ <
l r— -:.+ Wat,‘;.-_ ‘
‘ b
1 Time, (ms)

 

Figure 5.15 Acceleration time histories of steel ball impacting different surfaces

In Figures 5.15, the maximum acceleration of 800g occurs when the steel ball
impacts concrete. The peak acceleration for impact on WCPB is 220g and that for impact
on the wood board is 180g. The deceleration forces on the steel ball due to impact on the
WCPB and the wood board are close, while the deceleration force for impact on concrete
is about four times greater. The wood-cement particleboard layer reduces the impact

acceleration, and hence the maximum impact deceleration force, by about 72%.

118

Figure 5.16 and Figure 5.17 showed the velocity and displacement time histories
of the steel ball for the three different cases shown in Figure 5.1. In Figure 5.16, the peak
rebound velocities of the steel ball for impact on concrete, WCPB and wood are 1.2, 0.9
and 0.78 (mm/ms), respectively. Therefore, the wood-cement particleboard layer reduces
the peak rebound velocity by 25%. In Figure 5.17, the peak negative displacements
(maximum deformation of specimen due to impact) for impact on concrete, WCPB and
wood are 0.6, 1.6 and 2.0 (mm), respectively. It is the large deformation of wood and
WCPB that results in the reductions in peak acceleration and rebound velocity of the steel

ball.

 

Concrete

 

 

-2

Velocity, (mm/ms)
£- 1

 

 

 

Time, (ms)

 

 

Figure 5.16 Velocity time histories of steel ball

119

 

Concrete

   

 

WCPB Wood

 

Displacement, (mm)

 

Time, (ms)

 

 

Figure 5.17 Displacement time histories of steel ball

5.3 Multi-Stress Test on Wood-Cement Particle Tube

The multi-stress test was conducted by applying a compressive load on a wood-
cement composite ring. The purpose of this test was to verify the accuracy of simulations
(described later) to predict the response of wood-cement particleboard tubes under impact.
The motivation for using tubes is presented in Chapter 7.

In this research, the multi-stress test on the wood-cement particleboard ring was
conducted to investigate the mechanical response of the ring. The WCPB ring was cut
from a ﬂat board with an outer diameter of 110 mm, inner diameter of 60 mm and
thickness of 35 mm. The cement/wood ratio of the composite tube was 1.5.

Figure 5.18 shows the geometry of the test specimen. Two test specimens were
manufactured and tested. The ring was cut from a ﬂat 35 mm thick WCPB. The ﬁbers

therefore randomly oriented in the plane of the ring.

120

 

 

(mm)

60 (mm) 35 mm

 

110

 

Figure 5.18 Geometry of specimen

5.3.1 Test results and analysis

The compressive loading was applied using an MTS machine with a loading
speed of 0.01 mm/second. Two displacement measurements were recorded at the top of
the specimen’s external and internal circles as shown in Figure 5.19. The movement of
the loading crosshead of the MTS machine was used as the displacement of the top of the
test specimen. The loading direction was parallel to the cross section of the tube. The
load-deformation curve was obtained in this test. The compressive test setup was shown

in Figure 5.19.

121

 

Figure 5.19 Compression test setup

The failure mode of the compression test is shown in Figure 5.20. The WCPB
ring showed strong inelastic behavior. Even though the ring cracked and its shape

deformed from a circle to an ellipse, it did not collapse.

122

Compressive load (N)

 

Figure 5.20 Compressive failure of WCPB ring

 

4000
3500 r
3000
2500 r
2000 L
1500
1000
500

ﬁ

7

Tﬁ

 

 

 

1 Q; L l

 

Figure 5.21

0.5 1 1.5 2 2.5 3

Displacement (mm)

Load-displacement curve for multi-stress test on WCPB ring

123

Figure 5.21 shows the relationship between the vertical load and the vertical
displacement at the top of the specimen. The load-displacement curve indicates that the
ring displays elastic-degrading plastic behavior. The behavior is elastic up to a load of

2800 N after which the ring undergoes stiffness degradation.

Figure 5.22 shows the stress states at the points A and B and the stress
distributions along the cross sections 1, II and III in the composite ring under vertical
compressive load. The stress distributions explain the failure cracks observed in Figure
5.20. Figure 5.22 shows that the composite ring is in a multi-stress state under vertical

compression loading.

*
.4

    

 

 

Figure 5.22 Stress state and stress distribution

124

5.4 Summary

Results from a falling ball impact test shows that wood-cement particleboard
possesses good energy dissipation characteristics. The reduction in the impact force when
the steel ball impacts WCPB resting on concrete is similar to that for impact on a wood
board resting on concrete. Compared with the steel ball impacting a concrete block, the
wood-cement particleboard resting on concrete reduced the impact acceleration of the
steel ball, and hence the maximum impact deceleration force, by about 72%. The ability
of the WCPB to reduce such a large portion of the impact deceleration force suggests that
WCPB has a strong potential to dissipate energy.

The multi-stress test conducted on wood-cement particleboard ring samples
showed that the rings display elastic-degrading plastic behavior similar to the behavior

described in Section 3.3.3 for ﬂexural specimens.

125

CHAPTER 6
VALIDATION OF MATERIAL MODEL FOR WOOD-CEMENT

PARTICLEBOARD

6.1 Introduction

Finite element simulation of the falling ball impact test was performed to validate
the ability of the material models in LS-DYNA to adequately represent wood-cement
particleboard and concrete. If the simulations is successful, then preliminary studies on
the impact of a vehicle with a concrete crash barrier enhanced with wood-cement

particleboard panels can be performed using ﬁnite element simulation.

6.2 Finite Element Modeling of Falling Ball Impact Test

One of the important tasks in ﬁnite element simulation is ﬁnite element modeling.
The quality of modeling directly affects to the accuracy of the results of the ﬁnite element
simulation. The quality of modeling depends on many features, such as the power and
ﬂexibility of the modeling software, element modeling, material modeling, contact

modeling and so on.

6.2.1 Modeling software
In the simulation study, four commercial software packages were used. LS-
DYNA (Hallquist, 1999) is used for the impact analysis. The description of LS-DYNA

was presented in Section 2.13. HyperMesh (Altair Engineering, 1999) was used as a

126

preprocessor to develop the ﬁnite element model. HyperGraphic (Atair Engineering,
1999) and LS-POST (LSTC, 2002) were used for post-processing. LS-DYNA and LS-
POST are the products of Livermore Software Technology Corporation. HyperMesh and

HyperGraphic are the products of Altair Engineering, Inc.

6.2.2 Element modeling
The steel ball, wood-cement particleboard and concrete were modeled with 8-

node hexahedron solid elements.

6.2.3 Material modeling

In nonlinear ﬁnite element analysis of vehicle-barrier crashes, the principal
challenge is the capability of the material constitutive models to represent a realistic
response of the barrier under impact loading. In this study, concrete, wood-cement

particleboard and steel were modeled with the material models available in LS-DYNA.

6.2.3.1 Modeling concrete

LS-DYNA offers several concrete material models, Material Model 16
(MAT_PSEUDO_TENSOR) is a model with the minimum required input data (Hallquist,
1999).

Material Model 16 (MAT_PSEUDO_TENSOR) is a pseudo-tensor geological
model with an option for reinforcement. The model offers two major modes: a simple
pressure-dependent Mohr-Coulomb yield surface with the Tresca limit, and two “yield

versus pressure” functions with the capability to transit from one curve to another. The

127

latter one is used with the damage scale option as suggested in the User Manual
(Hallquist, 1999). The equation of state (EOS) type 8 is used, which describes the
volumetric response of the material. In this study, two-yield surface versus pressure
criteria were employed for modeling the concrete.

Material Model 16 is based on the stress-strain relation of concrete under
hydrostatic compression and a two yield surface limit. LS-DYNA uses a tri-linear
relation between hydrostatic stress and volumetric strain to model the concrete behavior

under hydrostatic compression. Figure 6.1 depicts the stress-strain relation adopted in LS-

DYNA.

:0
}

Hydrostatic stress
0)

 

 

l - 4 E
Volumetric strain

Figure 6.1 Hydrostatic stress vs. volumetric strain of concrete (schwer, 2001)

The following algorithm for generating the default EOS (Equation of State)
parameters is used to determine the pressure and volume strain pairs, and an associated
unloading bulk modulus for each of the four points comprising the pressure-volume curve

(Shewer, 2002).

128

Table 6.1 Equation of State Algorithrn for Generic Concrete (Shewer, 2002).

 

 

 

 

 

 

 

 

 

 

 

1’0”“ Bulk Unloading K...- Pressure I: Volume sum .9...
1 E/[3(1— 2v)] 0 0
2 120 f; 02/; —P, /KU,
3 300]: Be - Kuz (8v3 - £v2) .0068
4 Rs 'Ku3 (€124 —€v3) -010

The bulk unloading moduli are further modiﬁed to assure that the largest

unloading modulus is associated with the last point of the curve:

Km = max(KUl’KU3)
Kw = Km (6.1)
Kuz = KUl

and then only the elastic Km and largest unloading moduli KU4 are retained in the

generated EOS.
The elastic modulus of concrete was estimated from the formula provided by the
American Concrete Institute. The estimate of Young’s modulus for concrete, for a

normal weight density of 145 lbf/ft3, is given by the formula

E = 57,000,”; psi (6.2)
and typical values of Poisson’s ratio for concrete range from 0.15 to 0.2. In this study, a

Poisson ratio of 0.2 was used (Nilson, 1991).

129

LS-DYNA also uses the two-yield surface limit shown in Figure 6.2 as a failure
criteria. In Figure 6.2, the lower curve represents the initial yield surface and the upper
curve represents the evolution of the yield surface due to concrete hardening.

P

=a0+———
a1+azp

Omax

 

 

Yield

 
 

\

afailed = 00f +

 
 

alf-l-azp

 

 

Pressure

Figure 6.2 Yield stress vs. mean stress for concrete under hydrostatic compression

The input parameters of the concrete material used in this study are given in Table

6.2.

Table 6.2 Concrete Material Parameters

Densi 2.4 cm
Y ’s Modulus 21526

Poisson’s Ratio 0.2
've S 20

 

130

6.2.3.2 Modeling wood-cement particleboard

Wood-cement particleboard is modeled as an isotropic material due to the small
size of the wood particles. This simplifying assumption is used even though wood
cement particleboard has directional properties.

Material Model 24 (MAT_PIECEWISE_LINEAR_PLASTICITY) was selected
to model wood-cement particleboard. Material Model 24 is suited for elasto-plastic
material with an arbitrary stress-strain curve or tabulated stress-strain data (Hallquist,
1999).

In this material model, the stress-strain pairs are taken from test data and failure is
characterized by a maximum plastic strain limit. The stress-strain curve is deﬁned by the
effective stress and effective plastic strain. The effective plastic strain is a true plastic
strain equal to the total strain minus the elastic strain, while the effective stress is the true
stress corresponding to the effective plastic strain.

Figure 6.3 shows a typical piecewise linear curve of effective stress vs. effective
plastic strain. The piecewise linear curve deﬁnes the yield stress versus the effective

plastic strain. A nonzero yield stress is deﬁned when the plastic strain is zero.

131

“A

.,//—

Effective stress

 

 

’
Ep
Effective plastic strain

Figure 6.3 Effective stress vs. effective plastic strain curve (Hallquist, 1999).

The engineering stress-strain curve does not give a true indication of the
deformation characteristics of a material because it is based entirely on the original
dimensions of the specimen, and these dimensions change continuously during the test.

The true stress 0' is expressed in terms of engineering stress by

a = s(e +1) (6.3)

The true strain 8 may be determined from the engineering or conventional strain e

g = ln (8 +1) (6.4)

To ﬁnd the curve of true stress and true plastic strain, three steps were taken.

