the objective of this research was to explore the

AN ABSTRACT OF THE DISSERTATION OF

Jun Hee Kim for the degree of Doctor of Philosophy in Civil Engineeringpresented on December 22, 2003.Title: Performance-Based Seismic Design of Light-Frame Shearwalls

Abstract approved:

David V. Rosowsky

Performance-based design has gained interest in recent years among

structural designers and researchers. Performance-based design includes selection

of appropriate building sites, structural systems and configurations, as well as

analytical procedures used in the design process, to confirm that the structure has

adequate strength, stiffness and energy dissipation capacity to respond to the

design loads without exceeding permissible damage states. Although performance-

based seismic design has advanced for some materials and structural types, such as

steel and reinforced concrete buildings and bridges, its application to light-frame

structures remains largely unexplored.

The objective of this research was to explore the potential for the

application of performance-based engineering concepts to the design and

assessment of woodframe structures subject to earthquakes. Nonlinear dynamic

time-history analysis was used to predict the performance of shearwalls

considering a suite of scaled characteristic ordinary ground motions to represent

the seismic hazard. Sensitivity studies were performed to investigate the relative

Redacted for privacy

effects of damping, sheathing properties, fastener type and spacing, panel layout,

and other properties on the performance of wood shearwalls. In addition, the

effects of uncertainty in ground motions and variability in sheathing-to-framing

connection hysteretic parameters were investigated. Issues such as the contribution

of nonstructural finish materials, different seismic hazard regions, and construction

quality also were investigated and modification factors to adjust peak displacement

distributions were developed. The peak displacement distributions were then used

to construct performance curves and design charts as a function of seismic weights

for two baseline walls. Finally, fragility curves were developed for the baseline

walls considering different nailing schedules, corresponding allowable seismic

weights, and various overstrength (R) factors.

©Copyright by Jun Hee KimDecember 22, 2003All Rights Reserved

Performance-Based Seismic Design of Light-Frame Shearwalls

byJun Hee Kim

A DISSERTATION

submitted to

Oregon State University

In partial fulfillment ofthe requirements for the

degree of

Doctor of Philosophy

Presented December 22, 2003Commencement June 2004

Doctor of Philosophy dissertation of Jun Hee Kimpresented on December 22, 2003.

APPROVED:

Major Professor, Civil Engineering

Head of Department of Civil, Construction and Environmental Engineering

Dean of thet-Graduate School

I understand that my dissertation will become part of the permanent collection ofOregon State University libraries. My signature below authorizes release of mydissertation to any reader upon request.

Jun Hee Kim, Authr





ACKNOWLEDGEMENTS

The research presented here was made possible largely through grants from

the CUREE-Caltech Woodframe Project and the National Science Foundation

through Grant No. CMS-0049038. This financial support from these organizations

is acknowledged.

I would like to express my sincere appreciation to Dr. David Rosowsky for

his advice, guidance, passion, patience, encouragement, and financial support

throughout my graduate work at Oregon State University. I also would like to

thank my graduate committee members: Dr. Solomon Yim, Dr. Robert Leichti, Dr.

Thomas Miller, and Dr. Bartelt Eleveld for their advice and help in completing my

graduate program. Special thanks are due to Dr. Brian Folz for his valuable

assistance with the computer programs CASHEW and SAWS.

Many thanks to the Structural Reliability Research Group members and the

staff in Civil Engineering for their assistance during the course of this research.

I would like to thank my sister, Mun Hee Kim, and brother-in-law, Dr.

Daniel Kim, for their love and support. Also, I would like to thank my father,

Dong Chan Kim, who passed away two months ago, and my mother, Jung Sook

Mm, for their love, encouragement, support, and prayer. And finally, I wish to

thank my wife, Mi Soon, and two lovely children, Gyu Yeun and Gyu Tae, for

their love, patience, prayer and support.

TABLE OF CONTENTS

1. INTRODUCTION ...................................................................................................... 1

1.1 General ................................................................................................................. 1

1.2 Scope and objectives ............................................................................................ 4

2. BACKGROUND AND LITERATURE REVIEW.................................................... 6

3. NONLINEAR DYNAMIC TIME-HISTORY ANALYSIS .................................... 10

3.1 Computer programs ............................................................................................ 10

3.1.1 CASHEW.................................................................................................... 10

3.1.2SASH1 ........................................................................................................ 11

3.1.3SASHFIT .................................................................................................... 12

3.1.4 SAWS .......................................................................................................... 14

3.2 Ordinary ground motion records ........................................................................ 16

3.3 Distribution functions (exceedence probability curves) .................................... 17

4. ANALYSIS OF ISOLATED SHEARWALLS ....................................................... 22

4.1 Model configuration (isolated shearwall) .......................................................... 22

4.2 Sensitivity studies .............................................................................................. 26

4.2.1 Baseline sensitivity studies ......................................................................... 264.2.1.1 Ground motions.................................................................................... 274.2.1.2 Damping ............................................................................................... 284.2.1.3 Shear modulus of sheathing materials.................................................. 31

4.2.1.4 Fastener spacing ................................................................................... 324.2.1.5 Panel layout .......................................................................................... 354.2.1.6 Shake-table test walls ........................................................................... 354.2.1.7 Missing fasteners.................................................................................. 384.2.1.8 Model uncertainty ................................................................................. 40

4.2.2 Sheathing-to-framing connection hysteretic parameter variability ............. 474.2.3 Contribution of nonstructural finish materials ............................................ 55

4.2.3.1 Analysis of solidwall ........................................................................... 574.2.3.2 Analysis of walls with openings .......................................................... 62

4.2.4 Construction quality .................................................................................... 654.2.5 Effects of different seismic hazard regions ................................................. 73

TABLE OF CONTENTS (Continued)

4.3 Additional studies . 84

4.3.1 Development of modification factors ......................................................... 844.3.1.1 Sheathing-to-framing connection hysteretic parameter variability ...... 844.3.1.2 Construction quality ............................................................................. 894.3.1.3 Contribution of nonstructural finish materials ..................................... 99

4.3.2 Construction of performance curves and design charts ............................ 1074.3.2.1 Baseline walls .................................................................................... 107

4.3.2.1.1 Construction of performance curves ........................................... 108

4.3.2.1.2 Design charts ............................................................................... 111

4.3.2.2 Construction quality ........................................................................... 121

4.3.2.2.1 Construction of performance curves ........................................... 121

4.3.2.2.2 Design charts ............................................................................... 1224.3.2.3 Effects of different seismic hazard regions ........................................ 130

4.4 Performance-based design ............................................................................... 136

4.4.1 Incremental dynamic analysis ................................................................... 1364.4.2 Fragility curves ......................................................................................... 143

4.4.2.1 Fragility curve based on peak displacement ...................................... 143

4.4.2.2 Fragility curve based on ultimate force .............................................. 154

5. ANALYSIS OF SHEARWALLS IN COMPLETE STRUCTURES .................... 161

5.1 Model configuration......................................................................................... 161

5.1.1 Model configuration of one-story residential structure ............................. 1625.1.2 Model configuration of two-story residential structure ............................ 165

5.2 Shearwall performance in complete structures ................................................ 167

5.2.1 One-story structure .................................................................................... 1675.2.1.1 Performance of shearwalls with OSB only ........................................ 1675.2.1.2 Performance of shearwalls with NSF materials ................................. 170

5.2.2 Two-story structure ................................................................................... 1775.2.2.1 Performance of shearwalls with OSB only ........................................ 1775.2.2.2 Performance of shearwalls with NSF materials ................................. 182

5.2.3 Additional studies ..................................................................................... 1875.2.3.1 Interstory displacement ...................................................................... 187

5.2.3.2 Effect of partition walls ...................................................................... 192

TABLE OF CONTENTS (Continued)

gç

5.2.3.3 Performance comparison for isolated wall and wall in one-storystructure........................................................................................... 197

5.3 Performance-based design ............................................................................... 200

5.3.1 Incremental dynamic analysis ................................................................... 2005.3.2 Fragility curves ......................................................................................... 203

5.3.2.1 Fragility curve for one-story structure ............................................... 2035.3.2.2 Fragility curve for two-story structure ............................................... 206

6. CONCLUSIONS AND RECOMMENDATIONS ................................................ 214

6.1 Conclusions ...................................................................................................... 215

6.2 Recommendations ............................................................................................ 218

REFERENCES ........................................................................................................... 221

APPENDICES........................................................................................................... 230

LIST OF FIGURES

Figure

3.1 CASHEW modeling procedure.............................................................................. 11

3.2 Force-displacement response of a wood shearwall under cyclic loading.Hysteretic model is fit to test data for an 8 ft x 8 ft shearwall with 3/8-in. thickOSB sheathing panels (from: Durham, 1998) ...................................................... 13

3.3 Load-displacement curve using parameters determined by SASHFIT .................. 14

3.4 Code based target response spectra ........................................................................ 1 S

3.5 Development of probability-based design charts for shearwall selection .............. 21

3.6 Fitting a lognormal distribution to the sample CDF of peak displacements .......... 21

4.1 Components of typical woodframe shearwall ........................................................ 23

4.2 Detailed configurations of baseline solid wall (BW1) and walls withopenings(OWl and 0W2) .................................................................................. 25

4.3 Baseline wall sheathing configuration ................................................................... 27

4.4 Response (peak displacement) variability for the three limit states ....................... 29

4.5 Effects of viscous damping ratio () on peak displacement .................................. 31

4.6 Effect of assigned shear modulus (G) on peak displacement ................................ 33

4.7 Effect of fastener spacing on peak displacement (W 560 lbs/ft) ........................ 33

4.8 Effect of fastener spacing on peak displacement (W 840 lbs/ft) ........................ 34

4.9 Effect of fastener spacing on peak displacement (W = 1120 lbs/ft) ...................... 34

4.10 Effect of fastener spacing on peak displacement (W = 1400 lbs/ft) .................... 35

4.11 Effect of panel layout on peak displacement ....................................................... 36

4.12 Task 1.1.1 and task 1.1.2 walls ............................................................................ 37

LIST OF FIGURES (Continued)

Figure

4.13 Peak displacement distributions for task 1.1.1 and task 1.1.2 walls .................... 37

4.14 Effect of missing fasteners on peak displacement (10, 5 0/50) ............................ 39

4.15 Effect of missing fasteners on peak displacement (LS, 10/50) ............................ 39

4.16 Effect of missing fasteners on peak displacement (CP, 2/50) .............................. 40

4.17 Effect of model uncertainty on peak displacement distribution ........................... 43

4.18 Effect of model uncertainty on peak displacement (3"/6") .................................. 44

4.19 Effect of model uncertainty on peak displacement (3"/12") ................................ 44


4.21 Effect of model uncertainty on peak displacement (6"/6") .................................. 45


4.23 Comparison of peak displacement distributions for different nail parameters(W=5601bs/ft) .................................................................................................... 50

4.24 Comparison of peak displacement distributions for different nail parameters(W=8401bs/ft) .................................................................................................... 51

4.25 Comparison of peak displacement distributions for different nail parameters(W=ll2Olbs/ft) .................................................................................................. 51

4.26 Comparison of peak displacement distributions for different nail parameters(W= 1400 lbs/ft) .................................................................................................. 52

4.27 Effect of fastener parameter variability on peak displacement ............................ 54

4.28 Typical exterior wall cross-section ...................................................................... 56

4.29 Effect of nonstructural finish materials on peak displacement(W 560 lbs/ft) ..................................................................................................... 59


Figure

4.30 Effect of nonstructural finish materials on peak displacement(W= 840 lbs/fl) ..................................................................................................... 59

4.31 Effect of nonstructural finish materials on peak displacement(W= 1120 lbs/ft) ................................................................................................... 60

4.32 Effect of nonstructural finish materials on peak displacement(W 1400 lbs/ft) ................................................................................................... 60

4.33 Effect of nonstructural finish materials on peak displacement(W= 840 lbs/ft) ..................................................................................................... 61

4.34 Effect of nonstructural finish materials on peak displacement(W= 1400 lbs/fl) ................................................................................................... 61

4.35 Effect of nonstructural finish materials on peak displacement(W=281 lbs/fl) ..................................................................................................... 63



4.38 Effect of nonstructural finish materials on peak displacement(W 984 lbs/fl) ..................................................................................................... 64

4.39 Peak displacement distributions for construction qualities(BW1, OSB only) ................................................................................................. 68

4.40 Peak displacement distributions for construction qualities(BW1, OSB ± GWB) ........................................................................................... 69

4.41 Peak displacement distributions for construction qualities(BW1, OSB + Stucco) .......................................................................................... 69

4.42 Peak displacement distributions for BW1 considering differentconstructionqualities ........................................................................................... 70


Figure

4.43 Peak displacement distributions for OWl (OSB only) consideringdifferent construction qualities ............................................................................. 70

4.44 Peak displacement distributions for OWl (OSB + GWB) consideringdifferent construction qualities ............................................................................. 71

4.45 Peak displacement distributions for OWl (OSB + Stucco) consideringdifferent construction qualities ............................................................................. 71

4.46 Peak displacement distributions for OWl (OSB + GWB + Stucco)considering different construction qualities ......................................................... 72

4.47 Target response spectra for different seismic hazard regions .............................. 75

4.48 Comparison of earthquake record scaling to target response spectra .................. 76

4.49 Comparison of peak displacement between CCWP and SACearthquake records ............................................................................................... 80

4.50 Comparison of peak displacement between fault-normal and fault-parallelearthquake records ............................................................................................... 81

4.51 Comparison of peak displacement for different seismic hazard regions(@4"/12", W 1400 lbs/ft) ................................................................................. 82

4.52 Comparison of peak displacement for different seismic hazard regions(@6"/12", W 1400 lbs/ft) ................................................................................. 82

4.53 Selection of median and target peak displacement distributions ......................... 86

4.54 Change of peak displacement considering various mean values ofmodificationfactor ............................................................................................... 86