First, derive engineering stresses and strains,

Engineering stress: 3 = i (6.5)

0

132

 

 

 

 

Strain

Engineering strain: 8: L (6.6)
0
l
l -3000 ~
1
l
l a
i e
l 8
a
l (I)
l
l l
l o i . . . .
; o -o.05 -o.1 -o.15 -o.2
t
l

 

 

Figure 6.4 Engineering stress vs. engineering strain (W CPB)

Second, calculate the plastic strain,

Plastic strain: Plastic strain cp = total strain (e) — elastic strain (cc)

 

 

Le %= -——= - ’ (am

where ep= plastic strain, sy = yield stress, Py = yield load, A0 = original cross section area,

L0 = original length, and L = current length.

133

Third, using equation 6.3 and 6.4, calculate the effective stress (true stress) and effective
plastic strain (true plastic strain) and make sure a nonzero yield stress is deﬁned when the
plastic strain is zero. Figure 6.5 shows the relation between effective stress and effective

plastic strain.

 

 

 

o L
r . r T *F

0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16

T ﬂ

Effoctlve plastic drain

 

 

 

Figure 6.5 Effective stress vs. effective plastic strain (W CPB)

Table 6.3 shows the input data required for the material model of wood-cement

particleboard in this simulation.

 

 

 

 

 

 

 

Table 6.3 WCPB Material Data
Density 1.044 (g/crn )
Young’s Modulus* 400 (MPa)
Poisson’s Ratio 0.2
Load curve Shown in Figure 6.5
Yield stress (compressive) 17.2 (MPa)
Tensile cutoff stress 11.2 (MPa)

 

 

 

 

_ _
* Young’s Modulus is in the direction perpendicular to the board thickness.

134

6.2.3.3 Modeling steel

Steel material is oﬁen modeled with elastic, kinematic plastic or piecewise linear
plasticity models. When the steel ball impacts the concrete block from 6 inches above the
concrete, the maxiumum stress on the steel ball is well below the yield stress. Therefore,
Material Model 1 (MAT_ELASTIC) was used to model the steel ball. The Materail
Model 1 (MAT_ELASTIC) is a linear isotropic elastic material model (Hallquist, 1999).

The material properties of the steel ball are listed in Table 6.4.

Table 6.4 Material Data for Steel Ball

Densi 7.85 cm
Youn ’s Modulus 210 000

Poisson’s Ratio 0.2
Yield S 200

 

6.2.4 Contact modeling
The contact-surface algorithm in LS-DYNA is a penalty-based method. In the
falling ball impact simulation, there is one kind of contact interaction: steel ball to the
surface of the particleboard for the steel ball impacting the particleboard panel resting on
a concrete block, or steel ball to the surface of concrete for the steel ball directly
impacting the concrete block. The contact algorithms are set as follows:
0 For the steel ball impacting the particleboard panel resting on a concrete block: 1).
Set the particleboard’s top surface as the master surface and the surface of steel
ball as a slave surface; 2) Set the top surface of the concrete block as master

surface and the bottom surface of the particleboard as the slave surface. The

135

nodes on the slave surface cannot penetrate through the master surface. In LS-
DYNA, the Surface to Surface contact type was used.

0 For the steel ball directly impacting the concrete block, set the top surface of
concrete block as the master surface and the surface of the steel ball as a slave
surface. The nodes on the slave surface cannot penetrate through the master

surface. In LS-DYNA, the Surface to Surface contact type was used.

6.2.5 Boundary conditions

In the impact tests, the impact ball falls ﬁeely from a height of six inches (152.4
mm) above the impact target. Therefore, the initial velocity of the ball just prior to impact
is 17.2 m/s. In addition, the constraint condition imposed on the concrete block is that

the the nodes on the bottom of the concrete block are ﬁxed as shown in Figure 6.6.

 
    
  
 

Impact velocrtv Steel Ball

Particleboard

Concrete

Constrained
Figure 6.6 A h ILLUUVA (LIAM UVW‘M‘L‘J Wl‘ultions

136

6.2.6 Database output node
In order to compare the simulation results with the test results, in the simulation
model the results were output at a node at the exact location of the accelerometer

mounted within the steel ball. The output node location is shown in Figure 6.7.

Output node

  

Figure 6.7 Output node location

6.2.7 Impact simulation and result anslysis

Two types of impact simulations were conducted as described above. The ﬁnite
element models for these simulations are shown in Figure 6.8. The simulations were
carried out on a multi-processor Sun Unix workstation located in the Division of
Engineering Computing Services in the College of Engineering at Michigan State

University.

137

gnu
II
II
ianll
1:: :i:,.-::

 

a. Ball impacting particleboard resting concrete b. Ball impacting concrete

Figure 6.8 Finite element models

All the following test and simulation results, such as accelerations, velocities and
displacements, are in the vertical Z direction. Figure 6.9 shows the acceleration time
histories of the steel ball from the impact test and simulation for impact on the concrete
block. Figure 6.9 shows that the simualtion results have a very good agreement with the
impact test results except that there is about a 7% difference in peak the acceleration
between the test and simulation. The simulation results indicate that Mat__Pseudo_Tensor
model can be effectively applied to model concrete material.

Figure 6.10 shows the acceleration time histories of steel ball ﬁ'om the impact test
results and simulation for impact on WCPB resting on a concrete block. The ﬁgure
shows that the simulation result matches the test result quite well, especially ﬁ'om the
impact point to the peak acceleration. But, in the simulation results, the accelerations

following the peak acceleration display spurious oscillations.

138

The- spurious oscillations represent high ﬁ'equency motions that the analysis
model cannot acurately resolve and are mainly created by the algorithm of the analysis as
well as the viscosity of the material. The explicit central difference method that is used
in LS-DYNA does not automatically dissipate high frequency numerical oscillations. It is
possible to use implicit time integration in LS-DYNA and this will dissipate the high
frequency numerical oscillations. However, implicit time integration requires expensive
iterative procedures that are time consuming and expensive. In practice, because a certain
degree of spurious oscollation is not debilitating, employing explicit analysis and
enforcing an artiﬁcial damping control to smooth the oscillations may be a good chice.

In this falling ball problem, an explict analysis is applied and the spurious
oscillations were controlled with the LS-DYNA control command Control-Bulk-
Viscosity. The parameters used in the control command were the LS-DYNA
recommended default values related to sound wave propagation in general elasto-plastic
materials. In Figure 6.10, the spurious oscillations are limited and the simulation results
match the test results quite well.

In Figure 6.9, Figure 6.10, Figure 6.11 and Figure 6.12, t = 0 represents the
moment the steel ball ﬁrst touches the specimen’s top surface. The measured
acceleration time history was shiﬁed in accordance with this. In Figure 6.12, the zero
point in the displacement axis was deﬁned at the specimen’s top surface. Negative
displacement represents the deformations of the particleboard after the steel ball hits the
board’s top surface in the downward direction.

Figure 6.11 and Figure 6.12 show the velocity and displacement time histories

from the impact test and simulation for impact on the WCPB resting on a concrete block,

139

respectively. The simulation results match the results from the impact test quite well and
demonstrate that the nonlinear ﬁnite element analysis code LS-DYNA can effective
simulate the impact test on the wood-cement particleboard and that the wood—cement
particleboard and concrete can be modeled reasonably by the LS-DYNA elasto-plastic
material model (MAT_PIECEWISE_LINEAR_PLASTICITY) and geological material
model (MAT_PSEUDO_TENSOR), respectively.
The conditions at t = O are therefore
a(0) = g, v(0) = -l7.2 m/s, d(0) = 0.

where a = acceleration, v = velocity, and d = displacement.

 

900 1

i 30° J’ g —— Impact Test
700 1 m

600 4‘ —— Simulation

500 J

400 -«

300 J

200 ‘
1oo

"gnomes-aw "x _ "7'“ w‘_,_ v-7— Y——vvvwm

.4 -2 -100 o 2 4 6 8

(g)

 

 

Acceleration,

Time, (ms)

 

 

Figure 6.9 Acceleration time histories from impact test and simulation

for impact on the concrete block

140

 

Acceleration, (g)
g; 8
L! . L4_J*
\—

Tirne, (ms)

 

.. .r . i ll
fury-"Pu rr (,7 wrigﬂnl
. .- ,,

‘1“ .' “ 14: ‘i'lrir-
' J. 1'! I. ll" ww.‘

Impact test

Simulation

10 12

 

Figure 6.10 Acceleration time histories from impact test and simulation

for impact on WCPB resting on a concrete block

 

 

L 2

Velocity, (mm/ms)

 

Time, (ms)

   

Impact test

Simulation

 

 

Figure 6.11 Velocity time histories from impact test and simulation (WCPB)

141

 

 

10 J -— Impact test
""'"" Simulation

81

61

-4 -2 10 12

Displacement (mm)
M CD N
z," __.L_e__<
‘ _
O) -1
CD 1 t“

Time, (ms)

 

 

Figure 6.12 Displacement time histories from impact test and simulation (W CPB)

6.3 Finite Element Modeling of Multi-Stress Test

In order to verify that the material model in LS-DYNA can be used to model
wood-cement particleboard that is subjected to a multi-stress state, the ﬁnite element
simulation of the multi-stress test on the WCPB tube was conducted. The software of LS-
DYNA, Hypermesh, and LS-POST were used for simulation, preprocessing and post-

processing, respectively.

6.3.1 Element modeling
The wood-cement particleboard tube, loading cross head and supporting plate of
the loading machine were modeled using 8-node solid elements. The ﬁnite element

model is shown in Figure 6.13.

142

Loading crosshead

Composite tube

Supporting plate

 

 

Figure 6.13 Finite element model

6.3.2 Material modeling
The wood-cement particleboard tube was modeled with Material Model 24
(MAT_PIECEWISE_LINEAR_PLASTICITY). The loading crosshead and supporting

plate of the loading machine were modeled with Material Model 1 (MAT_ELASTIC).

6.3.3 Contact modeling

In the multi-stress test simulation, two contact interactions occurr between the
loading machine and the testing specimen. As the composite tube deforms, the top and
bottom surfaces of the composite tube contact with the surface of the loading crosshead
and the surface of the supporting plate, respectively. The contact algorithms are set as

follows:

143

0 Set the top surface of the composite tube as the slave surface and the surface of
the crosshead as the master surface. The nodes on the slave surface cannot
penetrate through the master surface. In LS-DYNA, the Surface to Surface
contact type was used.

0 Set the bottom surface of the composite tube as the slave surface and the surface
of the supporting plate as the master surface. The nodes on the slave surface
cannot penetrate through the master surface. In LS-DYNA, the Surface to

Surface contact type was used.

6.3.4 Boundary conditions

In this simulation, there are two different boundary conditions consisting of the
loading velocity and support conditions. The velocity condition was set to the loading
speed of 0.01 mm/second, and ﬁxity was imposed on the nodes along the bottom of the

supporting plate. The boundary conditions are shown in Figure 6.14.

lVelocity =0.01 min/second

  
   

III'IIIIIII Constraintnodes
“II-WIIIIIIII

Figure 6.14 Boundary conditions

6.3.5 Database output node
In order to compare the simulation results with the test results, the simulation
results were output at the node on the top surface of the tube that has initial contact with

the crosshead. The output node location is shown in Figure 6.15.

Output node

  

Figure 6.15 Location of output node

6.3.6 Multi-stress test simulation

Figure 6.16 shows the largest deformation of the ﬁnite element model in the
simulation. The shape of the deformed ﬁnite element model is close to the shape of the
deformed test specimen in Figure 5.20. The deformed ﬁnite element model does not
show cracks. Because LS-DYNA uses a Lagrangian formulation in which the mesh
moves with the material, cracks are generally not introduced in the ﬁnite element model
without special treatment. It is possible to utilize the DELETE command to make cracks
occur. The DELETE command deletes any elements in which elemental stresses or
strains exceed speciﬁed limits. However, in this research the focus was on the deformed

shape and the load-displacement curve of the ﬁnite element model, and the DELETE

145

command was not employed in the simulation. It should be noted that the analysis
properly accounts for failure of the material (i.e., cracks) by reducing the stiffness of
failed elements. Therefore, although cracks are not visible, the analysis properly accounts

for their presence.

 

Figure 6.16 Deformed FE model

Figure 6.17 shows the load-displacement curves obtained from the compressive
test and the simulation. The simulation results match the compressive test results quite
well. The simulation results indicate that the elasto-plastic material model
(MAT_PIECEWISE_LINEAR_PLASTICITY) adequately models the wood-cement

particleboard under a multi-stress state even though the material is truly orthotropic.

146

 

4000

 

 

3500 [
A 3000 IL Simulation
E 2500 .
(I)
“g 2000 L
1000 l
500 r.’
O L l _l 1 AA.
0 O 5 l 1.5 2 2 5 3
Displacement (mm)
Figure 6.17 Load—displacement curves from test and simulation
6.4 Summary

The nonlinear, explicit, three-dimensional ﬁnite element analysis code LS-DYNA
was applied to simulate a falling ball impacting a wood-cement particleboard resting on
concrete block and a concrete block as well as the multi-stress test on a composite tube.

The agreement between simulation results and experiments indicate that the
nonlinear ﬁnite element analysis code, LS-DYNA, could be successfully used to simulate
impact on wood-cement particleboard. The simulation results also indicated that the LS-
DYNA material models MAT_PIECEWISE_LINEAR_PLASTICITY and MAT_
PSEUDO_TENSOR can be effectively employed to model wood-cement particle board
and concrete, respectively. The simulation result of the multi-stress test indicates that
although the WCPB is an orthotropic material, the isotropic plasticity model in LS-

DYNA is able to provide very good results.

147

CHAPTER 7
FINITE ELEMENT SIMULATION OF VEHICLE CRASHING ON

CONCRETE BARRIER ENHANCED WITH WCPB

7.1 Introduction

Concrete crash barriers are designed to prevent vehicle passengers from injury
and provide certiﬁable levels of vehicle containment. However, accident data shows that
a large number of deaths still occur as a result of vehicles impacting concrete barriers. To
improve roadside saﬁey it may be feasible to install energy absorbing devices to enhance
the impact characteristics of existing concrete barriers so as to reduce injury to
passengers.

An enhanced barrier should be designed to absorb as much impact energy as
possible so as to minimize the deceleration forces on vehicles in the case of vehicle
impact and at same time maintain the integrity of the barrier. One approach is to install
cushioning material on the side surface of concrete barriers. However, this kind of
material should be durable and cost effective. Wood-cement particleboard is durable, cost
effective and its use in high volume is favorable to the environment. The potential of
using ECPB as cushioning material on concrete crash barriers is investigated in this
chapter.

The nonlinear, explicit, three dimensional ﬁnite element analysis code LS-DYNA
was used to perform crash simulation of vehicle against a concrete barrier enhanced with
wood-cement particleboard elements. The impact deceleration force was taken to be the

key parameter related to occupant injury. The geometry of WCPB elements that was

148

attached to the sides of concrete crash barriers were varied to minimize the impact forces
on the vehicle so as to reduce the risk of occupant injury during the crash. Three
different wood-cement particleboard element geometries were considered: a ﬂat board, a

wave shape and bundles of tubes.

7 .2 Simulation Test Design and Evaluation Criteria

7.2.1 Concrete barrier

A vertical concrete barrier was selected to conduct crash impact simulations. The
compressive strength was taken to be 4000 psi. The barrier had a height of 1066 mm, a
width of 190 mm and a length of 5000 mm. The longitudinal reinforcements consisted of
six steel bars with a diameter of 15 mm, and the vertical stirrups were made of steel bars
with a diameter of 8 mm at a spacing of 400 mm along the barrier’s longitudinal direction.

Figure 7.1 shows the geometry of the concrete barrier and steel reinforcement.

 

CD8

 

(1315

1 066
.‘.\
\V

 

 

 

 

 

I 3 <1

 

Figure 7.1 Geometry of the concrete barrier and steel reinforcement

149

7.2.2 Enhanced concrete barrier

The enhanced barrier contains a wood-cement particleboard facade installed on
the side surface of the existing concrete barrier where impacts occur. Figure 7.2 shows
the enhanced vertical concrete barrier with a ﬂat wood-cement particleboard facade. The
enhanced concrete barrier is expected absorb more impact energy and reduce the impact
deceleration force on vehicles during impact.

As both wood-cement particleboard and concrete belong to the family of cement-
bonded composites, they have compatible properties and can be easily bonded with
cement if necessary. However, it is proposed that the wood-cement particleboard facade
be fastened using metal fasteners on the side surface of the concrete barriers. This

installation will enable facades damaged by vehicle impacts to be replaced easily.

Vertical Concrete barrier

  

WCPB facade

Figure 7.2 Enhanced vertical concrete barrier

Three types of wood-cement particleboard facades were investigated. These

consisted of a ﬂat facade, a wave shape and bundled tubes. Figure 7.3 shows the cross

the section of particleboard facades.

150

 

 

a. Flat

Figure 7.3

Cross section of particleboard facades

7.2.3 Simulation test matrix

In this study, frontal crash impact simulations were investigated. In order to ﬁnd a
suitable solution for using wood-cement particleboard facades to improve the impact
characteristics of the existing concrete barrier, ﬁve different WCPB facades were
investigated. The vehicle impact speed was 30 mph or 13.4 mm/ms and the total
simulation time was set to 100 milliseconds. Table 7.1 shows the simulation matrix for
the frontal crash impact studies presented herein.

performed for different WCPB thickness and geometric parameters, and the ones

 

b. Wave shape

presented here yielded the best results.

@©®

c. Bundled tubes

Table 7.1 Physical Properties of WCPB Facades

A number of other studies were

 

 

 

 

 

 

 

 

 

Double Triple Triple-tube Four-tube
WCPB Shape Flat wave wave bundle bundle
Thickness (mm) 35 35 25 25 25
Geometric =225 l=225
parameters (mm) h=60 h=60 r =112 r =76

 

151

 

7.2.4 Evaluation conditions and criteria

The simulation conditions follow the guidelines in the National Car Assessment
Program (NCAP) and the National Cooperative Highway Research Program Report 350
(NCHRP350). As NCAP recommends, the safety features of a vehicle were evaluated by
a frontal crash on a rigid barrier at a velocity of 30 mph. The crashing vehicle model was
the 820C that was recommended by NCHRP350. The 820C model represents a small car
with a weight of 820 kg.

The evaluation criteria used in these simulations are based on the acceleration of
the vehicle during impact. The immediate causes of injury or death in trafﬁc accidents
are numerous, such as broken glass, ﬁre, crushing of the vehicle, hurling of the passenger
against the car, etc. However, a primary cause of injury is the acceleration a passenger
experiences. The acceleration criteria used in this chapter are recommended in
Information About Automobile Accidents (2003) and the Federal Motor Vehicle Safety
Standard (FMVSS 214, 2000):

0 People can sustain an acceleration of 20g without injury.
0 People can tolerate accelerations up to 80g without serious injury.

0 Pelvic acceleration should be less than 130g for a survivable injury.

7.3 Finite Element Modeling

The most difﬁcult task in ﬁnite element simulation is ﬁnite element modeling.
The quality of modeling directly affects the accuracy of the results of the ﬁnite element
simulation. The quality of the modeling depends on many elements, such as software,

element modeling, material modeling and so on.

152

In this simulation study, four commercial software packages were used. LS-
DYNA was used for crash analysis, HyperMesh was used as a preprocessor to develop

the ﬁnite element model, and HyperGraph and LS-POST were used for post-processing.

7.3.1 Element modeling

The wood-cement particleboard and concrete were modeled with solid elements,
and the steel reinforcement inside the concrete barrier was modeled with beam elements.
The extremities of the beam elements were merged with the nodes of the solid elements.
Thus, the reinforcement cannot debond from the concrete. In addition, it was assumed
that no frictional sliding occurs between the particleboard and the concrete surface. The
comer nodes of the elements for the particleboard were merged with the nodes of the
elements for concrete. Figures 7.4 and Figure 7.5 show the element connections. Figure

7.6 shows the mesh for the steel bar inside the concrete barrier.

Steel bar Concrete

Concrete

 

Figure 7.4 Concrete and steel bar elements Figure 7.5 WCPB and concrete elements

153

LEI _ r . kagm Steel bar inside concrete barrier

 

 

 

 

 

 

 

 

Figure 7.6 Steel reinforcement meshed with beam elements

7.3.2 Material modeling

In nonlinear ﬁnite element analysis of vehicle-barrier crashes, the principal
challenge is the capability of the material constitutive models to represent a realistic
response of the barrier under impact loading. In this study, concrete, wood-cement
particleboard and steel bars were modeled with the material models available in LS-

DYNA.

7.3.2.1 Modeling concrete
Material Model 16 (MAT_PSEUDO_TENSOR) was used to model concrete
material. The details of this material model were described in Chapter 6. The material

parameters for the concrete are shown in Table 7.2

Table 7 .2 Concrete Material Parameters

 

 

 

 

 

 

 

Density 2.4 (g/cm3)
Young’s Modulus 21526 (MPa)
Poisson’s Ratio 0.2

Com ressive Stren 20 MPa)

 

154

7 .3.2.2 Modeling wood-cement particleboard

Material Model 24 (MAT_PIECEWISE_LINEAR_PLASTICITY) was selected
to model the WCPB. As described in Chapter 6, the WCPB was assumed to be an
isotropic material, which is an approximation. For elements in tension, failure was
controlled with the modulus of rupture (or tensile strength). For elements in compression,
the failure was controlled with the compressive strength. All input data for the wood-
cement particleboard were exactly the same as the values used in Chapter 6 and are

reported in Table 7.3.

Table 7.3 WCPB Material Data

 

 

 

 

 

 

TJeTrsity 1.044 (g/cm3)
Young’s Modulus 400 (MPa)
Poisson’s Ratio 0.2
Yield stress (compressive) 17.2 (MPa)
Tensile cut off (MOR) 11.0 (MPa)

 

 

* Young’s Modulus is in the direction perpendicular to board thickness.

7.3.2.3 Modeling steel

Steel material is often modeled with elastic, kinematic plastic or piecewise linear
plasticity models. Preliminary simulations in this study indicated that the maximum stress
in the steel bars inside the concrete barrier was signiﬁcantly lower than the yield stress.
Therefore, Material Model 1 (MAT_ELASTIC) was used to model the steel
reinforcement. Material Model 1 (MAT_ELASTIC) is a linear isotropic elastic material

model (Hallguist, 1999). The material properties of the steel bar are listed in Table 7.4.

155

Table 7.4 Steel Bar Material Data

 

 

 

 

 

 

Density 7.85 (g/cm3)
Young’s Modulus 210,000 (MPa)
Poisson’s Ratio 0.2

Yield Strenﬂ 200 (MPa)

 

7.3.3 Contact modeling

The contact-surface algorithm in LS-DYNA is a penalty-based method as
described in Chapter 6. In vehicle crash simulation, the contact interactions are much
more complicated than in the falling ball impact problem modeled in Chapter 6.

In vehicle crash simulations, there are three kinds of contact interactions: vehicle
body to barrier surface, vehicle tires to the ground surface, and vehicle self contact
(component to component and component self contact due to folding). The contact
algorithms were set as follows:

0 For the contact interactions between vehicle and barrier surface, set barrier
surface as master surface and vehicle surface at impacting side as a slave surface.
The nodes on the slave surface cannot penetrate through the master surface. In
LS-DYNA, use the Surface to Surface contact type.

0 For the contact between vehicle tires and the ground surface, set the ground
surface as master surface, and the nodes on the tires as slave nodes. Use the Node
to Surface contact type.

0 For vehicle self contact, specify that nodes on one component of the vehicle
cannot penetrate through any other components of the vehicle. Use the Single

Surface (self contact) contact type.

156

7.3.4 Vehicle model

In this simulation study, a 1975 Honda Civic vehicle model (820C) was selected
to conduct crash simulation. The 820C model was released to the public by the National
Crash Analysis Center (NCAC). The 820C model has a weight of 820 kg and 5400
elements including shell elements, beam elements and solid elements. This vehicle model
was built for frontal crash impact and angle impact. The Federal Highway Administration
has approved the model for use in small vehicle trafﬁc crash simulation studies. Figure

7.7 is the ﬁnite element model of the vehicle.