4.55 Change of peak displacement considering various COV values ofmodification factor ............................................................................................... 87

4.56 Modification factors for sheathing-to-framing connection hystereticparameter variability ............................................................................................ 88


Figure Page

4.57 Graphical method for determination of modification factors inconstruction quality (BW1) ................................................................................... 91

4.58 Graphical method for determination of modification factors inconstruction quality (OWl).................................................................................. 91

4.59 Mean of modification factor for BW1 (OSB only) .............................................. 92

4.60 COV of modification factor for BW1 (OSB only)............................................... 92

4.61 Mean of modification factor for BW1 (OSB + GWB) ........................................ 93

4.62 COV of modification factor for BW1 (OSB +GWB) .......................................... 93

4.63 Mean of modification factor for BW1 (OSB + Stucco) ....................................... 94

4.64 COy of modification factor for BW1 (OSB + Stucco) ....................................... 94

4.65 Mean of modification factor for OWl (OSB only) .............................................. 95

4.66 COV of modification factor for OWl (OSB only) .............................................. 95

4.67 Mean of modification factor for OWl (OSB + GWB) ........................................ 96

4.68 COy of modification factor for OWl (OSB + GWB) ......................................... 96

4.69 Mean of modification factor for OWl (OSB + Stucco) ....................................... 97

4.70 COV of modification factor for OWl (OSB + Stucco) ....................................... 97

4.71 Mean of modification factor for OWl (OSB + GWB + Stucco) ......................... 98

4.72 COV of modification factor for OWl (OSB + GWB + Stucco) .......................... 98

4.73 Graphical method to develop deterministic modification factors innonstructural finish materials effects (BW1) ..................................................... 103

4.74 Graphical method to develop deterministic modification factors innonstructural finish materials effects (0W2) ..................................................... 104


Figure

4.75 Mean of deterministic modification factor for BW1 (OSB sheathing) .............. 104

4.76 Mean of deterministic modification factor for BW1 (Plywood sheathing) ....... 105

4.77 Mean of deterministic modification factor for BW1 (Plywood sheathing) ....... 105

4.78 Mean of deterministic modification factor for OWl (OSB sheathing) .............. 106

4.79 Mean of deterministic modification factor for 0W2 (OSB sheathing) .............. 106

4.80 Performance curve for BW1, OSB (3/8-in.), @3"/6" ......................................... 112

4.81 Performance curve for BW1, OSB (3/8-in.), @4"/12" ....................................... 112

4.82 Performance curve for BW1, OSB (3/8-in.), @6"/6" ......................................... 113

4.83 Performance curve for BW1, OSB (3/s-in.), @6"/12" ....................................... 113

4.84 Performance curve for BW1, OSB (3/s-in.), @3"/6", axes switched ................. 114

4.85 Performance curve for BW1, PWD (3/8-in.), 8d@3"/6" .................................... 114

4.86 Performance curve for BW1, PWD (3/8-in.), 8d@4"/12" .................................. 115

4.87 Performance curve for OWl, OSB (3/8-in.), @3"/3" ......................................... 115

4.88 Performance curve for OWl, OSB (3/s-in.), @4"/4" ......................................... 116

4.89 Performance curve for OWl, OSB (3/8-in.), @6"/6" ......................................... 116

4.90 Performance curve for OWl, PWD (31'8-in.), 8d@4"/4" .................................... 117

4.91 Performance curve for OWL PWD (3/8-in.), 8d@6"/6" .................................... 117

4.92 Effect of model uncertainty on performance curve for BW1,OSB(3/8-in.), @3"/6"......................................................................................... 118

4.93 95thPercentile design chart for BW1, JO (50/50) .............................................. 118


Figure

494 95thPercenti1e design chart for BW1, LS (10/50) ............................................. 119

495 95thPercentj1e design chart for OWl, JO (50/50) .............................................. 119

4.96 95thPercentile design chart for OWl, LS (50/50) ............................................. 120

4.97 Performance curve for BW1, OSB only ............................................................ 123

4.98 Performance curve for BW1, OSB + GWB ....................................................... 123

4.99 Performance curve for BW1, OSB + Stucco ..................................................... 124

4.100 Performance curve for OWl, OSB only .......................................................... 124

4.101 Performance curve for OWl, OSB + GWB ..................................................... 125

4.102 Performance curve for OWl, OSB + Stucco ................................................... 125

4.103 Performance curve for OWl, OSB + GWB + Stucco ...................................... 126

4.104 95thPercentile design chart for BW1, poor quality ......................................... 126

4.105 95thPercentile design chart for BW 1, typical quality ...................................... 127

4.106 95thPercentile design chart for OWl, poor quality ......................................... 127

4.107 95thPercentile design chart for OWl, typical quality...................................... 128

4.108 95t1'-Percentile design chart for BW1, (OSB + Stucco) ................................... 128

4.109 95tIiPercentile design chart for OWl, (OSB + GWB + Stucco) ..................... 129

4.110 Performance curve for BW1, seismic zone III (Seattle), @3"/12" .................. 131

4.111 Performance curve for BW1, seismic zone IV (LA), @3"/12" ....................... 131

4.112 Performance curve for BW1, seismic zone II (Boston), @4"/12" ................... 132

4.113 Performance curve for BW1, seismic zone III (Seattle), @4"/12" .................. 132


Figure iEiig


4.115 Performance curve for BW1, seismic zone II (Boston), @6"/12" ................... 133

4.116 Performance curve for BW1, seismic zone III (Seattle), @6"112" .................. 134


4.118 95tlPercentile design chart for BW1, LS (10/50) ........................................... 135

4.119 Typical IDA curve ........................................................................................... 139

4.120 Estimated collapse points by tangent slope ...................................................... 139

4.121 Set of IDA curves (BW1, group 1) .................................................................. 140



4.124 Set of IDA curves (Owl, group 1) .................................................................. 141

4.125 Set of IDA curves (Owl, group 2) .................................................................. 142

4.126. Set of IDA curves (Owl, group 3) ................................................................. 142

4.127 Peak displacement distributions for different R factors (3"/12", 10) .............. 146

4.128 Peak displacement distributions for different R factors (3"/12", LS) .............. 146

4.129 Peak displacement distributions for different R factors (3"/12", CP) .............. 147

4.130 Fragility curves for three different hazard levels (2"/ 12") ............................... 148

4.13 1 Fragility curves for three different hazard levels (3"/12") ............................... 149

4.132 Fragility curves for three different hazard levels (4"! 12") ............................... 149

4.133 Fragility curves for three different hazard levels (6"/12") ............................... 150


Figure

4.134 Fragility curves considering R = 2.5 (LS, 10/50 hazard level) ........................ 151



4.137 Fragility curves considering R 5.5 (LS, 10/50 hazard level) ........................ 153

4.138 Single fragility curve considering R = 4.5 (LS, 10/50 hazard level) ............... 153

4.139 Fragility curves considering different assumed R factors (LS, 3"/12") ........... 154

4.140 CDF for ultimate force with various R factors (3"/12", JO) ............................ 156

4.141 CDF for ultimate force with various R factors (3"/12", LS) ............................ 156

4.142 CDF for ultimate force with various R factors (3"/12", CP) ........................... 157

4.143 Fragility curve for ultimate uplift force with various R factors(3"/12", HTT 22)................................................................................................ 158

4.144 Fragility curve for ultimate uplift force with various R factors(3"/12", PHD2-SDS3) ........................................................................................ 158

4.145 Fragility curve for ultimate uplift force with various R factors(4"/12", LTT 20B) ............................................................................................. 159

4.146 Hold-down fragility curve considering ultimate uplift capacity ...................... 160

5.1 Plan view and section view for the one-story house model ................................. 163

5.2 Detailed wall configurations for the one-story house model ............................... 164

5.3 Elevation and plan view for two-story house model(from: Fischer et al., 2001) ................................................................................. 166

5.4 SAWS model of the one-story structure, OSB only ............................................ 168


Figure

5.5 Peak displacement distributions for shearwalls in one-story structure,OSB only (JO, 50/50 hazard level) .................................................................... 169

5.6 Peak displacement distributions for shearwalls in one-story structure,OSB only (LS, 10/50 hazard level) .................................................................... 169

5.7 Peak displacement distributions for shearwalls in one-story structure,OSB only (CP, 2/50 hazard level) ...................................................................... 170

5.8 SAWS model of the one-story structure, OSB and NSF materials(GWB and Stucco) ............................................................................................. 173

5.9 SAWS model of the one-story structure, OSB and GWB ................................... 173

5.10 Peak displacement distributions for shearwalls in one-story structure,OSB + GWB (10, 50/50 hazard level) ............................................................... 174

5.11 Peak displacement distributions for shearwalls in one-story structure,OSB + GWB (LS, 10/5 0 hazard level) .............................................................. 175

5.12 Peak displacement distributions for shearwalls in one-story structure,OSB + GWB (CP, 2/50 hazard level) ................................................................ 175

5.13 Peak displacement distributions for shearwalls in one-story structure,OSB + GWB + Stucco (10, 50/150 hazard level) ................................................ 176

5.14 Peak displacement distributions for shearwalls in one-story structure,OSB + GWB + Stucco (LS, 10/50 hazard level) ............................................... 176

5.15 Peak displacement distributions for shearwalls in one-story structure,OSB + GWB + Stucco (CP, 2/50 hazard level) ................................................. 177

5.16 SAWS model of the two-story structure, OSB only(from: Folz and Filiatrault, 2002) ....................................................................... 180

5.17 Peak displacement (relative to ground) distributions for shearwallsin two-story structure (JO, 50/50 hazard level) .................................................. 181


Figure iag

5.18 Peak displacement (relative to ground) distributions for shearwallsin two-story structure (LS, 10/50 hazard level) ................................................. 181

5.19 Peak displacement (relative to ground) distributions for shearwallsin two-story structure (CP, 2/50 hazard level) ................................................... 182

5.20 SAWS model of the two-story Structure, OSB and NSF materials(from: Folz and Filiatrault, 2002) ....................................................................... 185

5.21 Peak displacement (relative to ground) distributions for shearwallsin two-story structure (10, 50/50 hazard level) .................................................. 186

5.22 Peak displacement (relative to ground) distributions for shearwallsin two-story structure (LS, 10/50 hazard level) ................................................. 186

5.23 Peak displacement (relative to ground) distributions for shearwallsin two-story structure (CP, 2/50 hazard level) ................................................... 187

5.24 Comparison of peak displacements at first and second stories,OSB (JO, 50/50 hazard level) ............................................................................ 189

5.25 Comparison of peak displacements at first and second stories,OSB (LS, 10/50 hazard level) ............................................................................ 189

5.26 Comparison of peak displacements at first and second stories,OSB (CP, 2/50 hazard level) .............................................................................. 190

5.27 Comparison of peak displacements at first and second stories,OSB + GWB ± Stucco (10, 50/50 hazard level) ................................................ 190

5.28 Comparison of peak displacements at first and second stories,OSB + GWB + Stucco (LS, 10/50 hazard level) ............................................... 191

5.29. Comparison of peak displacements at first and second stories,OSB + GWB + Stucco (CP, 2/50 hazard level) ................................................. 191

5.30 SAWS model of one-story structure without partition walls,(OSB+GWB) .................................................................................................... 194


Figure

5.31 Peak displacement distributions for one-story structure,OSB + GWB (without partition walls), 10 (50/50 hazard level) ....................... 194

5.32 Peak displacement distributions for one-story structure,OSB + GWB (without partition walls), LS (10/50 hazard level) ....................... 195

5.33 Peak displacement distributions for one-story structure,OSB + GWB (without partition walls), CP (2/50 hazard level) ........................ 195

5.34 Comparison of peak displacement distributions for the effect ofpartition walls and NSF materials, (JO, 50/50 hazard level).............................. 196

5.35 Comparison of peak displacement distributions for the effect ofpartition walls and NSF materials, (LS, 5 0/50 hazard level) ............................. 196

5.36 Comparison of peak displacement distributions for the effect ofpartition walls and NSF materials, (CP, 2/50 hazard level) ............................... 197

5.37 Comparison of peak displacement distributions for isolated shearwall andshearwall in complete one-story structure (JO, 50/50 hazard level) .................. 199

5.38 Comparison of peak displacement distributions for isolated shearwall andshearwall in complete one-story structure (LS, 10/50 hazard level) .................. 199

5.39 Set of IDA curves for selected OSB-only walls with garage door opening(2EW) ................................................................................................................. 202

5.40 Set of IDA curves for selected OSB + NSF walls with pedestrian dooropening(2WW) .................................................................................................. 202

5.41 Fragility curves for the North wall (OSB only) in the one-story structure(without partition walls) ..................................................................................... 204

5.42 Fragility curves for the North wall (OSB + GWB) in theone-story structure (without partition walls) ...................................................... 205

5.43 Comparison of fragility curves for the North wall in theone-story structure (JO, 50/50, 1% drift limit) ................................................... 205


Figure

5.44 Fragility curve for wall with garage door opening, max

(relative to ground) at first story ........................................................................ 207

5.45 Fragility curve for wall with garage door opening, interstory drift ................... 208

5.46 Fragility curve for wall with garage door opening, ömax(relative to ground) at second story.................................................................... 208

5.47 Fragility curve for wall with pedestrian door opening, max

(relative to ground) at first story ........................................................................ 209

5.48 Fragility curve for wall with pedestrian door opening, 6max(relative to ground) at second story.................................................................... 209

5.49 Comparison of fragility curves for shearwafl in two-story structure(JO, 50/50, 1% drift limit) .................................................................................. 210

5.50 Comparison of fragility curves for shearwall in two-story structure(LS, 10/50, 2% drift limit) ................................................................................. 210

5.51 Fragility curves for shearwall with NSF materials (2EW) intwo-story structure ............................................................................................. 212

5.52 Fragility curves for shearwall with NSF materials (2WW) intwo-story structure ............................................................................................. 212