 

Figure 7.7 Finite element model of the1975 Honda Civic (820C)

7.3.5 Center of gravity of vehicle

In most car vehicles, the center of gravity (CG) of the car is located between the

two frontal seats and about 100 to 300 mm above vehicle’s ﬂoor. In this vehicle model,

157

its center of gravity is between the two frontal seats and 273 mm above the vehicle’s
ﬂoor or 57 8 mm above the ground.

In order to record the vehicle’s impact response, an accelerometer was mounted
at the center of gravity of the vehicle. Accelerations, velocities and displacements of the
vehicle were output from the accelerometer. Figure 7.8 shows the location of the center

of gravity and accelerometer.

 

Figure 7.8 CG and accelerometer location

7.3.6 Boundary conditions

Two kinds of boundary conditions were imposed in the crash simulation. One
boundary condition is the impact velocity of the vehicle. In this frontal crash impact
simulation, the impact velocity of the vehicle was 30 mph or 13.4 mm/ms. The impact

velocity was imposed on all nodes of the crash vehicle.

158

Another boundary condition is the constraint condition imposed on the concrete
barrier and the ground. The bottom of the concrete barrier and the ground are ﬁxed.
Therefore, all nodes at the bottom of the concrete barrier and on the ground elements

were ﬁxed in x, y and 2 directions. Figure 7.9 shows the boundary condition.

Velocity = 30 mph '

 

 

 

Constrained nodes

Figure 7.9 Boundary conditions of the FE model

7.4 Finite Element Simulation

As shown in Table 7.1, six different types of frontal vehicle-barrier crash
simulations were conducted. In each crash event, the deceleration force on the vehicle
during impact is the most important because this correlates with the severity of occupant
injuries. To signiﬁcantly reduce the deceleration force on the vehicle, the wood-cement
particleboard elements must dissipate a large amount of impact energy by undergoing

large deformations. Concrete barriers enhanced with wood-cement particleboard

159

elements having different geometric conﬁgurations were studied to determine which

conﬁguration is most effective in reducing impact forces.

7.4.1 Frontal crash set up

Six different types of frontal crash simulations with different enhanced barriers
were conducted using LS-DYNA. These included crash on (1) a plain concrete barrier
and crashes on the concrete barrier enhanced with WCPB facades having (2) ﬂat, (3)
double-wave shape, (4) triple-wave shape, (5) triple-tube bundle, and (6) four-tube
bundle facade, respectively.

The ﬁnite element models of the six different frontal crash simulations are shown

in Figures 7.10 to 7.15.

a.
are, .
/"v”//// 47/ ‘ '

I I
”’05:

 

Figure 7.10 Concrete barrier

160

 

Figure 7.11 Concrete barrier with ﬂat WCPB facade

////// "'I
///// ”ff/g,” "I”:
III/”I’ll,” If], 1:55.:

 

Figure 7.12 Concrete barrier with double-wave WCPB facade

161

 
      
 

x/.
//
///

244/4

3937 / /
I l/l/II/l/l/Ill:
.. 40?

s

‘l
-“ \‘II ‘
New“?

Figure 7.13 Concrete barrier with triple-wave WCPB facade

 

Figure 7.14

Concrete barrier with triple-tube bundle WCPB facade

162

.

‘l

.
§‘-

~

. 9.!

 

Figure 7.15 Concrete barrier with four-tube bundle WCPB facade

7.4.2 Simulation results

The nonlinear, explicit, three-dimensional ﬁnite element simulation for a vehicle
barrier crash creates two major types of outputs that include XY pair data and graphic
images of vehicle—barrier crash at every time step. For the analysis presented here the
time step was taken to be one millisecond.

Figure 7.16 to Figure 7.21 show the 820C vehicle crashing on a plain concrete
barrier, and a concrete barrier with each of the ﬁve WCPB facades shown in Figure 7.11
to 7.15.

Each ﬁgure shows snapshots at ﬁve different time steps: t = 0, 10, 30, 70 and 100

(ms). The crash characteristics at each of these time steps are as follows:

163

t = 0 ms: time step just before the vehicle impact the barrier.
t = 10 ms: time at which the frontal bumper is deformed to its maximum level and just
before the vehicle’s ﬁ'ontal head contacts with the barrier.
t = 30 ms: time at which the vehicle’s deceleration force reaches the maximum value.
t = 70 ms: time at which the vehicle just loses contact with the barrier.
t = 100 ms: time at end of the simulation.
The characteristics described above do not occur exactly at the indicated times
(t=10, 30 and 70 ms) for all cases but occurs within 1 or 2 millisecond of these times.
It should be noted that the facades do not appear deformed in Figures 7.18 to 7.21

because the cross section of the facades and barriers shown are outside the impact region.

164

t=0ms

 

 

 

 

 

 

 

 

 

 

 

Figure 7.16 Crash on concrete barrier Figure 7.17 Crash on ﬂat WCPB facade

165

t=0ms

 

 

 

10ms

t

 

 

 

t=30ms

 

 

 

= 70 ms

t

 

 

 

t=100ms

 

 

 

Figure 7.19 Crash on triple-wave

Figure 7.18 Crash on double-wave WCPB

WCPB facade

facade

166

t=0ms

 

 

10 ms

t

 

 

=30ms

t

 

 

=70 ms

t

 

 

 

100 ms

t

 

 

 

 

Figure 7.21 Crash on four-tube WCPB

20 Crash on triple-tube WCPB

Figure 7

facade

facade

167

Figure 7.22 to Figure 7.27 show the damage on the vehicle at t = 70 ms for impact

against different barriers.

 

Figure 7.22 Vehicle damage after crash on concrete barrier (at t = 70 ms)

 

Figure 7.23 Vehicle damage after crash on ﬂat WCPB facade (at t = 70 ms)

168

 

Figure 7.24 Vehicle damage after crash on double-wave WCPB facade (at t = 70 ms)

 

Figure 7.25 Vehicle damage after crash on triple-wave WCPB facade (at 70 ms)

169

 

Figure 7.26 Vehicle damage after crash on triple-tube WCPB facade (at t = 70 ms)

\\\\\\ \\
\\'\“\\\\\\\:‘, \S‘
\\ \\\\
\\\‘\\\“‘

 

Figure 7.27 Vehicle damage aﬁer crash on four-tube WCPB facade (at t = 70 ms)

170

These ﬁgures show the damage to the vehicle for each different crash case. These
are two noticeable characteristics in Figure 7.16 to 7.27: one is the upward lift of the
vehicle’s rear lift; another is the appearance of the vehicle damage. The simulation results
show that upward lift of the rear wheel ends and the appearance of external damage vary
from one to another. However, the appearance of the external damage is clearly related
to the barrier’s geometric conﬁguration. The ﬂatter the contact surface of the barrier, the
better the vehicle looks after impact. For good energy absorption, the facade’s height
should cover the height from the bottom of the bumper to the top of the vehicle’s frontal
head.

The plain concrete barrier and concrete barrier with the ﬂat WCPB facade
provide similar damage to the vehicle after impact than the other four types of facades.
The appearance of the damaged vehicle was the worst for the concrete barrier enhanced
with double-wave WCPB facade (Figure 7.18 and Figure 7.24). This was because the
height of the WCPB panel was smaller than the height of the vehicle’s frontal head. In
Figure 7.18, the double-wave WCPB facade only covers the bumper. The lower part of
the vehicle’s frontal head impacted the WCPB facade, whereas the upper part impacted
the concrete part of the barrier, thereby causing the worst damage appearance to the
vehicle’s head.

The acceleration, velocity and displacement time histories at the center of gravity
of the crashing vehicle for the six different barrier types described in Table 7.1 are shown

in Figure 7.28, Figure 7.29 and Figure 7.30, respectively.

171

 

 

 

 

 

 

300 ,
Double wave
250 *- . Triple-tube
E 200 . Hg Four-tube
v ‘ . \\
Triple-wave/ \r‘
E “‘\R.°°"~.....
a) 150 ‘4 \r . '0...
g ' Concrete x-
l -5 0.
"g. 100 ‘ Flat \\
.52
Q 50 -
0 T I l l
0 20 40 60 80 100
Time (ms)

Figure 7.28 Displacement time histories of vehicle

 

14 l Triple-tube
12 j

  
 
 
   
 

Double wave

 

Velocity (mm/ms)
O)

 

 

 

4 1 ‘ \ » Four-tube
2 4, Triple-wave—A,‘
0 IL I T T r \‘o. \ F\ r
-2 5 1o 15 2o 25 '
.4
_6 I

Time (ms)

Figure 7.29 Velocity time histories of vehicle

172

 

 

     
 

Acceleration (g)

Four-tube —'~" 1, ~
'° \ ‘. o

'100 i Triple-wave r"
\
-120 ~ Concrete

 

 

 

Figure 7.30 Acceleration time histories of vehicle

In Figure 7.28, the largest displacement occurs for impact on the triple-tube and
four-tube facades, and the smallest displacement occurs for impact on the plain concrete
barrier and the ﬂat WCPB facade. In Figure 7.29, the vehicle impacting the tube facades
required more time to reach zero velocity than that in the other four cases. This indicates
that the tube facades experienced larger deformation, dissipated more impact energy, and
were more effective in moderating the impact forces on the vehicle.

The vehicle decelerations during impact are shown in Figure 7.30. Localized
peaks appear in two groups: when the vehicle’s bumper contacts the barrier and when the
full vehicle head contacts the barrier.

In the ﬁrst group of localized peaks (between 5 and 10 ms), the deceleration for
impact on the concrete barrier reaches the maximum value at a time of about 8.5
milliseconds. The deceleration for impact on the ﬂat WCPB facade is essentially identical

to that for impact on the concrete barrier. For the double-wave and triple-wave facades,

173

the accelerations are essentially identical in the ﬁrst group of localized peaks, and the
maximum values are 70% lower than that for impact on the concrete barrier. For impact
on the tube facades, there is no obvious maximum in the ﬁrst group.

In the second group of localized peaks (between 22 and 26 ms), the maximum
decelerations for the ﬂat, triple-wave, four-tube and triple-tube facades were reduced by
7%, 9%, 20% and 30%, respectively, compared to the maximum deceleration for the
concrete barriers. For impact on the double-wave facade, however, the maximum
deceleration increased by 8% compared to that for impact on the concrete barrier. It is
evident that the geometric conﬁguration of the particleboard facade has a signiﬁcant
inﬂuence on the energy absorbing behavior of the enhanced barrier. The simulation
results show that with a good geometric conﬁguration of the particleboard facade it is
possible to reduce the peak deceleration by 30%.

The presence of a bumper on the vehicle makes its impact on the ﬂat WCPB
facade somewhat different from the impact of the steel ball studied in Chapter 5. The
steel ball was quite rigid and even a ﬂat WCPB facade was able to reduce the maximum
deceleration compared to impact on plain concrete. However, in the vehicle the bumper
absorbs most of the energy, especially during the early phase of the impact. This is why
in the ﬁrst group of localized peaks, which represents collision of the bumper, the
maximum accelerations for impact on the ﬂat WCPB facade and the plain concrete
barrier are essentially identical. In the second group of localized peaks, which represent
collision of the more rigid frontal head of the vehicle, there is a slight reduction in the

maximum deceleration for impact on the ﬂat WCPB facade compared to impact on the

174

plain concrete barrier. However, the decrease of 7% is still much smaller than the 72%
reduction observed for the steel ball in Section 5.2.

The results indicate that the triple-tube facade absorbs the most energy amongst
the conﬁgurations considered and consequently generates the lowest deceleration forces
that the vehicle passengers have to withstand during impact. The triple-tube facade is

most effective because it deforms the most as shown in Figure 7.31.

  

3-tube panel

Figure 7.31 Deformation of triple-tube facade after impact

As mentioned in Section 7.2.4, passengers in vehicles can tolerate a maximum
acceleration of 80g during impact without serious injury. The deceleration on passengers
should be smaller than the deceleration of the vehicle during impact because there is
relative movement between the vehicle and passengers. In this simulation, for the vehicle
impacting the three-tube facade the deceleration was 80g and the passengers inside this
vehicle should therefore experience decelerations smaller than 80g and not be subjected

to serious injury.