5.53 Comparison of fragility curves showing contribution of NSF materials,max (relative to ground) at first story ................................................................. 213

5.54 Comparison of fragility curves showing contribution of NSF materials,ömax (relative to ground) at second story ............................................................ 213

LIST OF TABLES

Table

3.1 20 Ordinary ground motion records and PGA values ............................................ 19

3.2 Structural performance levels and requirements for woodframe walls(from: Table C 1-3, FEMA 356) ........................................................................... 19

4.1 Sheathing-to-framing connection hysteretic parameters ........................................ 48

4.2 Comparable connection hysteretic parameters from other studies ........................ 49

4.3 Nail properties considered in this study ................................................................. 54

4.4 Matrix of walls used to investigate nonstructural finish material effects .............. 57

4.5 Definitions of three construction quality categories(from: Isoda et al., 2002) ...................................................................................... 66

4.6 Developed deterministic modification factor for construction quality .................. 67

4.7 Target response spectra for different seismic hazard regions ................................ 75

4.8 20 Ordinary ground motion records and PGA values (seismic zone IV, LA) ....... 76

4.9 20 Ordinary ground motion records and PGA values(seismic zone III, Seattle) ..................................................................................... 78

4.10 20 Ordinary ground motion records and PGA values(seismic zone II, Boston) ...................................................................................... 79

4.11 Analysis matrix for effects of different seismic hazard regions .......................... 79

4.12 Summary of modification factors considering construction quality .................... 90

4.13 Developed deterministic modification factor (ty) for contribution ofnonstructural finish materials effects ................................................................. 103

4.14 Fastener parameters used to develop performance curves anddesign charts for baseline walls ......................................................................... 108

LIST OF TABLES (Continued)

Table

4.15 Seismic weights calculated based on UBC '97 allowable unit shear values(Table 23-TI-I-i) ................................................................................................. 145

4.16 Capacities of hold-downs considered in this study ............................................ 155

5.1 Hysteretic parameters for the shearwall spring elements inone-story structure, OSB only ............................................................................ 167

5.2 Hysteretic parameters for the shearwall spring elements inone-story structure, OSB and NSF materials ..................................................... 172

5.3 Hysteretic parameters for the shearwall spring elements,OSB sheathing only (from: Folz and Filiatrault, 2002) ..................................... 179

5.4 Fitted hysteretic parameters for the SDOF shear element model of an8 ft>< 8 ft shearwall with stucco and gypsum wallboard(from: Folz and Filiatrault, 2002) ....................................................................... 183

5.5 Hysteretic parameters for the shearwall spring elements,OSB and NSF materials (from: Folz and Filiatrault, 2002) ............................... 184

5.6 Estimated collapse limit (from IDA) for shearwall in the completetwo-story structure ............................................................................................. 201

LIST OF APPENDICES

Appendix

A. Example showing convolution of hazard curve and fragility curve ...................... 231

B. Deterministic modification factors for construction quality ................................. 233

C. Scaling earthquake records to response spectra considering different scalingmethods.............................................................................................................. 236

D. Earthquake records used in this study ................................................................... 239

E. Peak displacement distributions considering different R factors .......................... 244

F. Fragility curves for baseline wall (BW1) considering different hazard levels ...... 250

G. Fragility curves for baseline wall (BW1) considering different R factorsand nailing schedules ......................................................................................... 257

H. CDF for baseline wall (BW1) considering ultimate force withvariousR factors ................................................................................................ 262

LIST OF APPENDIX FIGURES

Figure

A.1 Convolution of hazard curve and fragility curve ................................................ 232

C. 1 20 0GM records (CUREE) scaled over the plateau region of the responsespectrum(LS, 10/50) .......................................................................................... 237

C.2 20 0GM records (CUREE) scaled at a period of 0.2 sec to the responsespectrum(LS, 10/50) .......................................................................................... 237

C.3 20 0GM records (CUREE) scaled at a period of 0.5 sec to the responsespectrum(LS, 10/50) .......................................................................................... 238

E. 1 Peak displacement distributions considering different R factors(2"/12", JO) ........................................................................................................ 245

E.2 Peak displacement distributions considering different R factors(2"/12", LS) ........................................................................................................ 245

E.3 Peak displacement distributions considering different R factors(2"/12", CP) ....................................................................................................... 246

E.4 Peak displacement distributions considering different R factors(4"/12", 10) ........................................................................................................ 246



E.7 Peak displacement distributions considering different R factors(6"/12", 10) ........................................................................................................ 248



F.1 Fragility curves (R= 2.5, 2"/12") ........................................................................ 251

LIST OF APPENDIX FIGURES (Continued)

Figure

F.2 Fragility curves (R=2.5, 3"/12") ........................................................................ 251

F.3 Fragility curves (R = 2.5, 4"/12") ........................................................................ 252


F.5 Fragility curves (R 3.5, 2"/12") ........................................................................ 253

F.6 Fragility curves (R 3.5, 3"/12") ........................................................................ 253




F.10 Fragility curves (R 4.5, 3"/12") ...................................................................... 255

F. 11 Fragility curves (R = 4.5, 4"/12") ...................................................................... 256

F. 12 Fragility curves (R = 4.5, 6"/12") ...................................................................... 256

G.1 Fragility curves considering R = 2.5 (10, 50/50 hazard level) ............................ 258


G.3 Fragility curves considering R = 4.5 (JO, 50/50 hazard level) ............................ 259


G.5 Fragility curves considering R = 3.5 (CP, 2/50 hazard level) ............................. 260



H. 1 CDF for ultimate force with various R factors (2"/12", 10) ............................... 263

H.2 CDF for ultimate force with various R factors (2"/12", LS) ............................... 263

LIST OF APPENDIX FIGURES (Continued)

Figure

H.3 CDF for ultimate force with various R factors (2"/12", CP) .............................. 264

H.4 CDF for ultimate force with various R factors (4"/12", 10) ............................... 264


H.6 CDF for ultimate force with various R factors (4"/12", CP) .............................. 265

H.7 CDF for ultimate force with various R factors (6"/12", 10) ............................... 266


H.9 CDF for ultimate force with various R factors (6'712", CP) .............................. 267

LIST OF APPENDIX TABLES

Table

B. 1 Deterministic modification factors for construction quality ............................... 235

D.1 Set of LA ordinary ground motion records (CUREE project) ............................ 240

D.2 Set of LA earthquake ground motions with 10% probability ofexceedence in 50 years (SAC project) ............................................................... 241

D.3 Set of Seattle earthquake ground motions with 10% probability ofexceedence in 50 years (SAC project) ............................................................... 242

D.4 Set of Boston earthquake ground motions with 10% probability ofexceedence in 50 years (SAC project) ............................................................... 243

Performance-Based Seismic Design of Light-Frame Shearwalls

1. INTRODUCTION

1.1 General

Wood is the most common material used in one- and two-story residential

construction in the United States. Light-frame wood structures have a number of

advantages including aesthetics, beauty, construction cost and time, versatility,

flexibility in floor plans, and so forth. Most woodframe structures consist of floors,

walls, and roof systems tied together by fasteners. Shearwalls and diaphragms provide

the primary resistance to lateral forces in woodframe structures.

Light-frame wood structures generally have performed well with regard to life-

safety under natural hazard loadings such as earthquakes and hurricanes. Properly

built woodframe structures can withstand major earthquakes and hurricanes without

collapsing. However, costly damage (both nonstructural and secondary assemblies),

which can add significantly to the total economic loss in natural hazards, remains a

problem. Many woodframe structures designed to meet current standards (code

requirements) were damaged in recent natural disasters such as the Northridge

earthquake and hurricane Andrew [NAHB, 1993, 1994]. In the wake of these and

other events, the structural engineering community has come to recognize the

limitations of current design provisions, particularly with respect to damage

prevention. For example, current seismic design procedures for light-frame structures

[ICBO, 1997; AF&PA, 2001] require an estimate of the elastic fundamental period.

2

This may not be simple to estimate since woodframe buildings exhibit inelastic

response over the entire range of lateral deformation. Current strength-based code

procedures do not allow for a proper assessment of the safety of engineered buildings

considering the various limit states that these structures may have to meet during their

service-life. Therefore, the structural engineering community has started to embrace a

new design approach (termed "performance-based design") in order to address more

explicitly various performance requirements. Although performance-based seismic

design has advanced for some materials and structural types, such as steel and

reinforced concrete buildings and bridges [SAC, 1995; Wen and Foutch, 1997], its

application to light-frame structures remains largely unexplored.

In recent years, the concept of performance-based design has gained interest

among designers and researchers. Performance-based design includes selection of

appropriate building sites, structural systems and configurations, as well as analytical

procedures used in the design process, to confirm that the structure has adequate

strength, stiffness and energy dissipation capacity to respond to the design loads

without exceeding permissible damage states [SEAOC, 1999; FEMA, 2000 a,b; AISC,

2001]. The objective of performance-based design is to obtain a more reliable

prediction of structural behavior, quantifying and controlling the damage risk to an

acceptable level during the service-life of the structure [Moller et al., 2001].

Performance-based design has evident benefits. These benefits have to be made clear

in order for performance-based design to be an accepted alternative to present design

procedures.

3

Although performance-based design concepts are gaining acceptance in the

design community, these are a number of obstacles that must be overcome for

performance-based design to be widely accepted. Performance objectives (including

both performance levels and hazard levels) must be formulated in a probability-based

format to take proper account of the various sources of uncertainty. Uncertainties can

be classified as aleatory or epistemic. Aleatory uncertainties arise from inherent

variability in (e.g., material) properties, whereas epistemic uncertainties arise from a

deficiency in the knowledge base, including limited data or model uncertainties. In

design for natural hazards, the greatest source of uncertainty arises from the hazard

itself. For example, variability in the seismic hazard (as represented by a characteristic

suite of ground motions) contributes the greatest uncertainty to the predicted response

(peak displacement) of a woodframe shearwall. The ground motions are highly

variable in terms of peak ground acceleration, strong motion duration, frequency

content, and so forth. Other uncertainty sources could include the analytical models,

material and connection properties, construction materials, workmanship, and so on.

Efforts to develop performance-based design procedures must identify and quantify

sources of uncertainty to accurately evaluate the associated reliability (performance)

levels.

The objective of the proposed research is to explore the potential for the

application of performance-based engineering concepts to the design and

assessment of woodframe structures subject to earthquakes. To accomplish this, a

general methodology will be developed for assessing probabilistic response of

ru

woodframe structures. The eventual adoption of performance-based concepts in design

can lead to an improvement in performance, reduction in property destruction and

damage, improvement in durability, and reduction in maintenance costs of woodframe

structures. This research also can provide a technical basis for the development of

further performance-based design provisions for woodframe construction.

1.2 Scope and objectives

The focus of this research is on shearwalls in woodframe structures subject to

earthquake loading. The shearwalls are treated as isolated subassemblies (Chapter 4)

or as parts of complete systems (Chapter 5). Shearwalls comprise the vertical elements

in the lateral force resisting system of woodframe structures. They support the

horizontal diaphragms and transfer the lateral forces downward into the foundation. A

number of sheathing materials can be used to develop shearwall action in a light-frame

wall. These include wood structural panels such as OSB and plywood, gypsum

wallboard (interior finish material), and stucco (exterior finish material).

There are three main objectives in this research. The first is the development of

general methodology for assessing probabilistic response of wood shearwalls subject

to earthquake loading while considering the various parameters (ordinary ground

motion records, effects of nonstructural finish materials, construction quality, effects

of different seismic hazard regions, and sheathing-to-framing connection hysteretic

parameters) which affect shearwall performance. The shearwall response (peak

displacement) is obtained by nonlinear time history analysis using the analytical model

CASHEW and visually best-fit program, SASHFIT (detailed descriptions of both

programs are provided in Chapter 3). The second objective is the development of

probability-based (risk-consistent) design aids for woodframe shearwall design

(selection) in seismic regions. The resulting design aids (performance curves and

design charts) can be used in both design and evaluation applications. The third

objective is the application of fragility methodology, which can be used for design and

post-disaster condition assessment.

2. BACKGROUND AND LITERATURE REVIEW

In the early 1990's, several natural disasters struck opposite ends of the United

States. Hurricane Andrew struck the coast of Florida in 1992 and the Northridge

earthquake hit Southern California in 1994. These two large-scale natural hazards

caused tremendous damage to residential woodframe structures in these regions.

According to an NAHB survey, the main forms of damage to residential woodframe

structures were roof sheathing removal due to wind loading and damage to interior and

exterior finish materials due to earthquake loading [NAHB, 1993; 1994]. In seismic

events, shearwalls function mainly to resist lateral force, while in high-wind events,

roof systems function primarily as sheltering elements for the interior spaces of

buildings. In light of the costs of these recent natural disasters, many studies have

focused on mitigating damage through the development and implementation of

improved design procedures.

Wood shearwalls have been the subject of extensive investigation in recent

years. Numerous experimental tests have been conducted and both static and dynamic

analysis models have been developed to describe shearwall performance subject to

earthquake loading [Foschi, 1977; Tuomi and McCutcheon, 1978; McCutcheon, 1985;

Stewart, 1987; Cheung et al., 1988; Dolan, 1989; Filiatrault 1990; Dolan and Madsen,

1992; Durham, 1998; Dinehart and Shenton, 1998; Salenikovich, 2000; Dinehart and

Shenton, 2000; Folz and Filiatrault, 2000; 2001; 2002]. More recently, reliability

concepts have been applied to predicting shearwall performance under seismic

loading. Ceccotti and Foschi (1998) evaluated the earthquake design procedure for

7

woodframe shearwalls in the Canadian National Building Code using First-Order

Reliability Method (FORM) techniques. Paevere and Foliente (2000) investigated the

effect of hysteretic pinching and stiffness degradation on the peak displacement and

reliability of shearwalls using the Bouc-Wen-Baber-Noori (BWBN) model combined

with Monte Carlo Simulation. Rosowsky and Kim (2002a) proposed a risk-based

methodology for woodframe shearwall design considering a suite of earthquake

records and using a numerical model (CASHEW) and nonlinear dynamic time history

analysis. Another study by van de Lindt and Walz (2003) used a new hysteretic model

for dynamic analysis of wood shearwalls and fit the response to a Weibull distribution.

A large, multi-university project (the CUREE-Caltech Woodframe Project) with the

overall objective of developing improved analysis and design techniques for

woodframe structures is nearing completion at the time of this research. The project

included shake table tests of various woodframe assemblies and structures,

development of testing protocols, consideration of effects of anchorage and wall finish

materials, testing of nail and screw fastener connections, development of seismic

analysis software, reliability studies, and other aspects of woodframe structures

subject to earthquake loading [Camelo et al., 2002; Cobeen, 2001; Deierlein and

Kanvinde, 2003; Folz and Filiatrault, 2000; Fonseca et al., 2001; Isoda et al., 2001;

Krawinkler et al., 2000; Mahaney and Kehoe, 2002; McMullin and Merrick, 2001;

Rosowsky and Kim, 2002a]

As described above, many studies have focused on reducing the damage to

woodframe structures subject to natural hazards. Many of these are based on emerging

[I]

performance-based design concepts. The concept of performance-based design, in

actuality, is not new. The U.S. Department of Housing and Urban Development

experimented with what would later be known as performance-based design when

they sponsored a large research program ("Operation Breakthrough") to develop

criteria for design and evaluation of innovative housing systems [Performance, 1977].

This concept appeared again after the Northridge earthquake of 1994, where it became

apparent that buildings designed by code for life safety did not perform up to

performance expectations in other aspects.

Performance-based design, when implemented successfully, can contribute

effectively to the reduction of damage and associated losses, as well as improvement

in the performance and safety of structures under natural hazard loadings.

Performance-based design requires, most importantly, a realistic model for the

structural behavior under appropriately described natural hazard loadings. In addition,

tools are needed for the evaluation of probabilistic assessment of the response in order

to quantify the exceedence probability for each of the relevant performance states.

Performance-based design consists of four key features: performance levels,

seismic hazard levels, performance objectives, and confidence levels. Performance

levels are a state of defined and observable damage in a structure or structural

component. The performance goals should be based on reliability and uncertainty

principles. In other words, they should be based on calculated responses associated

with observed behaviors, and the acceptable risks should be determined in relationship

to other societal risks. Seismic hazard levels are representations of variation in suitable

parameters of the annual probability of exceedence. Performance objectives are the

coupling of performance levels with hazard levels. Confidence levels that the building

will satisfy the design requirements must then determined. Performance objectives

must be translated into engineering quantities to establish acceptance criteria, defined

as limiting values in the response parameters that become targets for the design

[AISC, 2001]. The expression of performance requirements is one of the most

significant challenges in developing performance-based design concepts.

10

3. NONLINEAR DYNAMIC TIME-HISTORY ANALYSIS

3.1 Computer programs

3.1.1 CASHEW

The CASHEW (Cyclic Analysis of Shearwalls) program was used in this study

to evaluate the dynamic response of woodframe shearwalls (treated as isolated

subassemblies). Specifically, the response quantity of interest was peak displacement

(or "drift") at the top of the wall. CASHEW is a numerical model capable of

predicting the load-displacement response of wood shearwalls under quasi-static

cyclic loading, and was developed under Task 1.5.1 (Analysis Software) of the

CUREE-Caltech Woodframe Project {Folz and Filiatrault, 2000, 2001]. With

information on shearwall geometry, material properties, and the hysteretic behavior of

the individual fasteners, CASHEW can be used to calibrate the parameters of an

equivalent SDOF system (modified Stewart hysteretic model). This is done, for

example, using the CUREE-Calteeh loading protocol developed under Task 1.3.2 of

the CUREE-Caltech Woodframe Project. The equivalent SDOF hysteretic model can

then be used to predict the global cyclic response of a shearwall under arbitrary quasi-

static cyclic loading or, using a nonlinear dynamic time-history analysis program, and

actual ground motion records. The CASHEW modeling procedure is illustrated in

Figure 3.1. Details of the numerical modeling procedure, the loading protocol, the

system identification procedure used to define the equivalent SDOF model parameters

is provided elsewhere [Folz and Filiatrault, 2000, 2001].

11

With assumed structural Single set of hystereticand connection parameters parameters for given wa/I

EquivalentI Nonlinear SOOF

CUREE bas ICASHE

llarProtocJ Program

J

A...

Ground Scaling _____Scaledground HImotion Procedurej moti

Ordinary ground motions UBC design spectrum Suite of 0GM records scaledcharacterizing seismic NEHRP guidelines for specific performance levelhazard in southern CA (LS-10/50, 10-50/50)

Figure 3.1 CASHEW modeling procedure

3.1.2 SASH!

SASH 1 is a nonlinear dynamic time history analysis program used to analyze

shearwalls under actual earthquake ground motions. The shearwall is modeled as a

single degree-of-freedom nonlinear oscillator using a modified Stewart hysteretic

model. The global shearwall hysteretic parameters used as input to SASH 1 are

obtained from CASHEW or SASHFIT (described next). The mass and damping ratio

as a percentage of critical, along with the earthquake record scaled appropriately for

the target hazard levels, also are required input for SASH 1. The SASH 1 program then

performs a nonlinear dynamic time history analysis to predict the peak relative

displacement of shearwall. The program also generates peak relative acceleration,

peak relative velocity, peak absolute acceleration, and peak force at top of wall.

12

3.1.3 SASHFIT

SASHFIT is a spreadsheet program which can be used to develop a set of

hysteretic parameters for the behavior of a single fastener or an entire (isolated)

shearwall. It requires the complete cyclic test data (i.e., load-deformation curve) for

the particular fastener or assembly. SASHFIT is developed based on the following

equation [Foschi, 1977]:

sgn(8) (F0 + K0 S)x [i exp( K0 s)i F0 5j

F= sgn(8)xF+r2K0[8_sgn(s)x8J (3.1)

0,

where, F = global force, 8 = deformation, 8u = deformation at ultimate load, Sp' =

deformation at failure, F0 = force intercept of the asymptotic line, K0 initial

stiffness, rjK0 = asymptotic stiffness under monotonic load, F = ultimate load, and

r2K0 = post ultimate strength stiffness under monotonic load. This equation was

developed based on monotonic loading, so further consideration of cyclic loading is

required. Figure 3.2 shows a load-deformation curve under an arbitrary cyclic loading.

In this figure, r3K0 = unloading stiffness, r4K0 re-loading pinched stiffness,

degrading stiffness K = KO(6O/8m)a, 6m /38, a and /3= hysteretic parameters for

stiffness degradation, and = final unloading displacement.

13

30

20

z-

Uci)00LI

-10

-20

-30

-

lj (o,F)

1rK- I

GF0

K

E 14KO__120- [Cyclic Loading Protocoj

E80-

40-0)

Eci)

<-, 0-

DD

-80-1111111 111111 I I I 11111111 I III I 1111111

-80 -40 0 40 80 120Displacement, (mm)

Figure 3.2 Force-displacement response of a wood shearwall under cyclic loading.Hysteretic model is fit to test data for an 8 ft x 8 ft shearwall with 3/g-in. thick OSBsheathing panels (from: Durham, 1998)

The hysteretic response of a typical shearwall exhibits the same defining

characteristics (pinched behavior, strength and stiffness degradation, etc.) as those of

the individual sheathing-to-framing connector under cyclic loading [Dolan and

Madsen, 1992]. Consequently, the hysteretic model presented here, which was used to

represent the hysteretic behavior of sheathing-to-framing connectors [Folz and

Filiatrault, 2001], also can be used to represent the global hysteretic response of a

shearwall under cyclic loading with appropriate model parameter values. The

spreadsheet application, SASHFIT, developed as part of this research, can be used to

identify the ten hysteretic parameters for the shearwall directly from full-scale cyclic

14

shearwall test data. Alternatively, SASHFIT can be used to identify the ten hysteretic

parameters of the individual fasteners, and CASHEW can be used to determine the

global hysteretic parameters of the overall shearwall. Figure 3.3 shows a comparison

of one example of the model based on the parameters determined by SASHFIT with

the original cyclic load-displacement curve.

C',a

0-J

1r\

Displacement (in.)

Figure 3.3 Load-displacement curve using parameters determined by SASHFIT

3.1.4 SAWS

The SAWS (Seismic Analysis of Woodframe Structures) program was

developed to predict the seismic response of a complete structure [Folz and Filiatrault,

2002]. In this model, the light-frame structure is composed of two primary

components: rigid horizontal diaphragms and nonlinear lateral load resisting shearwall

elements. In the modeling of the structure, it is assumed that both the floor and roof

15

elements have sufficiently high in-plane stiffness to be considered rigid elements. This

is expected to be a reasonable assumption for typically constructed diaphragms with a

planar aspect ratio on the order of 2:1, as supported by experimental results from full-

scale diaphragm tests [Philips et al., 1993]. The actual three-dimensional building is

degenerated into a two-dimensional planar model using zero-height shearwall spring

elements connected between the diaphragms and the foundation. All diaphragms in the

building model are assumed to have infinite in-plane stiffness [Folz and Filiatrault,

2002].

The SAWS program executes linear dynamic analysis, nonlinear dynamic

analysis, and quasi-static pushover analysis on the woodframe building model with

information of building configuration, masses of system, hysteretic wall parameters,

viscous damping parameters, and earthquake records. The program also can be used to

predict the response of structures consisting of shearwalls with nonstructural finish

materials such as stucco and gypsum wallboard. The hysteretic parameters for the

shearwalls in the structure can be obtained by CASHEW or SASHFIT. Selected

hysteretic parameters for nonstructural finish (NSF) materials are provided in the

SAWS report [Folz and Filiatrault, 2002], however they also can be developed using

SASHFIT if full-scale test data are available. The SAWS report only provided

hysteretic parameters for an 8 ft. x 8 ft. wall with NSF materials (stucco and gypsum).

Therefore, these values must be adjusted for the length of the actual wall and for the

presence of door and window openings [Folz and Filiatrault, 2002].

16

3.2 Ordinary ground motion records

The ground motions considered in this research were obtained from Task 1.3.2

(Loading Protocol) of the CUREE-Caltech Woodframe Project (CCWP). The suite of

20 ordinary ground motion (0GM) records are assumed to be representative of the

10% in 50 years (10/50) hazard level for California conditions and formed the basis

for the development of the CUREE-Caltech loading protocol {Krawinkler et al., 2000;

Filiatrault and Folz, 2001]. Two limit states, life safety (LS) and immediate occupancy

(10), were the focus of this study. The life safety limit state is paired with the 10%

probability of exceedance in 50 years (10/50) hazard level, while the immediate

occupancy limit state is paired with the 50% probability of exceedance in 50 years

(50/50) hazard level [FEMA, 2000a,b]. For the life safety (LS, 10/50) limit state

analyses, each record was scaled such that its mean 5% damped spectral value

between periods of about 0.12 and 0.58 seconds matched the UBC design spectral

value of 1.lg for the same period range {ICBO, 1997]. For the immediate occupancy

(10, 50/50) limit state, the records were scaled according to the procedure

recommended in the NEHRP Guidelines [FEMA, 2000a,b]. Seismic zone 4 and soil

type D were assumed for most cases in this study. The code provided target response

spectra are shown in Figure 3.4 with three hazard levels (JO, LS, and CP). The scaled

peak ground accelerations for the 20 records are shown in Table 3.1.

The evaluation of the collapse prevention (CP, 2/50) limit state, which is

paired with the 2% probability of exceedance in 50 years (2/5 0) hazard level [FEMA,

2000a,b], is the subject of some discussion. While not considered as extensively as the

17

LS (10/50) and 10 (50/50) limit states here, the CP (2/50) limit state (generally

associated with a 2% in 50 years hazard level) was considered in selected stages of

this research. The peak ground accelerations for these records also are shown in Table

3.1. Again, the focus of this study was on the life safety (LS, 10/50) and immediate

occupancy (JO, 50/50) limit states.

3.3 Distribution functions (exceedence probability curves)

The greatest source of variability (or more specifically, the largest contribution

to the variability in peak response) arises from the ground motions themselves (i.e.,

the suite of 20 ordinary ground motion records characterizing the seismic environment

in California). It was therefore decided to present the peak displacements obtained

using each of the ground motions, scaled as appropriate for the limit state, in the form

of a sample cumulative distribution function (CDF). The relative contribution of the

ground motion variability to the overall response (drift) variability will be addressed in

the next section. These distribution functions provide a convenient method for

estimating probabilities of exceedence, or "non-performance." That is, one can quickly

evaluate the probability of the maximum shearwall drift exceeding (or not exceeding)

a prescribed level at the hazard level defined by the suite of ground motions. In this

study, the prescribed performance levels correspond to the FEMA 356 drift limits (see

Table 3.2). Once the peak displacement distributions are determined, they can be post-

processed into a form useful for design and/or assessment for given target

probabilities. This process is illustrated in Figure 3.5 and will be described later in

Section 4.3.2.