175

7 .4.3 Comparison of WCPB, concrete and steel triple-tube facades

To further investigate the contribution of the shape and material of the tube facade
in reducing the impact acceleration, crash simulations onto a concrete barrier enhanced
with a triple-tube facades made of steel and concrete were conducted. The geometry of
the concrete triple-tube facade was the same as that of the WCPB triple-tube facade. For
the steel triple-tube facade, two cases were studied with each tube having the same outer
diameter as that of the WCPB tube but with thicknesses of 3 mm and 5 mm. Figure 7.32
to Figure 7.35 show the crash images at the largest tube deformation for each case. As
evident, in these ﬁgures, the WCPB facade has modest deformation (maximum
displacement in vehicle moving direction = 32 mm), the concrete tube facade has very
little deformation due to its rigidity (maximum displacement in vehicle moving direction
= 1.2 mm), and the 3 mm thick steel tubes have very large deformation because of its
high ﬂexibility (maximum displacement in vehicle moving direction = 220 mm), and the
5 mm thick steel tubes deform more than the WCPB tubes but less than 3mm thick steel

tubes (maximum displacement in vehicle moving direction = 123 mm).

176

 

Figure 7.32 Vehicle crashing on WCPB triple-tube facade

 

Figure 7.33 Vehicle crashing on concrete triple-tube facade

177

 

Figure 7.34 Vehicle crashing on 3 mm thick steel triple-tube facade

 

Figure 7.35 Vehicle crashing on 5 mm thick steel triple-tube facade

178

 

3 mm thick steel tubes

 

  

Concrete ' " .
-80 _ tubes / '. 5 mm thick steel tubes

 

Acceleration (g)

 

40° i WCPB tubes 11,,
-120 l ‘J

_140.rs

 

 

Figure 7.36 Acceleration time histories of vehicle for four triple-tube facades

Figure 7.36 shows the acceleration time histories at the center of gravity of the
crashing vehicle for the various types of triple-tube facades, the peak acceleration for
concrete triple-tube facade is 120g, which is almost the same as the acceleration of the
vehicle that impacts a plain concrete barrier. Therefore, the concrete tube facade does not
improve the impact characteristics of the modiﬁed concrete barrier. The peak
accelerations for the two steel triple-tube facades are 65g and 68g for the 5 mm and 3 mm
thicknesses, and are lower than the peak acceleration for the WCPB triple-tube facade
(80g). While steel performs better than WCPB, the WCPB performs adequately, is
cheaper, and its use promotes environmental conservation. Figure 7.36 indicates that
both the shape as well as the energy absorption characteristics of the material affect the

performance of the facade.

179

7.5 Summary

Large deformation, dynamic nonlinear ﬁnite element analysis was used to
evaluate the performance of a concrete barrier enhanced with ﬁve different wood-cement
particleboard facades subjected to vehicle impact.

The results of the analysis indicate that it is possible to improve the energy
absorption capability of a concrete crash barrier by adding appropriately conﬁgured
wood-cement particleboard facades. The simulation results indicate that the triple-tube
facade reduces the deceleration forces on the vehicle the most. For a frontal impact of a
Honda Civic vehicle traveling at 30 mph, the triple-tube WCPB facade reduces the
vehicle deceleration during impact to 80g compared to 120g when no facade is used.
Occupants of vehicles sustain serious injuries when the deceleration during impact
exceeds 80g. Therefore, the use of the triple-tube WCPB facade can prevent serious
injury to vehicle occupants during a frontal impact at 30 mph.

Comparisons of the performance of WCPB, concrete and steel triple-tube facades
indicate that although the steel tubes are most effective in reducing the deceleration

forces, the WCPB tubes also perform quite well, while the concrete tubes are ineffective.

180

CHAPTER 8

CONCLUSIONS

8.1 Introduction

The work reported in this dissertation include a new manufacturing process for
making high performance wood-cement particle composites, material characterization of
the composites subjected to bending, compression and impact loading, and results of
dynamic large deformation nonlinear ﬁnite element analysis for simulating impact tests
on wood-cement particleboard and simulating a vehicle impacting a concrete barrier
enhanced with wood-cement particleboard facades. Background information about
wood-cement composites and computational simulation of a vehicle impacting traﬂic
crash barriers also is provided. The most signiﬁcant ﬁndings of this study as well as a
discussion of experimental and analytical future research that will advance this research

further are summarized in this chapter.

8.2 Development of Wood-Cement Particleboard
8.2.1 Manufacturing process

A new process for manufacturing wood-cement particleboard is presented and
consists of pressure control and fully saturated curing to produce high-performance
wood-cement particleboard. In particular, the fully saturated curing procedure improves

the bonding between the irregular surface of wood particles and the cement matrix. As
181

a result, the bending strength of wood-cement particleboard reaches 11.7 MPa and is

higher than the cormnonly accepted limit to 10 MPa (Wolfe, 1996).

8.2.2 Effect of cement/wood ratio to bending strength

Bending tests were performed to establish the relationship between bending
strength and the cement/wood ratio. The particleboard with a cement/wood ratio of 1.5
yields the maximum bending strength (or modulus of rupture) and may be considered as
the optimum cement/wood ratio. The results also indicate that the bending strength of
WCPB increases as the cement/wood ratio increases from 1.0 to 1.5 and decreases when
the cement/wood ratios increases further from 1.5 to 4.0.

The test results also indicate that the modulus of elasticity and density of the
particleboard increases almost linearly as the cement/wood ratio increases.

The toughness index of 6.0 achieved from bending tests indicates that the
wood-cement particleboard has a higher energy dissipation capacity compared to steel
ﬁber reinforced concrete which has a toughness index of 5.0. Therefore, the

wood—cement particleboard can be used when energy dissipation is desired.

8.2.3 Compressive properties
The compressive load-deformation curve of WCPB displays signiﬁcant softening
non-linearity indicating that the material has the capability to absorb energy. The material

properties of WCPB are not isotropic but have directional dependencies due to the
182

orientation of the wood particles resulting from pressing during the manufacturing
process. Short column specimens failed predominantly in shear under compressive
loading irrespective of the orientation of the particles in the columns.

The test results indicated that WCPB with a cement/wood ratio of 1.5 exhibited a
compressive strength comparable to that of normal concrete. However, the strain at peak
load was ten to ﬁfty times larger than that of normal concrete depending on the direction
of the compressive loading. The ability to sustain such large deformations implies that
WCPB can be used for applications where energy dissipation is desired.

A WCPB tube was also tested under compressive loading applied along a diameter.
The test results indicated that the tube had a large plastic deformation. Under a
multi-stress state the wood-cement particleboard can also be modeled as an elasto-plastic

material.

8.2.4 Impact characteristics

A falling ball impact test was used to study the impact characteristics of WCPB.
The falling ball impact test was performed on a plain concrete block, WCPB resting on
the concrete block, and a wood board resting on the concrete block. The time histories
of acceleration, velocity and displacement for impact on WCPB are much closer to the
corresponding time histories for impact on the wood board rather than for impact on plain
concrete. The wood-cement particleboard resting on concrete reduced the maximum

impact deceleration force by about 72% compared to the deceleration force for impact on

183

concrete. The ability of the WCPB to reduce the impact deceleration force by such a

large amount suggests that WCPB has a strong potential to dissipate energy.

8.3 Computational Simulations of Impact on Wood-Cement

Particleboard

Simulation results indicate that the nonlinear ﬁnite element analysis code,
LS-DYNA, can be successfully applied to simulate the mechanical tests on wood-cement
particleboard. The simulation results agree very well with the results of both the impact
test and multi-stress test.

The simulation results also indicate that the LS-DYNA material models could be
effectively applied to model the concrete and wood-cement particleboard. The simulation
results further indicated that the wood-cement particleboard displays elastic-degrading
plastic behavior under a multi-stress state and assuming that the WCPB is an isotropic
elasto-plastic material in ﬁnite element analysis produces reasonable results even though
the real material has directional properties.

Dynamic nonlinear ﬁnite element analysis was applied for evaluating the designs
of a concrete crash barrier enhanced with wood-cement particleboard facades subjected
to vehicle impact. The results of the simulations indicate that it is possible to maximize
the energy absorption ability of the system by selecting an appropriate geometric
conﬁguration for the wood-cement particleboard facade. The concrete barrier enhanced

with triple-tube WCPB facade provides the highest level of crash energy absorption. The
184

geometry of the concrete barrier enhanced with the triple-tube wood-cement
particleboard facade assures controllable deformations and reduces the potential injury to

vehicle occupants.

8.4 Future Research
Future research, development, and applications of wood-cement particleboard are
necessary in order for enhanced concrete crash barriers with WCPB facade to become a
reality. The future works to study include the development of:
0 Better protective coating system (ﬁber-resin coating or resin coating) for WCPB
facade to strengthen its durability under freeze-thaw.
0 Better material model to accurately characterize the material properties of
wood-cement particleboard.
0 Optimization of the geometric conﬁguration of wood-cement particleboard facade
used to enhance concrete crash barriers.
0 Development of connection details for securing the WCPB facade to the concrete
barriers.
Because of the complexity of the behavior of concrete barriers enhanced with the
triple-tube WCPB facade, a ﬁeld crash test of a vehicle impacting an enhanced concrete

crash barrier is also recommended in the future research.

185

APPENDICES

186

APPENDIX A

AN 820C VEHICLE IMPACTING A CONCRETE BARRIER ENHANCED WITH

A 3-TUBE WOOD-CEMENT COMPOSITE PANEL

This appendix presents the input ﬁle of an 820C vehicle impacting a concrete
barrier enhanced with a 3-tube wood-cement particle panel. For simpliﬁcation, the
original 596 pages input ﬁle was simpliﬁed into current input ﬁle. The current input ﬁle
only presents the major command syntax but eliminating most of the data related to
Nodes and Elements. The input ﬁle is in keyword format in LS-DYNA Version 960 and

generated using HyperMesh Version 5.1 and Text Editor as well.

‘KEYWORD
I"TITLE

820C VEH IMPACT SIMULATION WITH VERTICAL CONCRETE BARRIER ENHANCED

WITH WCPB
$$ HM_OUTPUT_DECK created 18:06:59 12-10-2003 by HyperMesh Version 5.1
$5 Generated using HyperMesh-Ls-dyna Template Version : 5.0-l
*CONTROL_TERMINATION
$$ ENDTIM ENDCYC DTMIN ENDENG ENDMAS
100.0 0 0.0 0.0 0.0

‘CONTROL_TIMESTEP
$$ DTINIT TSSFAC lSDO TSLIMT DTZMS LCTM
MSIST

0.00.89999998 0 0.0 -0.00111 0 l
*CONTROL_SHELL
SS WRPANG ITRIST IRNXX ISTUPD THEORY BWC
PROJ

45.0 2 -1 0 2 2 l
*CONTROL_HOURGLASS
$$ [HO OH
5 0.1

*CONTROL_BULK_VISCOSITY
$5 Q2 Q1

1.5 0.06
*CONTROL_CONTACT
$3 SLSFAC RWPNAL ISLCHK SHLTHK PENOPT THKCHG
ENMASS

0.1 0.0 2 O 1 0 1

187

ERODE

MITER

ORIEN

$$ USRSTR USRFRC NSBC S INTERM XPENE SSTHK ECDT
TIEDPRJ

0 0 0 0 0.0 0 0 0
*CONTROL_PARALLEL
$$ NCPU NUMRHS ACCU

4 0 1
*CONTROL_COUPLIN G
$5 UNLENG UNTIME UNFORC TIMIDL FLIPX FLIPY FLIPZ
SUBCYL

l .0 1 .0 0.0 0.0 0 0 0 l

’CONTROL_OUTPUT
$5 NPOPT NEECHO NREFUP IACCOP OPIFS IPNINT IKEDIT

l 3 0 0 0.0 0 100 0
*CONTROL_ENERGY
$$ HGEN RWEN SLNTEN RYLEN

2 2 2 2

$$DATABASE_OPTION -- Control Cards for ASCII output
*DATABASE_NODOUT

1.0
*DATABASE_GLSTAT
1.0
*DATABASE_MATSUM
1.0
*DATABASE_RBDOUT
1.0
*DATABASE_BINARY_D3PLOT
$5 DT/CYCL LCDT BEAM NPLTC
2.0 0 0 0
*DATABASE_BINARY_D3THDT
$$ DT/CYCL LCID
2.0 0
*DATABASE_BINARY_D3DUMP
$8 DT/CYCL
40000.0
*DATABASE_BINARY_RUNRSF
$5 DT/CYCL
10000.0