2

1.8

1.6

1.4

1.2

0)

(I)

0.8

0.6

0.4

0.2

- - -

I

([ \

OL

0

Collapse Prevention (2%/5oyrs)

Life Safety (10%/50yrs)

Immediate Occupancy (50%I5Oyrs)

0.5 1 1.5 2 2.5 3 3.5 4Period (sec)

Figure 3.4 Code based target response spectra

Iv

EQ Event&Year File Station

Peak Ground Acceleration (g)Raw Scaled

OG 1 10(50/50)

LS(10/50)

CP(2/50)

SUP1 Brawley 0.116 0.264 0.604 0.985SUP2 El Centro Imperial County Center 0.258 0.255 0.584 0.973SUP3 PlasterCity 0.186 0.174 0.398 0.643NOR2 Beverly Hills 14145 Mulhol 0.416 0.205 0.470 0.759NOR3 CanogaPark-TopangaCanyon 0.356 0.261 0.599 0.967

North d NOR4 Glendale Las Palmas 0.357 0.206 0.472 0.762NOR5 LA-Hollywood Storage 0.231 0.210 0.482 0.778NOR6 LA (North) Faring Road 0.273 0.266 0.609 0.984NOR9 NorthHollywood-Coldwater 0.271 0.212 0.485 0.783

NOR1O Sunland-MtGleasonAve 0.157 0.206 0.472 0.762LP1 Capitola 0.529 0.185 0.423 0.683LP2 Gilroy Array #3 0.555 0.206 0.473 0.764

Loma Prieta LP3 Gilroy Array #4 0.417 0.227 0.520 0.840(1989) LP4 Gilroy Array #7 0.226 0.179 0.410 0.662

LP5 Hollister Differential Array 0.279 0.181 0.415 0.670LP6 Saratoga West Valley 0.332 0.262 0.600 0.969

Cape Mendocino CM1 Fortuna Boulevard 0.116 0.231 0.530 0.856(1992) CM2 Rio Dell Overpass 0.3 85 0.232 0.532 0.859

Landers LAN! DesertHotSprings 0.154 0.237 0.542 0.875(1992) LAN2 Yermo Fire Station 0.152 0.174 0.399 0.644

Table 3.1 20 Ordinary ground motion records and PGA values

Structural Performance Levels

Elements]

Typej

Collapse Prevention] Life SafeF Immediate

OccupancyConnections loose, Moderate loosening Distributed minorNails partially of connections and hairline cracking ofwithdrawn. Some minor splitting of gypsum and plaster

Primary splitting of members veneersmembers andpanels. Veneers

Wood Stud dislodged

Walls Sheathing sheared Connections loose, Same as primaryoff Let-in braces Nails partially

Secondary fractures and withdrawn. Somebuckled. Framing splitting ofsplit and fractured. members and

panels.

Drift 3% transient or 2% transient; 1% transient;permanent. 1% permanent 0.25% permanent

Assumed hazard level and 2/50 (2% in 50 10/50 (10% in 50 50/50 (50% in 50mean return period years), years), 474 years years), 72 years

2475 years

Table 3.2 Structural performance levels and requirements for woodframe walls (from:

Table C1-3, FEMA 356)

20

Figure 3.6 illustrates the construction of the sample cumulative distribution

function (CDF) and the fitted distribution function Fx(x) for the peak displacement of

one wall considering both the life safety (LS, 10/50) and immediate occupancy (JO,

50/50) limit states. Each point represents the peak drift obtained from a nonlinear

time-history analysis for a particular ground motion record. The results are then rank-

ordered to construct the sample cumulative distribution function (CDF), which is then

fit to a lognormal (LN) distribution given by:

F(x) (3.2)

where 'Do = standard normal cumulative distribution (CDF) function, X = logarithmic

mean, and = logarithmic standard deviation. The LN parameters (X, ) are obtained

using a maximum likelihood procedure. In addition to providing a good fit, the LN

distribution is the most convenient distribution form for fragility analysis as well as

consideration of model uncertainty, as will be discussed later. Figure 3.6 can be used,

for example, to evaluate the probability that a wall of this type will exceed a certain

peak drift, again assuming California seismic hazard conditions. For example, the

probability that the peak drift will exceed the FEMA 356 drift limit for life safety (LS,

10/50) of 2% is about 0.03 (or a 97% non-exceedence probability). As can be seen

from this figure, the probability of non-performance of this wall considering the

immediate occupancy (JO, 50/50) limit state is very small. The JO (50/50) distribution

function is considerably below the 1% drift limit prescribed by FEMA 356.

21

SASH11

PeakScaledground displacement

ds Lprogram

jdtribution,Fx(x)motion

Nonlinear time-history analysis

Seismic hazardModified Stewart hysteretic model Response distribution for

characterizationParameters from CASHEW given seismic weight

II

J

PerformanceDesign charts for curves

shearwall selection (peak drift vs.seismic weight)

One set for each One set for each combination ofnon-exceedence Structural parameters (sheathing,probability level Fastener type, fastener spacing)

Figure 3.5 Development of probability-based design charts for shearwall selection

0.9

0.8

0.7

0.6

0.4

0.3

0.2

0.1

C)

+ *

+

IO(5lI50)*

*

*

LS(10/50)

. I I

I I

* I8ft.

Ill+H i * I

C'4 8ft.*

+ d ii BW (8 >< 8), 8d3"I6,+ f OSB(3/8"),ED= /8',

G =l8Oksi,ç=2%,+ *W = 1400 lbs/ft (50 kN total)

0 0.5 1 1.5 2 2.5 3

6max (in.)

Figure 3.6 Fitting a lognormal distribution to the sample CDF of peak displacements

22

4. ANALYSIS OF ISOLATED SHEARWALLS

4.1 Model configuration (isolated shearwall)

The focus of this chapter is on the performance of shearwalls in woodframe

structures under real earthquake loading. These shearwalls are treated as isolated

subassemblies in this chapter. Shearwalls acting as part of a complete structure are

discussed in Chapter 5.

Woodframe shearwalls typically have three major components: dimension

lumber framing, sheathing panels, and fasteners. The dimension lumber framing

elements generally are nominal 2 in. x 4 in. or nominal 2 in. x 6 in. sawn lumber

pieces. These are oriented horizontally (plates and sills) and vertically (studs) with

only nominal nailing to hold the framework together. The top plate and end studs

generally consist of double members, while the sole plate and the interior studs are

single members. Studs are generally spaced at 16 in. or 24 in. on center. Hold-downs

are used to prevent overturning of the wall and ensure a racking mode of deformation.

For exterior sheathing, structural panels such as oriented strand board (OSB) and

plywood are most commonly used. Gypsum wallboard is most commonly used for

interior sheathing and stucco is a widely used finishing material, particularly in

California. Sheathing panel thicknesses vary, but 3/8-in., 15/32-in., and 7/16-in. are the

most common in typical lightframe shearwall construction. Sheathing panels are

usually 4 ft. x 8 ft. in size and are installed either vertically or horizontally. Blocking

is often used when panels are installed horizontally. The sheathing is attached to the

framing with dowel-type fasteners such as nails (most common), screws, or staples,

23

although adhesives sometimes are used. These fasteners are typically spaced at regular

intervals with fastener lines around the perimeter of the sheathing panels more densely

spaced than throughout the sheathing panel interior. Figure 4.1 depicts the components

of typical woodframe shearwall.

i:rnr.r.r

Top-plate Sheathing-to-framing connector

Sill Sheathing material

Figure 4.1 Components of typical woodframe shearwall

Three (typical) shearwall configurations were considered in this study. The

first baseline wall (BW1) was a solid 8 ft. x 8 ft. wall built with 2 in. x 4 in. nominal

lumber, 4 ft. x 8 ft. OSB sheathing materials (3/8-in. thicimess) oriented vertically and

various nailing schedules (3"/12", 4"/12", and 6"112"). Double top plate and end studs

were assumed with single sole plate and interior studs. The stud spacing was 16 in.

and properly installed hold-downs were assumed to be present. The second baseline

24

wall (OW!) was a long wall with a 16 ft. garage door opening (see Figure 4.2) This

wall was built with 2 in. x 4 in. nominal framing lumber, vertically oriented OSB

sheathing (3/8-in. thickness) at the wall ends, a solid header over the opening, and

various nailing schedules. Hold-down anchorage was assumed to be present and

properly installed. The third baseline wall (0W2) was a 16 ft. long wall with a

pedestrian door opening (see Figure 4.2). The construction parameters were similar to

those of OWl.

25

- - - ----11-. S

II I

II I III I

.

II I

II III I

II

.

II

IIBV\/1 S

II i I

II III I

S S S

II I

II I

II

II

II

II

II I

_11___

.

S S S

.

I S I

4ft. 4ft.

8ft.

11l

. S S

S

: :

I I

I I

-----Il----

owl

loft.

.

.

S S

S

I

3ft. 3ft.4-16 ft.

-1

I I I

S

S.

.

S

I II I II

II

I II

I

I

I II

I

I I

S es S

0W2

3ft.

..

: :

S S

.: .

3ft. 3.5 ft. 3.5 ft. 3ft.

16 ft.

Figure 4.2 Detailed configurations of baseline solid wall (BW1) and walls with

openings (OWl and 0W2)

26

4.2 Sensitivity studies

4.2.1 Baseline sensitivity studies

Sensitivity studies were performed to investigate the relative contribution of

both aleatory (inherent) and epistemic (knowledge-based) uncertainties to the

estimation of shearwall peak displacement. Some factors affecting performance of

wood structures are inherently random (aleatory) in nature, and thus are irreducible at

the current level of engineering analysis {Ellingwood et al., 2003]. Examples would

include strength of wood in tension or in compression parallel-to-grain, shear modulus

values for sheathing materials and fastener hysteretic parameters. Others arise from

the assumptions made in the analysis of the system and from limitations in the

supporting databases. In contrast to the aleatory uncertainties, these knowledge-based

(epistemic) uncertainties depend on the quality of the analysis supporting databases,

and generally can be reduced, at the expenses of more comprehensive (and costly)

analysis. Sources of epistemic uncertainty in light-frame wood construction include

modeling error (CASHEW), two-dimensional models of three-dimensional buildings,

seismic mass, and probabilistic models of uncertainty estimated from small data

samples. The effects of these sources of uncertainty (variability) on shearwall response

(peak displacement) are described in the following sections. The four solid walls (all 8

ft. x 8 ft.) considered in the sensitivity studies are shown in Figure 4.3. The baseline

wall used for most comparisons is the 8 ft. < 8 ft. solid wall with two sheathing panels

oriented vertically, designated BW1.

II I

II I

II I IIII I II

II

I II

II

II

II I IIII I IIII I IIII I II

III

III

I

II

I

II

II

II

II

I II

I II

I II

I II

I

I

I

II

II

I IIII

I III II

III III III III II

I II

I II

II

I II

I II

I II

II

I II

I II

I II

I II

I

I II

I

4ft. 4ft.

8 ft.

Baseline Wall 1 (BW1)

18 ft.

ir--II II

II

II

4ft.

8III

III I

--

il


rt.

27

-

4ft.

-

I I

I

-

4ft

8

8ft.- - -


-4ft.- 8

I II

4ft. 4ft.-

8 ft.


Figure 4.3 Baseline wall sheathing configuration

4.2.1.1 Ground motions

ft.

The greatest contributor to response variability is expected to be the ground

motions. The 20 ground motions taken to be representative of the seismic hazard in

southern California are highly variable (in terms of spectral acceleration, strong

motion duration, frequency content, etc.). Figure 4.4 shows the variability of relative

!41

response (peak displacement) obtained using the CASHEW modeling procedure

described in Section 3.1.1 for one given wall configuration, assuming the Durham nail

hysteretic parameters (described in Section 4.2.2), and considering the three limit

states: life safety (LS, 10/50), immediate occupancy (JO, 50/50), and collapse

prevention (CP, 2/50). The distributions shown for 10 (50/50), LS (10/50), and CP

(2/50) are obtained using the 20 ordinary ground motion records scaled to the

appropriate UBC design spectral values (1.1 g) and NEHRP guidelines {TCBO, 1997;

FEMA, 2000a, b]. The distribution for CP_NF is obtained using the six near-fault

records also identified as part of the CCWP. These near-fault records were not scaled

because insufficient knowledge exists at this time to scale near-fault records to return

period specific hazard levels [Krawinkler et al., 2000]. The FEMA 356 drift limits are

shown for comparison. The response variability clearly increases at higher hazard

levels. The evaluation of the CP (2/50) limit state is the subject of some debate, and

the focus of this study is on the LS (10/50) and JO (50/50) limit states.

4.2.1.2 Damping

Damping ratios for woodframe structures are often presumed to be in the range

of 2% to 8%. Fischer et al. (2001) performed shake-table tests of full-scale woodframe

structures as part of Task 1.1.1 of the CUREE-Caltech Woodframe Project (CCWP)

and found equivalent viscous damping ratios of 3.1% at ambient levels, increasing to

12% at PGA = 0.22g shaking, and decreasing to about 6% at PGA 0.5g shaking.

As part of CCWP Task 1.3.3 (Dynamic Characteristics of Woodframe

Structures), damping ratios were determined to be in the range of 2.6% to 17.3%, with

an average of 7.2% [Camelo et al., 2001]. However, much of this is likely to be

hysteretic damping, which is accounted for directly in the hysteretic model.

I

0.9

0.8

0.7

0.6

0.5LL

0.4

0.3

0.2

0.1

0

0

IO(50/50)

LS(10/50) --

I- KCPNF (2/50)

/ CP (2/50)

/

Bft.

,' / BWI (8' x 8'), 8d3"/6", OSB(3/8"),

ED = I8", G = 180 ksi, ç = 2%,

W=1400 lbs/ft (50 kN total)

2 4 6 8 10 12 14 16 18 20

6max (in.)

Figure 4.4 Response (peak displacement) variability for the three limit states

The damping parameter in the nonlinear time-history analysis program is the

nominal viscous damping value, as a percent of critical. This is why measured

damping values are often in excess of 15%, while most people assume about 2%

(viscous) damping in their models [Foliente, 1995]. In fact, the viscous damping is

thought by some to be much less, perhaps effectively zero, in woodframe shearwalls,

particularly at high peak ground acceleration (PGA) values. Some people use 0.1%

(for example) to avoid singularity problems in the analyses.

Discussion among Element 1 researchers in the CCWP suggested it may be

appropriate to use different damping values for different ground motion intensities.

For example, the full-scale building tests at University of California at San Diego

suggest viscous damping levels of about 7% for low-intensity shaking, and close to

zero for strong shaking. Using BW1 (and considering both JO, 50/50 and LS, 10/50),

three different approaches to the assignment of viscous damping are considered: (1)

constant damping, 0%, 1% and 2%, is assumed for all cases; (2) 7% is assumed for the

records scaled for 10 (50/50) and 0.1% is assumed for the records scaled for LS

(10/50) and (3) damping is assumed to vary linearly from 0.1% to 7%, inversely

proportional to the PGA of the scaled record. The effect of these damping assumptions

on the peak displacement distribution is shown in Figure 4.5 for the two different

performance levels. For the development of performance curves and design charts,

described later in this dissertation, a single damping value of 2% of critical is

assumed. However, based on the results shown in Figure 4.5, it may be conservative in

future studies to assume a lower value of damping for higher intensity ground motions

(i.e., records scaled for LS, 10/50).

31

0.9

0.8

0.7

1x

0.5LL

0.4

0.3

0.2

0.1

17//Z'

/f :LL"ç = 0% l0, 50/50) /,," = 0% (LS, 10/50)

= 1% (10, 50/50) //'/ = 0.1% (LS, 10/50)

= 2% (l$D, 5O/5O),i-c.: = 1% (LS, 10/50)

Variable (10, 5oI1'J ,,N Variable (LS, 10/50)

= 7% (10] 50/50)""2% (LS, 10/50)

1'I 1" I

I;,,;

Ii

01 iii

0 0.5

8ft

II 8ft.

BW1 (8' >< 8'), 8d@3"/6",

OSB(3/8"), G = 180 ksi,

W = 1400 lbs/ft (50 kN total)

1 1.5 2 2.5 3

max (in.)

Figure 4.5 Effects of viscous damping ratio () on peak displacement

4.2.1.3 Shear modulus of sheathing materials

The assigned shear modulus (G) for wood sheathing panels ranges from about

60 90 ksi (0.41 0.62 GPa) for plywood, and about 180 - 290 ksi (1.24 - 2.00 GPa)

for OSB [Plywood, 1998; Wood Handbook, 1999]. The shear modulus increases with

panel thickness. It was shown that the variability in shear modulus (G) contributes

very little to the variability in response (peak displacement), and hence, shear modulus

can be treated as a deterministic quantity. A sensitivity study can be performed to

evaluate the effect of shear modulus on peak displacement, however it must be

recognized that in addition to changing shear modulus, the change in thickness and

sheathing material will affect the fastener hysteretic parameters. Since only the

Durham (1998) nail data (see Section 4.2.2) is used for this part of study, only the

32

shear modulus is changed for the purposes of this comparison. Figure 4.6 shows the

effect of assigned shear modulus on the peak displacement distribution, considering

baseline wall BW1 and the life safety (LS, 10/50) limit state. The effect of shear

modulus, considering sheathing thickness varying from /8 /8 in., is seen to be

relatively small; however, to properly investigate its effect will require the appropriate

fastener hysteretic parameters. From this point forward, the shear modulus (G) is

assumed to be deterministic with values of 60 ksi for plywood and 180 ksi for OSB.

42.l.4 Fastener spacing

The number of fasteners is clearly one of the most significant factors affecting

shearwall performance under earthquake loading. The arrangement (spacing) of

fasteners also is important. These factors influence specific fastener failure modes

(i.e., which fasteners are worked hardest) as well as the overall energy dissipating

characteristics of the shearwall. Figures 4.7 through 4.10 show the effect of fastener

spacing on peak displacement with various assumed seismic weights, considering

BW1 and the life safety (LS, 1 0/5 0) limit state. A practical drift limit of 4 in. also is

shown on these figures. Based on the comparison of 6"/6" and 6"/12" nailing

schedules, field nailing schedule has little effect on the performance of shearwalls

compared to the effect of edge nailing schedule, particularly at higher values of

seismic weight (see Figures 4.7 through 4.10). Fastener spacing obviously is a

significant design parameter for woodframe shearwalis and will be treated as such in

the design charts developed in Section 4.3.2.

33

0.9

0.8

0.7

0.6

0.5U-

0.4

0.3

0.2

0.1

n

G 290

G=235ksL ';;/

G = 218 ksi

G=180ksN//

7/ ______

/1 ________8ff

//8ff.

BW1 (8 x 8), 8d3'I6",

/1 OSB(3/8"),ED= /8",0 W5OkN,ç=2%,

/[S (10/50)

0 0.5 1 1.5 2 2.5 3

(in)

Figure 4.6 Effect of assigned shear modulus (G) on peak displacement

S.

5;

0.7

0.6

0.5

0.4

0.3

0.2

0.1

It

8d@4"/l/

I I ' 8d6"I6"8d@3"/12" ... I /

I 8d@6"/12"

/' : //

0/

Li /I

i 0/I

1i 8ff.

/ I'>i /

/ ,'o / 8ff.

1 / / BW1 (8' x 8'), OSB (/"), ED =

/ ',/ G=180ksi,=2%,LS(10/50),

_J / 0 W = 560 lbs/ft (20 kN total)

0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6

ömax (in.)

Figure 4.7 Effect of fastener spacing on peak displacement (W = 560 lbs/fl)

34

0.9

0.8

0.7

0.6

0.5

0.4

0.3

0.2

0.1

n

8d©4"/12" / ,, I

8d@3flhlr

///8d@fi2"

I

/8ft._Oft.

/ / ,'// BW1 (8 8'), OSB (3/), ED =/8",

/ / ,/ G=l8Oksi,ç=2%,LS(10/50),/ W = 840 lbs/ft (30 kN total)

0 0.5 1 1.5 2 2.5 3 3.5 4

ömax (in.)

Figure 4.8 Effect of fastener spacing on peak displacement (W = 840 lbs/ft)

0.9

0.8

0.7

0.6

0.5U-

0.4

0.3

0.2

0.1

U

I --/ V -/ - -

8d@4"/12" I -. --

8d©3"/12"xr8d@6"/6"

/

11/

j / ,'/ 8ft.

/ I ,'gII V/ / ' BW1 (8' x 8'), OSB (/"), ED =,/ G = 180 ksi, = 2%, LS (10/50),

- W = 1120 lbs/ft (40 kN total)

0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5

3max (in.)

Figure 4.9 Effect of fastener spacing on peak displacement (W 1120 lbs/ft)

35

0.9

0.8

0.7

0.6

0.5U-

0.4

0.3

0.2

0.1

n

78d@4"/12"

/8d@3"/12"/ Th'-i,/'

/ '1I / 8d@6"/12"

I

' /// / ,,

/ /HI

// 1 __8ft.f,r / ,

BW1 (8' x 8'), OSB (/8"), ED /8,

/ / ,, / G 180 ksi, = 2%, LS (1 0/50),

- W = 1400 lbs/ft (50 kN total)

0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5

6max (in.)

Figure 4.10 Effect of fastener spacing on peak displacement (W = 1400 lbs/fl)

4.2.1.5 Panel layout

Figure 4.11 shows the effect of sheathing arrangement on the peak

displacement distribution. The horizontal panel arrangement (with blocking) performs

better than the vertical panel arrangement. Despite having more fasteners, BW4

performs the least well because of the additional discontinuity in the sheathing and

hence reduced overall rigidity of the wall.

4.2.1.6 Shake-table test walls

The procedure described in Section 3.1 also can be used to construct peak

displacement distributions for walls that have been tested on a shake-table. This can

serve a number of purposes, among them (1) to validate the CASHEW model, and (2)

36

to provide information on expected shearwall response under dynamic loading. The

full-scale walls considered in Task 1.1.1 (UCSD) and Task 1.1.2 (UC- Berkeley) were

analyzed using this procedure. These walls were designed and built specifically for

shake-table testing under specific ground motions (and with specific seismic weights).

Figure 4.12 shows the three walls with the panel layouts. Figure 4.13 shows the peak

displacement distributions determined using the CASHEW modeling procedure for the

east and west walls tested in Task 1.1.1 and the rear wall tested in Task 1.1.2 (for two

different seismic masses, the smaller value corresponding to a post-retrofit condition).

The performance of the Task 1.1.2 rear wall (with retrofit) exhibits significantly lower

peak displacements, in part due to the reduction in seismic weight.

1

0.5U-

0.4

0.3

0.2

0.1

0

0

BW4BW3BW1

, BW1 BW2

8ft. 88.

1//1 / 8ff. 8ff.7/ BW3 BW4

88. 8ft.

I/,,L 8d©3"16", OSB (I8"), ED =18',

G = 180 ksi, ç = 2%, LS(10150),


0.5 1 1.5 2 2.5 3 3.5 4

ömax (in.)

Figure 4.11 Effect of panel layout on peak displacement

I 03r\, I. I. I L.QL VVGIII

in

37

96n.

32in. 46in. 36in. 46in. 32in.

192 in.

Task 1.1.1 West Wall

I asK 1.1.2 Kear Wall (Norm Wan)

Figure 4.12 Task 1.1.1 and task 1.1.2 walls

0.9

0.8

0.7

O.5

0.4

0.3

0.2

0.1

C'

/ Task 1.l.1 west Wall -I W = 82(t) lsIft (58.4 kN tota9-4 ,V

/OSB, G 180 ksi,

/ task 1.1.1 East Wall/ / W = 856 lbs/ft (58.4 kN total)

__- / / OSB, G = 180 ksi, 8d@3"/12"Task 1.1. Rear Wll /w = ii od lbs/ft (1 7.8kN tl)PWD,GF6Oksl,8d 4I1,With Retfofit ,' / / Task 1.1.2 Rear Wall/ / W=22171bs/ft(315.SkNtotal)/ / PWD, G 60 ksi, 8d@4"/12"

/

-7"

-

ED /8, t3/,

= 2%, LS (10/50)

in.

0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 5.5 6

max (in.)

Figure 4.13 Peak displacement distributions for task 1.1.1 and task 1.1.2 walls

4.2.1.7 Missing fasteners

As a preliminary study to investigate the effect of construction tolerances

(errors), the effect of missing fasteners was investigated using the CASHEW modeling

procedure. Such issues of construction quality are thought by some to significantly

affect the performance of woodframe assemblies, particularly under dynamic loading

[Seible et al., 1999]. Using BW3, the effect of missing fasteners or fastener lines in

critical locations is investigated. The results are shown in Figures 4.14 through 4.16

for the three limit states (immediate occupancy, life safety and collapse prevention),

respectively. While certainly not a comprehensive study, the results in these figures

provide some indication of the relative importance of ensuring the design fastener

schedule on the shearwall performance. Notice that while the nail along the sole plate

has a significant effect on performance, the fact that overturning anchors are present

reduces the effect of missing nails since the sole plate nails only resist wall racking

forces. Further discussion of construction tolerance issues is provided in Section 4.2.4.

39

1

0.9

0.7

0.6

0.5I-I-

0.4

0.3

0.2

0.1

nt

d@3"16" [2M]

/

8c3 '/6" [4M]

8d3"/6" [3M]

8d©/6 MM]8d©3"/è"

/ 3/6" baseline wall nail patternElM] missing left side vertical nail line in S1

// /,' [2M] missing horizontal blocking at mid-height

[3M] missing every other nail along sole plate[4M] missing entire nail line along sole plate

Iii," 1s21s31

J / ,, (0 SI I

0)I

III

86.II

I BW3 (8' x 8'), 8d@3"/6", OSB (/8"),

/ '0 ED=3/8" G=180ksi,=2%, 10(50/50),


0 0.5 1 1.5 2

max (in.)

Figure 4.14 Effect of missing fasteners on peak displacement (JO, 50/50)

0.9

0.8

0.7

0.6

0.5U-

0.4

0.3

0.2

0.1

U

-- --------"-8d©3"/6" [2M]

-

8d@3"16" [4M} -,

8d@3"/6" [3M]

8d@3"/6" MM] /

8d@3"16" k, I / L"/6" baseline wall nail pattern[1 M] missing left side vertical nail line in S1

I[3M] missing every other nail along sole plateI [2M] missing horizontal blocking at mid-height

missing entire nail line along sole plate

I s Is3 I

I

186.

I//'ci

I

I Si I

I I

811.

BW3 (8' < 8'), 8d@3'/6", OSB (3/s),

0ED = /8", G = 180 ksi, = 2%, LS (10/50),W = 1400 lbs/ft (50 kN total)

0 0.5 1 1.5 2 2.5 3 3.5 4

ömax (in.)

Figure 4.15 Effect of missing fasteners on peak displacement (LS, 10/50)

1

0.9

0.8

0.7

0.6

0.5U-

0.4

0.3

0.2

0.1

8d@3"16" [2M/

8d©3"/6" [4M]

8d@3"/6" [1M] /

8d©3"16" /8d©3"/6" [3M] / 36",taseline wall nail pattern/ ,,11M missing left side vertical nail line in S1

/ r[2MJ missing horizontal blocking at mid-height

/ J1M] missing every other nail along sole plate/ / , [4M] missing entire nail line along sole plate

/,',' S2S3/ // 8ft.S1

/// / ' BW3 (8 > 8), 8d@3"16", OSB (3/),

/ / - ED = G = 180 ksi, ç = 2%, CP (2/50),W 1400 lbs/ft (50 kN total)

0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 5.5 6

max (in.)

Figure 4.16 Effect of missing fasteners on peak displacement (CP, 2/50)

4.2.1.8 Model uncertainty

The numerical model (CASHEW) contributes to the epistemic uncertainty in

the analysis. This uncertainty arises from modeling assumptions and simplifications,

either idealizations or approximations. One example is the assumption in CASHEW of

rigid tie-downs. Model uncertainty should therefore account for variations in (or

perhaps lack of) anchorage. While it is expected that this uncertainty and its effect on

dynamic response would be greater than some of the physical parameters such as

fastener hysteretic parameters and sheathing properties, it is not obvious how it would

compare to the effect of variability in the ground motions (seismic hazard). Model

uncertainty often is taken into account using an error term (s), such that X* =

where X = model predicted response (a random variable) and X = the random

41

response taking into account model uncertainty. The model error term, s, can be

modeled as a random variable. (If both and X are lognormal variables, then X also is

lognormal.)