*NODE
l 67 1 .50823 85485 -767.1 162518609266499999970496

*MAT_ELASTIC

188

$HMNAME MATS 1MAT1_1

11.0120E-O7 200.0 0.33 0.0
*MAT_ELASTIC
$HMNAME MATS 3MAT1_3
38.1280E-O6 200.0 0.33 0.0
‘MAT_ELASTIC
SHMNAME MATS 4MAT1_4
42.2000E-06 80.0 0.33 0.0
*MAT_ELASTIC
$HMNAME MATS 5MAT1_5
52.2000E-O6 80.0 0.33 0.0
’MAT_ELASTIC
$HMNAME MATS 6MAT1_6
68.1140E-O6 200.0 0.33 0.0
*MAT_ELASTIC
$HMNAME MATS 8MAT1_8
81 .897OE-06 200.0 0.33 0.0
*MAT_ELASTIC
$HMNAME MATS 20MAT1_20
207.86OOE-06 200.0 0.33 0.0
*MAT_ELASTIC
SHMNAME MATS 21MAT1__21
217.8600606 200.0 0.33 0.0
‘MAT_ELASTIC
$HMNAME MATS 22MAT1_22
227.8600606 200.0 0.33 0.0
*MAT_ELASTIC
SHMNAME MATS 24MAT1_24
241 .0120E-07 200.0 0.33 0.0
*MAT_ELASTIC
SHMNAME MATS 26MAT1__26
262.7780608 200.0 0.33 0.0
*MAT_PLASTIC_KINEMATIC
SHMNAME MATS 2MAT3_2
27.8600606 200.0 0.33 0.207
0.0 0.0 0.0 0
‘MAT_PLASTIC_KINEMATIC
SHMNAME MATS 7MAT3_7
77.8600606 200.0 0.33 0.207
0.0 0.0 0.0 0

*MAT_PLASTIC_KINEMATIC

189

SHMNAME MATS 9MAT3_9

938000607 10.0 0.35 0.0207 0.1
0.0 0.0 0.0 0
*MAT_PLASTIC_KINEMATIC
$HMNAME MATS 10MAT3_10
105.6050607 200.0 0.33 0.207 2.0
0.0 0.0 0.0 o
*MAT_PLASTIC_KINEMATIC
$HMNAM6 MATS rrMAT3__11
117.8600E-06 200.0 0.33 0.207 2.0
0.0 0.0 0.0 o
*MAT_PLASTIC_KINEMATIC
$HMNAME MATS 12MAT3_12
1278600606 200.0 0.33 0.207 2.0
0.0 0.0 0.0 0
*MAT_PLAST1C_KIN6MATIC
SHMNAME MATS 13MAT3_13
135.7410607 30.0 0.33 0.02 0.03
0.0 0.0 0.0 0
‘MAT_PLASTIC_KINEMATIC
$HMNAME MATS 14MAT3_14
1478600606 200.0 0.33 0.207 2.0
0.0 0.0 0.0 0
*MAT_PLASTIC_KINEMATIC
$HMNAM6 MATS 15MAT3_15
1578600606 200.0 0.33 0.207 2.0
0.0 0.0 0.0 0
*MAT_PLASTIC_K1N6MATIC
SHMNAME MATS 17MAT3_17
1778600606 200.0 0.33 0.207 2.0
0.0 0.0 0.0 0
*MAT_PLAST1C_K1N6MAT1C
SHMNAME MATS 18MAT3_18
1878600606 2000.0 0.33 2.07 2.0
0.0 0.0 0.0 0
*MAT_PLASTIC_KINEMATIC
$HMNAME MATS 19MAT3_19
1978600606 200.0 0.33 0.207 2.0
0.0 0.0 0.0 0
*MAT_PLASTIC_KINEMATIC
$HMNAME MATS 23MAT3_23

190

237.8600E-06 200.0 0.33 0.207 2.0 0.0

0.0 0.0 0.0 0
*MAT_PLASTIC_KINEMATIC
$HMNAME MATS 25MAT3_25
251.5720605 200.0 0.33 0.207 2.0 0.0
0.0 0.0 0.0 0
*MAT_NULL
SHMNAME MATS lOZNULL
1024.0000607 200.0
*MAT_RIGID
$HMNAME MATS 16MAT20_16
161.4650E-06 200.0 0.33 0.0 0.0 0.0
1.0 0 0
0.0 0.0 0.0 0.0 0.0 0.0

*MAT_PIECEWISE_LINEAR_PLASTICITY
$HMNAME MATS 108board

1081 .0400E-06 0.37 0.3 0.017
0 0
$$ HM Entries in Stress-Strain Curve = 8
0.0034 0.0068 0.0096 0.0123 0.0146 0.0178 0.0188
0.0 0.01 0.02 0.033 0.0475 0.0808 0.1168

*MAT_RESULTAN T_PLASTIC1TY
SHMNAME MATS 101MAT_BAR

1017.8340E-06 200.0 0.33 0.207 0.0

I"PART
SHMNAME COMPS lPartl
SHMCOLOR COMPS 1 1
material type # 1 (elastic)

l 1 l 0 0 0
*PART
SHMNAME COMPS 2Part2
$HMCOLOR COMPS 2 2
material type # 3 (Kinematic/Isotropic elastic-plastic)

2 2 2 0 0 0
1"PART
SHMNAME COMPS 3Part3
SHMCOLOR COMPS 3 3
material type # l (elastic)

3 3 3 O 0 0
*PART
SHMNAME COMPS 4Part4

191

0.0142
0.151

0.3

SHMCOLOR COMPS 4 4
material type # 1 (elastic)

4 4 4 0
‘PART
$HMNAME COMPS 5Part5
SHMCOLOR COMPS 5 5
material type # l (elastic)

5 5 5 0
‘PART
SHMNAME COMPS 6Part6
SHMCOLOR COMPS 6 6
material type # 1 (elastic)

6 6 6 0
*PART
SHMNAME COMPS 7Part7
$HMCOLOR COMPS 7 7
material type # 3 (Kinematic/Isotropic elastic-plastic)

7 7 7 0
‘PART
SHMNAME COMPS 8Part8
$HMCOLOR COMPS 8 8
material type # l (elastic)

8 8 8 0
"PART
$HMNAME COMPS 9Part9
$HMCOLOR COMPS 9 9
material type # 3 (Kinematic/Isotropic elastic-plastic)

9 9 9 0
‘PART
SHMNAME COMPS lOParth
SHMCOLOR COMPS 10 10
material type # 3 (Kinematic/Isotropic elastic-plastic)

10 10 10 0

*PART
SHMNAME COMPS llPartll
SHMCOLOR COMPS ll 11

material type # 3 (Kinematic/Isotropic elastic-plastic)
ll 11 ll 0

’PART
$HMNAME COMPS 12Part12
SHMCOLOR COMPS 12 12

192

material type # 3 (Kinematic/Isotropic elastic-plastic)

12 12 12 0

*PART

SHMNAME COMPS 13Part13

$HMCOLOR COMPS 13 13

material type # 3 (Kinematic/Isotropic elastic-plastic)
13 13 13 0

*PART

$HMNAME COMPS l4Part14

$HMCOLOR COMPS l4 l4

material type # 3 (Kinematic/Isotropic elastic-plastic)
l4 14 14 0

*PART

$HMNAME COMPS lSPartlS

SHMCOLOR COMPS 15 15

material type # 3 (Kinematic/Isotropic elastic-plastic)
15 15 15 0

*PART

SHMNAME COMPS 16Partl6

$HMCOLOR COMPS 16 1

material type # 20 (Rigid)
l6 l6 l6 0

*PART

$HMNAME COMPS 17Part17

SHMCOLOR COMPS 17 1

material type # 3 (Kinematic/Isotropic elastic-plastic)
17 l7 17 0

*PART

$HMNAME COMPS 18Part18

SHMCOLOR COMPS l8 2

material type # 3 (Kinematic/Isotropic elastic-plastic)
18 18 18 0

*PART

$HMNAME COMPS 19Partl9

$HMCOLOR COMPS 19 3

material type # 3 (Kinematic/Isotropic elastic-plastic)
19 19 19 0

*PART

SHMNAME COMPS 20Part20

SHMCOLOR COMPS 20 4

material type # l (elastic)

193

20 20 20 0

‘PART
$HMNAME COMPS 21Part21
SHMCOLOR COMPS 21 5
material type # l (elastic)

21 21 21 0
*PART
SHMNAME COMPS 22Part22
SHMCOLOR COMPS 22 6
material type # l (elastic)

22 22 22 0
I"PART
$l-IMNAME COMPS 23Part23
SHMCOLOR COMPS 23 7
material type # 3 (Kinematic/Isotropic elastic-plastic)

23 23 23 0
‘PART
$HMNAME COMPS 24Part24
SHMCOLOR COMPS 24 8
material type # l (elastic)

24 24 24 0
*PART
$HMNAME COMPS 25Part25
SHMCOLOR COMPS 25 9
material type # 3 (Kinematic/Isotropic elastic-plastic)

25 25 25 0
‘PART
$HMNAME COMPS 26Part26
$HMCOLOR COMPS 26 10
material type # l (rigid)

26 26 16 0
‘PART
$HMNAME COMPS lOlBARl
$HMCOLOR COMPS 101 15
barl

101 101 101
‘PART
SHMNAME COMPS lOZBARZ
SHMCOLOR COMPS 102 15
bar2
102 101 101

194

*PART
$HMNAME COMPS
$HMCOLOR COMPS
bar3

103 101
‘PART
$HMNAME COMPS
SHMCOLOR COMPS
bar4

104 101
*PART
$HMNAME COMPS
SHMCOLOR COMPS
Barrier end surface

105 102
*PART
SHMNAME COMPS
SHMCOLOR COMPS
Barrier side surface

106 102
*PART
SHMNAME COMPS
$HMCOLOR COMPS
Concrete barrier

107 103
*PART
$HMNAME COMPS
$HMCOLOR COMPS
board

108 103
*PART
SHMNAME COMPS
$HMCOLOR COMPS

1 10 l 02
*PART
SHMNAME COMPS
$HMCOLOR COMPS

112 102
*SECTION_BEAM

103BAR3
103 15

101

1043AR4
104 15

101

IOSSURF_CROSS

105 1
102
106SIDESURF
106 9
102

107CONCRETE
107 l
103
108board
108 14
108
l lOcontath
1 10 4
102
1 12contact1
1 12 5
102

195

$HMNAME PROPS 18SectBeaml 8

18 l 0.0 2.0
75.0 75.0 71.0 71.0

*SECTION_SHELL

$HMNAME PROPS ZSectSh112
2 2 0.0 3
l .0 1.0 1.0 1 .0

*SECTION_SOLID
$HMNAME PROPS lSectSldl
l l 4

*INITIAL_VELOCITY

$HMNAME LOADCOLS lVelocity
SHMCOLOR LOADCOLS l 1
7 0
13.4 0.0 0.0 0.0

$8 Base Accelerations and Angular Velocities
*LOAD_BODY_Z

SHMNAME LOADCOLS 3LoadBody_3
$HMCOLOR LOADCOLS 3 1
1 0.0 0
*ELEMENT_BEAM
43317 18 1 8 7
43318 18 8 9 7

’ELEMENT_SHELL_THICKNESS
9 2 108 1 1 1 123
1.54 l .54

*SET_PART_LIST

$HMSET
SHMNAME SETS lset_part
l
l 2 3 4
9 10 l l 12
17 18 19 20
25
*SET_PART_LIST
$HMSET

196

O.

1.0

0.0
0

121

1.54

13
21

0.0
0.0

0.0

0.0

1.54

14
22

0
0 0
7 8
15 16
23 24

$HMNAME SETS

28et_part.l

11
19

6Set_6

4304
4371
4450
4493
4572
4642
4682
4743

12
20

4330
4373
4452
4510
4574
4644
4699
4745

6setSegment_6

6 6

l 3car_tube

13 l

2
1 2
9 10
17 18
25
*SET_NODE_L1ST
$HMS6T
$HMNAM6 SETS
6
4300 4302
4355 4369
4422 4424
4489 4491
4544 4570
4623 4625
4678 4680
4724 4741
*SET_SEGMENT
$HMNAM6 CSURFS
SHMCOLOR CSURFS
6
‘CONTACT_NODES_TO_SURFACE
rurwntA866<3RLNJPS
SHMCOLOR GROUPS
10

13
21

4332
4390
4454
45 12

4646
470 l

*CONTACT_AUTOMATIC_SINGLE_SURFACE_TITLE
$664304146<3anrps

SHMCOLOR GROUPS
5singlesurface

2
0.15 0.
1.0

*RIGIDWALL_PLANAR

031
1.0

$HMNAME GROUPS

$HMCOLOR GROUPS

6

452.93851-567.67354
*BOUNDARY_SPC_NODE

SHMNAME LOADCOLS

5singlesurface
5 2
2
0.0 0.0
6rigid
6 12

0.0

1.0

-l .0 452.93851-567.67354

2NodalConstraints. 1

14
22

43 34
4392
447 l
45 14
4602
4660
4703

1.0

0.0

1.0

7 8

15 16
23 24
435 1 4353
4394 4420
4473 4475
4540 4542
4604 462 1
4662 4664
4720 4722

0.01 .0000E+20

1.0

SHMCOLOR LOADCOLS 2 5
5297 0 1 l 1

*DATABASE_HISTORY_NODE

SHMNAME OUTPUTBLOCKS 1nodeth
2788110 2788108 2787921 2787919

*DEFINE_CURVE

SHMNANIE CURVES lLoadCurvel
$HMCOLOR CURVES l l
$HMCURVE l l LoadCurvel
l 0 1.0 1.0 0.0
0.0 0.0
3.0 0.00981
400.0 0.00981

$5 UNSUPPORTED CARDS - STARTS
*MAT_PSEUDO_TENSOR
103 2.4E-06 8.971 0.2
0.02 -1.0 0.0 0.0 0.0 0.0
0.0 0.0 0.0
$5 UNSUPPORTED CARDS - ENDS
I"END

198

0.0

0.0

0.0

APPENDIX B

A STEEL BALL IMPACTING ON WOOD-CEMENT PARTICLEBOARD
OVERLAYING A CONCRETE BLOCK

This appendix presents the input ﬁle of a steel ball impacting a wood-cement
particleboard overlaying a concrete block. For simpliﬁcation, the original 477 pages input
ﬁle was simpliﬁed into the current demonstration ﬁle. The current ﬁle only presents the
major command syntax but eliminating most of the data related to Nodes and Elements.
The input ﬁle is in keyword format in LS-DYNA Version 960 and generated using
HyperMesh Version 5.1 and Text Editor as well.

*KEYWORD

$$ HM_OUTPUT_DECK created 02:33:56 05-26-2003 by HyperMesh Version 5.1
$$ Generated using HyperMesh—Ls-dyna Template Version : 5. l
*CONTROL_TERMINATION

$8 ENDTIM ENDCYC DTMIN ENDENG ENDMAS

10.0 0 0.0 0.0 0.0
*CONTROL_TIMESTEP
$$ DTINIT TSSFAC ISDO TSLIMT DTZMS LCTM
ERODE MSIST
0.00.89999998 0 0.0 0.0 0 l 0

*CONTROL_SHELL
$8 WRPANG ITRIST IRNXX ISTUPD THEORY BWC MITER PROJ

45.0 2 -1 0 2 2 1 0
*CONTROL_HOURGLASS
$$ IHQ QH
5 O. 1
*CONTROL_BULK_VISCOSITY
$$ Q2 Q1
1 .5 0.06

*CONTROL_CONTACT
$$ SLSFAC RWPNAL ISLCHK SHLTHK PENOPT THKCHG ORIEN
ENMASS

0. 1 0.0 2 0 1 0 1
$5 USRSTR USRFRC NSBCS INTERM XPENE SSTHK ECDT
TIEDPRJ

199

0 0 0 0 0.0
*CONTROL_PARALLEL
$$ NCPU NUMRHS ACCU
4 0 1
*CONTROL_COUPLING
$$ UNLENG UNTIME UNFORC TIMIDL
SUBCYL
1.0 1.0 0.0 0.0 0
*CONTROL_OUTPUT
$$ NPOPT NEECHO NREFUP IACCOP
l 3 0 0 0.0
*CONTROL_ENERGY
$3 HGEN RWEN SLNTEN RYLEN
2 2 2 2
$$DATABASE_OPTION - Control Cards for ASCII output
*DATABASE_NODOUT
0.02
*DATABASE_GLSTAT
0.02
*DATABASE_MATSUM
0.02
‘DATABASE_RBDOUT
0.02
l"DATABASE”BlNARYmD3PLOT
$$ DT/CYCL LCDT BEAM NPLTC
1.0 0 0 0
‘DATABASE_BINARY_D3THDT
$$ DT/CYCL LCID
2.0 0
I"DATABASE_BINARY_D3DUMP
$$ DT/CYCL
40000.0
I"DATABASE_BINARY__RUNRSF
$$ DT/CYCL
10000.0
*NODE

F LIPX

OPIFS

23.2759301877E-15 -53.5 -0.503

*MAT_PLASTIC_KINEMATIC
$11MNAM6 MATS lsphere
17.9000606 210.0 0.28 0.351

200

3.4

FLIPY

IPNINT
100

0.0

FLIPZ

IKEDIT

0

0.0 0.0 0.0

*MAT_NULL
SHNINAME MATS 3Null

38.4000E-07
*MAT_PIECEWISE_LINEAR_PLASTICITY
$HMNAME MATS 2board

21 .0400E-06 0.37 0.3 0.017

1 0

$$ HM Entries in Stress-Strain Curve = 0
*PART
$HMNAME COMPS lsphere
$HMCOLOR COMPS 1 4

l l 1
*PART
$HMNAME COMPS 2board
$HMCOLOR COMPS 2 2

2 2 2
*PART
$HMNAME COMPS 3concrete
$HMCOLOR CONIPS 3 8

3 3 4
‘PART
$HMNAME COMPS 4sphere_face
$HMCOLOR COMPS 4 9

4 4 3
*PART
$HMNAME COMPS 5board_face
$HMCOLOR COMPS 5 7

5 5 3
*SECTION_SHELL
$HMNAME PROPS 4sphere_face

4 2 0.833 2 0.0 0.0

0.1 0.1 0.1 0.1 0.0

$HMNAME PROPS 5board_face

201

200.0

0.3

5 2 0.833 2 0.0 0.0 0

0.1 0.1 0.1 0.1 0.0
*SECTION_SOLID
$HMNAME PROPS lsphere
1 1 4
$HMNAME PROPS 2board
2 l 4
$HMNAME PROPS 3concrete
3 l 4
*ELEMENT_SHELL
l 1586 4 2375 2373 2374 2374
l 1585 4 2222 2262 2263 2263
11584 4 2254 2221 2255 2255
11583 4 2095 2092 2118 2118
*ELEMENT_SHELL
l 1978 4 2288 2306 2295 2278
1 1977 4 2282 2275 2294 2297

*ELEMENT_SOLID
1420 1 1359 165 172 207 207 207 207 207
1476 l 1401 356 315 357 357 357 357 357

*CONTACT_AUTOMATIC_SURFACE_TO_SURFACE

$HMNAME GROUPS lsurftosurf
$HMCOLOR GROUPS 1 l
4 5 3 3 0 0
0.0 0.0 0.0 0.0 0.0 0 0.01.0000E+20
1.0 1.0 1.0 1.0 1.0 1.0
*BOUNDARY_SPC_NODE
$HMNAME LOADCOLS lspc
$HMCOLOR LOADCOLS 1 14
8462 0 l l l 1 1 l
*INITIAL_VELOC1TY_NODE
$HMNAME LOADCOLS 2velocity
$HMCOLOR LOADCOLS 2 10
2699 0.0 0.0 -l .72
2698 0.0 0.0 -1.72

I"DATABASE_HISTORY_NODE
$HMNAME OUTPUTBLOCKS lNOUT
399 558 2295

202

*DEFINE_CURVE

$HMNAME CURVES lcurvel
$HMCOLOR CURVES l l
$HMCURVE l 0 curvel
l 0 1.0 1.0 0.0 0.0
0.0 0.0034

0.0102 0.0067

0.0206 0.0095
$$ UNSUPPORTED CARDS - STARTS
*MAT_PSEUDO_TENSOR
4 2.4E-06 8.971 0.2
0.02 -l .0 0.0 0.0 0.0 0.0 0.0
0.0 0.0 0.0
$$ UNSUPPORTED CARDS - ENDS
*END

203

0.0

BIBLIOGRAPHY

204

BIBLIOGRAPH

Altair Engineering Inc. 1999. HyperMesh User Manual.
Altair Engineering Inc. 1999. HyperGraphics Usser Mannual.
American National Standards, Particleboard, 1993.

American Society for Testing and Materials, ASTM D 1037-78, 1996, Standard method
of evaluating the properties of wood-based ﬁber and particle panel materials.

Balaguru, P. N. and Shah, S. P., 1992, “Fiber-Reinforced Cement Composites”,
McGraw-Hill, Inc.

Bathe, K., 1996, “Finite Element Procedures”, Prentice-Hall, Inc.

Bison-Werke Bahre and Greten GMBH & Co. KG, 1977, “Cement-bonded particleboard
plant integrated with low-cost housing production unit”.

Blankenhor, RR. 1994, “Compressive strength of hardwood-cement composites”, Forest
Products Journal, Vol. 44, No. 4, pp. 59-62.

Bligh, Roger, 2002. “Project Shows predictive use of computer simulation.” Texas
Transportation Researcher.

Bodig, J ozsef and Jayne, B. A. 1982, “Mechanics of Wood and Wood Composites”,
Van Nostrand Reinhold Company Inc. New York, NY.

Bresler, B. and Pister, KS, 1958. “Strength of Concrete under Combined Stresses.”
Journal of the American Concrete Institute, 55(20), 321-345.

CETRIS, 2005, Web publication at http: //www.cetris.com.

Chen, W. F. and Saleeb, A. F ., 1982, “Constitutive Equations for Engineering Materials:
Volume 1, Elasticity and Modeling”, John Wiley & Son, Inc.

Chen, W. F ., and Han, D. J ., “Part IV: Concrete Plasticity.” Plasticity for Structural
Engineers. New York: Springer-Verlag New York Inc.

Cook, R. D., Malkus, D. S. and Plesha, M. E., 1989. “Concept and Applications of Finite
Element Analysis”. Third edition, John Wiley & Sons, Inc.

205

Cooper, RA. 1994. “Disposal of treated wood removed from service: The issues in
environmental considerations in manufacture, use, and disposal of preservative-
treated wood,” Forest Products Society(Madison, Wisconsin), pp. 85-90.

Coutts, R.S.P.,l986 “Sticks and Stones”, Forest Products Newsletter, CSIRO Division of

Chemical and Wood Technology (Australia), Vol.2, No.1, January, pp. 1-4
D103 7-78.

Desai, C. S. and Siriwardane, H. J ., 1984, “Constitutive Laws for Engineering Materials
with Emphasis on Geologic Materials”, Prentice Hall, Inc. New Jersey.