In order to determine moments of the model error term, it is necessary to have

a series of full-scale test results to which model predictions can be compared. (Note

that additional uncertainty is introduced in moving from actual field construction to

laboratory test conditions). Only limited full-scale test results are presently available

for which direct comparisons can be made to model predictions using CASHEW.

However, since the Durham fastener parameters for the spiral nail (see Section 4.2.2)

were used in this part of the study, the model predictions obtained using CASHEW

can be compared with the full-scale wall tests conducted by Durham (1998). Folz et al.

(2001) presented a comparison of these results considering peak displacement under

both a cyclic loading protocol and the Landers earthquake. The results suggest an

"error" term having a mean of about 1.0 and a COV between 15% and 30%,

depending on how many peaks in the Landers analysis are used in the comparison

[Rosowsky and Kim, 2002]. Lognormal parameters (2w, ) can be determined using the

method of moments.

To examine the effect of model uncertainty on the dynamic response, one

could compare peak displacement distributions, as done previously. The relative

contribution of the model error also could be evaluated by considering its effect on the

peak displacement distribution parameters. Since a lognormal distribution is fit to the

peak displacements obtained using a suite of 20 scaled ordinary ground motion

42

records, the modified distribution, taking into account the model error term, is given

by F (4= s x F (4. The lognormal parameters (2, ) for the modified distribution

F; (x) are given by:

= + (4.1)

(4.2)

Figure 4.17 shows the effect of model uncertainty on the peak displacement

distribution for BW1, for LS (10/50) and Figures 4.18 through 4.22 show the effect of

including these error terms on the peak displacement distributions for various nail

spacings and two seismic weights. The effect of including the model uncertainty

(error) diminishes as the response variability increases, e.g., for large nail spacings

(see Figures 4.21 and 4.22). In that case, the response variability is completely

dominated by the variability in the seismic hazard (i.e., the suite of ground motion

records). The model error may be significant, however, for peak displacement

distributions having less variability, such as those with tighter nail spacings. Note that

these steeper distributions (lower variability) are typically seen at lower values of peak

displacement, i.e., for walls that would meet the FEMA 356 peak drift criteria with

high probability. In these cases, the uncertainty introduced by model error may be on

the same order of magnitude as the response variability arising from the suite of

ground motions.

The effect of model error associated with the CASHEW program may be

significant. The model error also may not be uniform over all displacement ranges

(i.e., degree of nonlinearity). This is a potentially significant source of uncertainty and

43

should be studied further, prior to the development of final design recommendations.

This will require additional comparisons between full-scale tests and CASHEW model

predictions. Results from tests such as those conducted at UCSD (Task 1.3.1) and UC-

Irvine (Task 1.4.4) could be useful in this regard. Consideration also could be given to

the differences between laboratory tests and actual field conditions. This also could be

taken into account through a model error term.

1

S.

0.8

0.7

[SIr

0.5U-

0.4

0.2

0.1

8d@3"/6" (COV=0%)

fN 8d@3"16" (COV=15%)

8d@3"16" (COV=30%)

8ff.

BW (8 x 8'), 8d@36", OSB (3/.),

ED = /8, G = 180 ksi, = 2%,

W = 1400 lbs/ft (50 kN total), LS (10/50)

0 0.5 1 1.5 2 2.5 3 3.5 4

ömax (in.)

Figure 4.17 Effect of model uncertainty on peak displacement distribution

0.9

0.8

0.7

0.6

LL

0.4

0.3

0.2

0.1

n

:IT2lbS/ft (40 kN total)

8d3"/6" (COV=0%)8d©3"/6' (COV=15%)8d@3"/6" (COV=30%)

W = 560 lbs/ft (20 lN total)8d@3"/G" (COV=0°(o)8d@3"/6" (COV=15%)8d@3"/6" (COV=30%)

811.

C)

811.

OW (8'>< 8'), 8d@3'I6",

OSB(3/e"),ED=3/8",=2%,G = 180 ksi, LS (10150)

0 0.5 1 1.5 2 2.5 3

max (in.)

Figure 4.18 Effect of model uncertainty on peak displacement (3"/6")

0.9

0.8

0.7

0.6

0.5

0.4

0.3

0.2

0.1

0

1/" //YNI; II' W = 1120 lbs/ft (40 kN total)

8d@3"/1" (COV=0%)'N //' 8d@3"/1" (COV=15%)

N.J' 8dQ3"/1"

,TN W = 560 lbs/ft (20 N total)'J 8d@3"/12" (COV=Ø%)

"I 8d©3"/12" (COV=15%)8d@3"/12" (COV=0%)

_8ft.,' /1 OW (8' x 8'), 8d@3"112",

' / OSB(3/8"),ED=3/8",=2%,

- = 180 ksi, LS (10/50)

0 0.5 1 1.5 2 2.5 3

max (in.)


45

0.9

0.8

0.7

0.6

o.5

0.4

0.3

0.2

0.1

1)

1'W=ll2Olbs/ft(4okNtotal)

,//' 8d©4"/12"(COV=0%)II: /' 8d@4"/12" (C0V15%)

8d@4"/12" (C0V30%)

W = 560 IbIft(20 kN1 total)8d@4"/l 2'COV=0°/)8d©4"/12Y(COV=15%)

8d@4"f (COV=30/0)

:1j

8ft.

/ BW (8' x 8), 8d©4"/12", OSB (/8")

ED = /8, G = 180 ks = 2%, LS (1 0/50)

0 0.5 1 1.5 2 2.5 3 3.5 4

max (in.)


0.9

0.8

0.7

0.6

0.5

0.4

0.3

0.2

0.1

r)

W=5601bs/ft(20kNtotal)(// 8d@6"I6" (COV=0%)

8d©6"/6" (COV=15%)1120 lbs/ft (40 kN total)8d@6"/6"

8L

811.

BW (8' x 8'), 8d@6"/6", OSB (3/),

ED = /8", G = 180 ksi, = 2°I, LS (10/50)

0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5

6max (in.)

Figure 4.21 Effect of model uncertainty on peak displacement (6"/o")

0.9

0.8

0.7

0.6

0.5

0.4

0.3

0.2

0.1

n

- -

- -

'VV = 560 lbs/ft (20 kN total)

/,' 8d@6"12" (COV=0%)

8d@6"/12"

8d©6"12" (COW39%'o) W = 1120 lbs/ft (40 kN total)

/ 8d@6"/12" (COV=0%)

/ 8d@6"/12" (COV=15%)8d©6"/12" (COV=30%)

HY8ft.1

BW (8 x 8), 8d©6"/12", OSE (/8"),

ED = '8, G = 180 ks, = 2%, LS (10/50)

0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 5.5 6

max (in.)


47

4.2.2 Sheathing-to-framing connection hysteretic parameter variability

As part of Task 1.4.8.1 of the CUREE-Caltech Woodframe Project (CCWP), a

database of sheathing-to-framing connection hysteretic properties was developed for a

wide range of fastener types, sheathing materials, and boundary conditions. Once

validated, this database can provide valuable information for the development of

performance curves and design charts (discussed later in Section 4.3.2) for a broad

range of structural types. For the purposes of this study, however, the fastener

parameters (see Table 4.1) provided by Durham (1998) for an spiral nail and by Folz

(2001) for an 8d box nail were used. (Additional parameters from Dolan (1989) also

will be used later.)

The Durham parameters were obtained from tests of 3/8-In. OSB sheathing,

having an assigned shear modulus of 218 ksi, attached to nominal 2 in. x 4 in. framing

members with pneumatically-driven 2 in. long spiral nails. The Folz parameters were

obtained from tests of 3/8-in. OSB sheathing, attached to nominal 2 in. x 4 in. framing

members with pneumatically-driven 8d box gun nails. Table 4.1 shows the sheathing-

to-framing connection hysteretic parameters obtained from the experimental studies by

Durham (1998) and Folz (2001). Further information on these hysteretic model

parameters may be found elsewhere [Durham, 1998; Fischer et al., 2001; Folz and

Filiatrault, 2000, 2001].

Nail K0 r1 r2 r3 r4 F0 F1J

a

DurhamW spiral 3.203I 0.061 -0.078 1.40 0.143

0.169 I 0.032 0.4920.8 1.1kips/in

I I I kips kips I in.I

Folz° 8d box 4.870.049 -0.049 1.40 0.015

0.180 I 0.042 I 0.500.8 1.1

Igun

Ikips/in kips kips in.

Table 4.1 Sheathing-to-framing connection hysteretic parameters

(1)

Values obtained by Durham (1998). Fasteners were 2 in. long, power-driven spiral nails attaching i-in. OSB to framing members.(2) Values obtained by Folz (2001). Fasteners were 8d box gun nails attaching 3/8-in. OSB to framingmembers.(3) Protocol did not include cyclic behavior near the ultimate capacity of the wall, resulting inunrealistically high values of r4 compared to other studies. Therefore, the value ofr4 was changed in thisstudy to 0.05 (see Folz and Filiatrault, 2000).

Selected data from other studies were considered in order to investigate the

contribution of fastener parameter variability on performance (peak displacement) of

shearwalls. First, a comparison was made between results obtained using the Durham

and Folz nail data and comparable sets of parameters developed in the CCWP Task

1.4.8.1 (see Table 4.2). Fonseca et al. (2002) obtained sheathing-to-framing

connection hysteretic parameters for numerous connection types and compiled a

database. This database was used in the sensitivity studies to investigate the variability

of sheathing-to-framing connection hysteretic parameters. Several parameters were

considered in that study such as sheathing types and nail types, sheathing panel

direction (perpendicular and parallel), edge distances, and the effect of overdriven

nails; however only Douglas Fir-Larch (DF-L) framing lumber was used. Testing was

conducted using the simplified basic loading history developed in Task 1.3.2 of

CCWP [Krawinkler et al., 2000]. Ten specimens were tested for each combination of

parameters. Sampling was done at a rate of 20 points per second, and ten hysteretic

parameters (for use in CASHEW) were determined for each specimen tested. Also

shown in this table are the parameters obtained by Dolan (1989) for a comparable

plywood product.

Durham' Folz2 Task 1.4.8.1 (Fonseca et. al.,)3 Dolan4Institutions IJBC IJCSD BYU UBC

Parameter Units /8" OSB /8" OSB '8 OSB3/8"OSB 3/8"OSB

Plywood

K0lups

'In 3.2034 4.8700 2.9746 3.5898 4.0341 5.1791

r1 0,0610 0,0490 0.0740 0.1099 0.1220 0.0496r2 -0.0780 -0.0490 -0.0774 -0.1459 -0.0753 0.0595r3 1.4000 1.4000 2.4933 1.6240 1.3495 1.4000r4 0.1430 0.0150 0.0724 0.1363 0.1334 0.0265r4 0.0500 0.0150 0.0724 0.0500 0.0700 0.0265F0 jp_ 0.1688 0.1800 0.1344 0.1229 0.1318 0.2271F1 kips 0.0317 0.0420 0.0418 0.0431 0.0442 0.0409

in 0.4921 0.5000 0.2502 0.1385 0.1573 0.3150a 0.8000 0.8000 0.6000 0.6000 0.6000 0.8000

1.1000 1.1000 1.1000 1.1000 1.1000 1.1000

Table 4.2 Comparable connection hysteretic parameters from other studies

Values obtained by Durham (1998). Fasteners were 2 in. long, power-driven spiral nails attaching i-in. OSB to SPF framing members.2)

Values obtained by Folz (2001). Fasteners were 8d box gun nails attaching 3/8-in. OSB to framingmembers.3)

Fastener hysteretic parameters (average values shown) obtained by Fonseca (2001). Fasteners were8d common nails attaching 3/3-in. OSB to DFL framing members. Loading was perpendicular to thegrain. Three different OSB manufacturers were considered.(4)

Fastener hysteretic parameters (average values shown) determined using results obtained by Dolan(1989). Fasteners were 8d common nails attaching 3/g-in. plywood to SPF framing member, 3/8-in, edgedistance, loading perpendicular to grain.

Values of r4 changed per Note 3 in Table 4.1.

Figures 4.23 through 4.26 present a comparison of peak displacement

distributions for the Durham OSB data set (spiral nail), the Folz OSB data set (8d box

gun nail) and the three BYU (Task 1.4.8.1) OSB data sets (8d cooler nail) with various

assumed seismic weights and considering the life safety (10/50) hazard level. The

baseline 8 ft. x 8 ft. solid wall (BW1) with two sheathing panels oriented vertically

and a 3"/l 2" (edge/field) fastener schedule was considered. The sample distribution

functions in these figures provide some indication as to the relative sensitivity of

50

results to assumed fastener parameters, which increases dramatically for larger

demands (seismic weights). The peak displacement curve developed using the Durham

nail parameters (spiral nail) generally is close to the median of the peak displacement

distributions throughout the range of seismic weights considered. This median peak

displacement curve will be used to develop a modification factor for sheathing-to-

framing connection hysteretic parameter variability in Section 4.3.1.1.

1

I

0.5

0.4

0.3

0.2

0.1

ii

/,8d cooler nail_3 ' /'/

8d box nail /

,, /' /

/: / / /

!,," // /

// // /

/,," // /

,/'//

8d cooler nail_i8d cooler nail 28d spiral nail

8ft

8 ft.

BW(8' x 8'), 8d©3"/12", OSB (/"),ED318",G l8Oksi,ç2%,W = 560 lbs/ft (20 kN total), LS (10/50)

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1

6max (in.)