Dinwoodies, J .H. et a1, 1991 “The long-term performance of cement bonded
particleboard” Proceedings of inorganic bonded wood and ﬁber composite

materials. Forest Product Society, Madison, Wisconsin.

Drucker, D. C. and W. Prager, 1952, Soil Mechanics and Plastic Analysis of Limit
Design, Quart. Applied Mathematics, Vol.10, No.2, pp. 157-165.

Du Bois, Paul A., 1999, “Crashworthiness Engineering Course Notes.”
Eltomation Web Publication, http: //www. e1tomation.nl.

Environmental Building News, 1997, “Disposal: The Achilles’ Heel of CCA-Treated
Wood”, Vol. 6, No.3, March.

Environmental Building News, 1998, “Wood-Chip And Cement Wall Form”, Vol. 7, No.
3, March.

Environmental Building News. 1992. “Wood-ﬁber Cement Blocks”, Vol. 1, No. 3.

Fact Sheet, 2002, “Wood: A strategy for Developing Recycling Market in Ireland”,
Environmental Protection Agency, Ireland.

F alk, R. H., 1994, “Housing Products From Recycled Wood”, Proceedings of the First
International Conference of CIB TG 16, Nov. 6-9, Tampa, Florida, USA.

Federal Motor Vehicle Safety Standards (FMV SS), 2000, No. 214. USA.

Ferguson, P. M.; Breen, J. E. and Jirsa, J. O., 1988, “Reinforced Concrete
Fundamentals”, Fifth edition, John Wiley & Son, Inc.

Fuwape, J oeseph Adeola, 1995,“The effect of wood/cement ratio on the strength
properties of cement-bonded particleboard from spruce”, Journal of Tropical
Forest Products.

206

Gertjansen, R.O., et a1. 1988. “Assessment of preservative-treated aspen waferboard
after 5 years of ﬁeld exposure. Forest Products J oumal, 39(4).

Hallguist, John O., 1999, LS-DYNA Theoretical Manual, Livermore Software
Technology Corporation.

Hallguist, John O., 1999, LS-DYNA User Manual, Livermore Software Technology
Corporation.

Hallquist, J. O. and Benson, D. J ., 1989, “DYNA3D — Explicit Finite Element Program
for Impact Calculations”, Proceedings of symposium of Crashworthiness and
Occupant Protection in Transportation Systems, ASME Annual Meeting, AMD
Vol. 106.

Haygreen, J .G. and Bowyer J .L., 1989, Forest Products and Wood Science, an
Introduction, Second Edition, Iowa State University Press /AMES.

Health Canada, 2003, “Fact Sheet on Chromated Copper Arsenate (CCA) Treated Wood”.

Hsieh, S.S., E.C. Ting and W.F. Chen. “A plasticity-Fracture Model for Concrete.”
International Journal of Solids and Structures, 18(3), 181-197.

Hsu, W.E. 1994, “Cement-bonded particleboard from recycled CCA-treated and virgin
wood”, Proceedings of Inorganic-Bonded Wood and Fiber Composites Materials,
University of Idaho, Moscow, Idaho, 4: 49-56. John Wiley & Sons, Inc.
http://www.nbc10.com/money/2 l97698/detai1.html.

HyperMesh. 2000. Finite Element Analysis Software, Altair Engineering Inc. Version
5.1.

HyperGraphics. 2000. Data Processing Software, Altair Engineering Inc. Version 5.1.

Information about automobile accidents, 2003.
http://www.physics.ubc.ca/~outreach/phys420/p420 96/danny.

Johnson, J. P., and Skynar, M., 1989, “Automotive Crash Analysis Using the Explicit
Integration Finite Element Method”, Proceedings of symposium of
Crashworthiness and Occupant Protection in transportation Systems, ASME
Annual Meeting, AMD Vol. 106.

Karam, G.N. et a1, 1992, “Evaluation of commercial wood-cement composites for
sandwich-panel facing,” J. of Materials in Commercial Engineering 6(1).

Kupfer, H., Hilsdorf, H. K. and Rusch, H.,l969. “Behavior of concrete under biaxial
stresses.” ACI Journal 66(8).

207

Kurﬁel, Fred and Moslemi, A.A., 1995, “Wood panel makers consider cement-ﬁber to
boost proﬁts”, Sept. Wood Technology.

Laws, V., 1983, “On the mixture rule for strength of ﬁber reinforced cements,” Journal
of Material Science Letters, 2, , pp. 527-531.

Lee, W.C., 1985, “Effect of cement/wood ratio on bending properties of cement-bonded
southern pine excelsior board,” Wood and Fiber Science, 17(3), pp. 361-364.

Little, R. W., 1999, “Elasticity.” Dover Publications, Inc., Mineola, New York.
Livermore Software Technology Corporation (LSTC). 2002. LS-POST User manual.

LS-DYNA Version: 960. 2000. Finite element analysis software. Livermore Software
Technology Corporation.

LS-POST. 2000. Data processing software. Livermore Software Technology Corporation.

Mackerle, J ., 2003. “Finite Element Crash Simulations and Impact-Induced Injuries: An
addendum. A Bibliography (1998-2002).” The Shock and Vibration Digest, July.

Marer, P.J. and Grimes, M. 1992. Wood Preservation. University of California

Statewide integrated pest management project. Division of Agriculture and
Natural Res. Pub. 3335.

Marra, G. G. 1972, “The future of engineered wood materials.” Quest 9(4): 18

McDevitt, C. F., 2003, “Basics of Concrete barriers.”
http://www.tflrtc.gov/pubrds/marprOO/concrete.htm.

Mehta, P. Krunar and Monteiro, Paula J. M., 1986, “Concrete: structure, properties and
materials”, second edition, Prentice Hall, Inc. New Jersey.

Micklewright, J. T. 1993. Wood Preserveation Statistics, 1991. American Preserver’s
Association. Woodstock, MD.

Miller, DR, 1987, “Wood-cement composites: Interactions of wood components with
Portland cement”, Doctoral Dissertation, University of Idaho.

Milligan, Ann, 2004, “Cement and wood-wool combine to improve Philippine housing”,
Partners in Research for Development, No. 14. Australia.

Moslemi, A.A. and Hamel, MP. 1989, “Fiber and particleboards bonded with inorganic
binders”, Forest Products Research Society, Madison, WI.

208

Moslemi, A.A. and Pﬁster, SC. 1987, “The inﬂuence of cement/wood ratio and cement
type on bending strength and dimensional stability of wood-cement composite
panel”, Wood and Fiber Science, 19(2), pp. 165—175.

Moslemi, A.A., 1974, Particleboard Volume 2. Southern Illinois University Press,
Carbondale IL.

NBC newsletter, 2003, “CCA Pressure-Treated Wood”,

NCHRP Report 350: Recommended Procedures for safety Performance Evaluation of
Highway Features. TRB, National Research Council, Washington. DC. 1993.

Neithalath, N., Weiss, J. and Olek, J., 2003. “Acoustic Performance and Damping
Behavior of Cellulose-Cement Composites.” ﬁom 1284 School of Civil
Engineering, Purdue University.

Nilson, H., 1991, “Design of Concrete Structures,” 11th edition, McGraw-Hill Inc.
RBAC Newsletter, 2002. “Recycling Woods: Value-Added Opportunities for Recycled
Wood”. Recycling Business Assistance Center, A cooperative effort of the N. C.

Department of Environment and Natural Resources and the N. C. Department of
Commerce. Summer.

Nilsson, Larsgunnar, et a1, 1998, “Simulation of truck and car impacting a roadside
barrier,” Proceedings of 5th LS-DYNA International Conference.

Ottosen, NS. 1977. “A failure criterion of concrete.” ASCE Journal of Engineering
Mechanics. 103 (EM4), pp. 527-535.

Oyagade, AD. 1990, “Effect of cement/wood ratio on the relationship between cement
bonded particleboard density and bending properties”, Journal of Tropical Forest
Science 2(2), pp. 211-219.

Personal email communication with Dr. Len Shewer, 2002, “Algorithm of generating
concrete bulk modulus in LS-DYNA.”

Plaxcico, C.A., 1997, “Simulation of a vehicle impacting a modiﬁed thrie-beam
guardrail”. Transportation Research Record 1599.

Prestemon, DR. 1976, “Preliminary evaluation of a wood-cement composite.” Forest
Products Journal, Vol. 26, No.2.

Reddy, J. N., 1993. “An Introduction of the ﬁnite element method.” Second edition,
McGraw-Hill, Inc.

Roadside Hazards, 2001 , Insurance Institute for Highway Safety.

209

Sarja, A., 1988, “Wood ﬁber reinforced concrete”, Natural Fiber Reinforced Cement and
Concrete, Edited by R.N. Swamy, blackie and Sons Publication, pp. 63-91.

Schmidt, R. 1994, “Increased wood-cement compatibility of chromate-treated wood.”
Forest Products Journal, 44(7/8): 44-46.

Schwer, Len, 2001, “Laboratory Tests for Characterizing Geo-materials”, Lecture notes
for Geomaterials subjected to impact loads.

Schwer, Len, 2002, Personal email communication.

Shojaati, M., 2003. “Correlation between injury risk and impact severity index ASI.”
Swiss Transportation Research Conference.

Simatupang, M.H., 1990. “Properties and durability of rapid curing cement-bonded
particleboard, Manufactured by addition of carbonation,” Proceeding 2“d
International Symposium on Vegetable Plants and Their Fibers as Building
Materials, Sept. Salvator, Bahia, Brazil. pp. 239-247.

Sorfa, P. 1984, “Properties of wood-cement composites.” Journal of Applied Polymer
Science, Applied Polymer Sym. 40, pp. 209-216.

Stalker, l. N. 1993. Disposal of treated wood after service. Proceeding CWPA: 159-174.

Suzuki, K. 1993. “Field evaluation of composite products, six years in service. In:
Protection of wood-based composite products.” Forest Products Society (Madison,
Wisconsin).

Tasuji, M.E., Slate, F.O., Nilsion, A.H., 1978, “Stress-strain response and fracture of
concrete in biaxial loading,” AC1 Journal, Vol. 75, 306-312.

US. Consumer Products Safety Commission (COSC), 2002, Qestions and Answers about
CCA-Treated Wood, http://www.cpsc.go_\d.

Venta, George J ., 1997, “Gypsum fiberboard: a high performance specialty board.”
Proceedings of inorganic bonded wood and ﬁber composite materials. Forest
Product Society.

Vesenjak, M and Ren, Z., 2003. “Improving the Road Safety with Computational
Simulations”. 4th European LS-DYNA Users Conference.

Vick, C.B., R. C. DeGreoot and J. Yongquist. 1990. “Compatibility of nonacidic

waterbone preservatives with phenol-formaldehyde adhesive. Forest Products
Journal, 40 (2).

210

Willarn, K.J., and ER Wamke. 1975. “Constitutive Model for the triaxial Behavior of
Concrete.” IABSE19. Proceedings of the IABSE Seminar on Concrete Structure
Subjected to Triaxial Stresses, Bergamo, Italy, May, IABSE.

Wolfe, Ronald W., 1997, “The use of recycled wood and paper in building applications,”
Proceedings No.7286, Forest Products Society.

Wolfe, Ronald W., 1999, “Durability and strength of cement bonded wood particle
composites made from construction waste, " Forest Products Journal, 49(2):24-31.

Won, J ong—Pil, 1995. “Accelerated processing of high performance wood cement
composites”, Doctoral dissertation, Michigan State University

Yonten, Karma, et a1, 2002, “An evaluation of constitutive models of concrete in LS-
DYNA ﬁnite element code,” 15th ASCE Mechanics Conference.

Zaouk, Abdullatif K, et a1, 1995, “Evaluation of a multipurpose pickup truck model using
full-scale crash data with application to highway barrier impacts,”
FHWA/NHTSA National Crash Analysis Center Report.

Zhou, Y. and Karndem, DP. 2001. “Effect of cement/wood ratio on the properties of

cement-bonded particleboard using CCA-treated wood removed from service.”
Forest Products J ourna.

211

IIIIIIIIIIIIIIIIIIIIIIIIIIII

111311111111111111111111111111111111111