Figure 4.23 Comparison of peak displacement distributions for different nailparameters (W = 560 lbs/ft)

51

0.9

0.8

0.7

0.6

0.5LL

0.4

0.3

0.2

0.1

n

-

/7,', ,7'

8d cooler nail 3 / ,' 8d cooler nail_i/ / 8d cooler nail 2

8d box nail / // 8d spiral nail

,f/ / //

/ //

': ///

!' / // 8ft.

18ft

/,' / / BW (8 x 8'), 8d@3"/12", OSB (/"),/ // ED=3/8",G=l8oksi,ç=2%,

W = 840 ibs/ft (30 kN totai), [S (10/50)

0 0.3 0.6 0.9 1.2 1.5

max (in.)

Figure 4.24 Comparison of peak displacement distributions for different nailparameters (W = 840 lbs/ft)

0.9

0.8

0.7

0.6

0.5U-

0.4

0.3

0.2

0.1

n

8d spiral nail /,7 /8d cooler nail 3 //

8d box nail/7 //

H,/

/ /

/,, /I//

8d cooler nail_i

8d cooler nail 2

8ft.

BW (8 x 8), 8d3"/12",OSB(3/8"),ED= /8',G=180ksi,=2%,W = 1120 ibs/ft (40 kN totai),LS(10/50

0 0.5 1 1.5 2 2.5 3

6max (in.)

Figure 4.25 Comparison of peak displacement distributions for different nailparameters (W = 1120 lb s/ft)

52

1

0.9

0.8

0.7

0.5U-

0.4

0.3

0.2

0.1

8d cooler nail 3

8d box nail

I

/

,,///

8d cooler naiL 2

8d spiral nail

8L.

8ft.

BW1 8' x 8'), 8d@3'/12", OSB (/"),

ED = /", G = 180 ksi. LS (10/50),

W1400 lbs/ft (50 kN total), = 2%

0.5 1 1.5 2 2.5

max (in.)3 3.5 4

Figure 4.26 Comparison of peak displacement distributions for different nailparameters (1N = 1400 lb s/ft)

Next, variability among parameter sets was investigated. Since many of the

fastener hysteretic parameters are strongly correlated, it is not possible to explicitly

consider the effect of individual parameter variability on the predicted shearwall

response. Instead, the results from the individual connection tests are considered as

sets of parameters. That is, rather than using the average values (obtained by averaging

the results from 10 individual tests per series), the hysteretic parameters fit to each

individual connection specimen are considered. The peak displacement distribution

results can be compared to those obtained using the average values, thereby providing

some indication of the relative contribution of fastener parameter variability. This was

done using the 3/g-in. OSB MFG1 data set from Task 1.4.8.1 (see Table 4.2). The

results are shown in Figure 4.27 for the LS (10/50) limit state, and assuming 4496 lbs,

53

6744 ibs, and 8992 lbs seismic weights. Note that only five curves are shown (in

addition to curve obtained using the average values) for each seismic weight.

CASHEW could not provide a convergent solution for the other five sets of

parameters. Whether this suggests a lack of robustness of the CASHEW program or a

problem fitting the test data obtained in Task 1.4.8.1 (there are still questions

regarding edge distance effects in this data, for example) remains to be determined.

Still, the limited results shown in Figure 4.27 provide some insight into the relative

contribution of fastener parameter variability to the predicted response. The peak

displacement distribution could be modified by a factor to account for fastener

parameter variability. For example, one could apply a factor to the logarithmic mean

value, denoted by 2, to adjust the median response. Alternatively, a positive factor

could be applied to the logarithmic standard deviation, denoted by E, to adjust the

uncertainty in the response. This is conceptually similar to the treatment of model

error, which is discussed in Section 4.2.1.8.

For the remainder of this study, the sheathing-to-framing connection hysteretic

parameter sets obtained by Durham and Folz for 3/8-in. OSB, and determined from

data obtained by Dolan for 3/8-in. plywood, are used without further consideration of

fastener parameter variability. (Note that the fasteners are different; see Table 4.3.)

54

Nail type Test location (investigator) Length (in.) Shank diameter (in.)8d box gun nail UCSD (Folz) 21/2 0.113

Spiral nail UBC (Durham) 2 0.1058d cooler nail BYU (Fonseca et al.) 2/ 0.113

8d common nail UBC(Dolan) 2'/2 0.131

'lable 4.3 Nail properties considered in this study

1

0.9

0.8

0.7

0.5U-

0.4

0.3

0.2

0.1

n

[ //(W=1120 lbs/ft

I//I

W= 840 lbs/ft(30 total)

/1! W = 560 lbJ/ft.(20 kN

total?

/

I//I '. 8 ft.

88.

J /,.'BW1 (8 >< 8), 8d@3 /6, OSB (/8

CN ED=318",G=l8Oksi,1=2%,LS(1O/50)

0 0.5 1 1.5 2 2.5 3 3.5 4

max (in.)

Figure 4.27 Effect of fastener parameter variability on peak displacement

4.2.3 Contribution of nonstructural finish materials

Woodframe structures are built with a wide variety of architectural finishes on

the walls. Two of the most common wall finishes in woodframe structures are gypsum

wallboard and stucco (Portland cement plaster). Modern structures usually rely upon

OSB or plywood shearwalls for lateral strength. They are seldom the final surface of

the wall and are usually covered with either gypsum wallboard or stucco (Portland

cement), for both appearance and fire resistance [McMullin and Merrick, 2001].

Finish materials such as stucco or gypsum wallboard usually are not

considered to have significant structural capacities and thus are neglected in design.

However, results from both the shearwall tests and the full-scale shake-table tests

conducted as part of Element 1 of the CUREE-Caltech Woodframe Project (CCWP)

suggest the presence of stucco, albeit under fairly ideal conditions (i.e., well applied,

uncracked, undamaged by moisture or other environmental actions), may be beneficial

from the standpoint of performance (drift). In the case of shearwalls, a well applied

stucco layer has the effect of making the sheathing panels perform as a single rigid

body. In the case of the full-scale structures, the stucco has the additional effect of

providing shell action around corners. The result in both cases is substantially reduced

drifts. Recent studies also indicate that the presence of finish materials in shearwalls

decreases deflection capacity and increases strength and initial stiffness [Gatto and

Uang, 2002]. Also, stucco applied to the sheathing panel appears to restrain sheathing

nail withdrawal and partially restrain nail head rotation [Cobeen, 2001]. This further

suggests the effect of finish materials such as gypsum wallboard and stucco may

56

indeed be significant and should be considered in developing performance-based

design guidelines. A cross-section of a typical wood shearwall with nonstructural

finish (NSF) materials is shown in Figure 4.28.

Figure 4.28 Typical exterior wall cross-section

Gypsum wallboard

Framing member

OSB or Plywood

Stucco

It is difficult to use CASHEW to account directly for the behavior of

nonstructural finish materials such as stucco and gypsum wallboard since it may not

be possible to determine a particular nailing schedule and shear modulus that can

capture the performance of the nonstructural finish materials. Therefore, the results of

three recent experimental tests of wall with nonstructural finish materials were used to

investigate this issue using SASHFJT rather than the CASHEW modeling procedure.

The three experimental testing programs were taken from: (1) CCWP Task 1.3.1, (2)

CCWP Task 1.4.4, and (3) the CoLA test program [Gatto and Uang, 2002; Pardoen

57

et.al., 2001, 2002]. These tests results were used to capture the global shearwall

hysteretic parameters using the visual best-fit program SASHFIT. The material

combinations and test programs considered are summarized in Table 4.4.

Project Sheathing materials + NSF Shearwall Loadingprotocol

CUREE1.3.1

OSB (3/)

solid wall(8 ft. x 8 ft.)

CUREE

OSB (I8") + GWB ('/2")

OSB (/8") + Stucco (/8")

PWD (15/)

PWD (15/) + GWB ('/")PWD (15/) + Stucco (/8")

CUREE

(/8") wall with garagedoor opening(16 ft. x 8 ft.)

_OSB

OSB (/8")+ Stucco ("8")

OSB (3/8")+GWB ('/2") + Stucco (/8")1.4.4 OSB (3/) wall with pedestrian

door opening(16 ft. x 8 ft.)

OSB (/8") + Stucco (/8")

OSB (/8") + GWB ('/2") + Stucco (/8")

CoLA

3 ,'PWD ( ) solid wall(8 ft. x 8 ft.)

SPD'PWD (/8") + GWB ('/2")

PWD (/8") + GWB ('/2", 2 sides)Table 4.4 Matrix of walls used to investigate nonstructural finish material effects

(1)

Sequential phased displacement loading protocol

4.2.3.1 Analysis of solid wall

Results from two experimental test programs (CUREE Task 1.3.1 and CoLA)

were used to investigate the performance of an 8 ft. x 8 ft. solid shearwall with

nonstructural finish materials. The test programs considered the same shearwall

configuration, nailing schedule, and sheathing materials. The test data were obtained

from the CUREE (Task 1.3.1) and CoLA testing program. Each shearwall

configuration was tested two or three times in these programs. Only the worst case

results for each shearwall configuration were selected to study NSF materials effects

on the performance of the isolated solid baseline shearwall, BW1 (8 ft. x 8 ft.). Two

sheathing material types, 3/g-in. OSB and 15/32-in. plywood, were used in Task 1.3.1

and 3/8-In. plywood was used in the CoLA tests. The fastener schedules were 4"112"

(edges/field) with a double row at the end studs for the Task 1.3.1 and 4"/12" for the

CoLA tests. Both tests had the same thickness of nonstructural materials: Y2-in.

gypsum wallboard and 7/8-in, stucco. The material properties of the stucco and gypsum

wallboard are found elsewhere [Gatto and Uang, 2002; SEAOSC, 2001].

The peak displacement curves showing the effects of NSF materials are shown

in Figures 4.29 through 4.32 for various assumed seismic weights. As expected, NSF

materials greatly enhance the performance of shearwalls. In particular, the presence of

stucco serves to greatly reduce peak wall displacement. Figures 4.33 and 4.34 also

show the effect of two-sided gypsum wallboard on shearwall behavior. The use of

gypsum wallboard on both sides of the wall, as is done for interior partition walls, is

considerably more effective than one-sided GWB.

59

0.9

0.8

0.7

0.6

0.5LL

0.4

0.3

0.2

0.1

PWD+Stucco1 , OSBOSB+GWB ,' PWD

OSB + Stucco --/ / PWD + GWB

I // /

/./

/ BW (8' 8', 8d@4"/12", 2%,

/ / OSB (/"), PWD (15/..) GWB (/2'), Stucco (I8"),_,' W = 560 lbs/ft (20 kN total), LS (10/50)

0 0.2 0.4 0.6 0.8

max (in.)

Figure 4.29 Effect of nonstructural finish materials on peak displacement (W 560lbs/fl)

0.9

0.8

0.7

0.6

0.5

0.4

0.3

0.2

0.1

n

: -----

n:::::*EX

,:'T T,/<: + GWB

OSB + Stucco OSB

// / ,' :,/

/1 8ft.

1/

,/ / 8ft.

/ / / BW (8' x 8'), 8d@4"/12", = 2%,

',, ,'/ / OSB (3/), PWD (/32"), GWB (l/2) Stucco (7/),

W = 840 lbs/ft (30 kN total), LS (1 0/50)

0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6

6max (in.)

Figure 4.30 Effect of nonstructural finish materials on peak displacement (W=1 840lhs/ft)

0.9

0.8

0.7

0.6

0.5

0.4

0.3

0.2

0.1

it

OSB+GWB- ..,

PWD + Stucco

OSB + Stucco

1//i ;':/°

4I

*

PWDGWBPWD

OSB

8ft

8ft.

BW (8 x 8), 8d@4"/12",OSB (/8"), PWD (15/32),GWB (/2"), Stucco (/8"),

= 2%, LS (10/50),W = 1120 lbs/ft (40 kN total)

0 0.5 1 1.5 2 2.5 3

ömax (in.)

Figure 4.31 Effect of nonstructural finish materials on peak displacement (W= 1120lbs/ft)

0.9

0.8

0.7

0.6

0.5

0.4

0.3

0.2

0.1

It

OSB+GWB/ ,,'PWD+Stucco 4'-' PWD+GWB

OSB+Stucco' k

//,,,,, //8 ft.

I, '/ // BW (8 x 8), 8d@4'112", = 2%,, LS (10/50)

I, /,,./ OSB (3/), PWD (15/32W), GWB (1/2)Stucco (7/), W = 1400 lbs/ft (50 kN total)

0 0.5 1 1.5 2 2.5 3 3.5 4

6max (in.)


61

1

0.9

0.8

0.7

0.6

0.5

0.4

0.3

0.2

0.1

C)

:/ PWD

7PWD+GWB

I. PWD GWB (both sides)

7 / SW (8' x 8), 8d@4"/1 2", = 2%,/ PWD (/"), GWB (1/2), LS (10/50)

_J' ,' W = 840 lbs/ft (30 kN total)

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1

ömax (in.)


0.9

0.8

0.7

0.6

0.5U-

0.4

0.3

0.2

0.1

1)

/PWDPWD+WB

/ PWD + GWB (both sides)

// 80.' / BW (8' x 8'), 8d@4'/12",// )=2%, LS(10/50)


0 0.5 1 1.5 2 2.5 3

max (in.)


62

4.2.3.2 Analysis of walls with openings

Two types of shearwalls with openings were considered, both 8 ft. high x 16 ft.

long, one having a garage door opening and the other having a pedestrian door

opening. Both walls had vertically oriented sheathing at the ends and a solid header

over the opening, and were tested under Task 1.4.4 of the CUREE-Caltech

Woodframe Project (CCWP). As before, only the worst case results for each shearwall

configuration were selected from the CUREE Task 1.4.4 experimental test results to

investigate the contributions of NSF materials for walls with openings. The walls had

3/8-in. OSB, ½-in, gypsum wallboard, and 7/8-in. stucco. The nailing schedules

(edge/field) were 3"/12" for the wall with the garage door opening and 6"/12" for the

wall with the pedestrian door opening. The material properties of the stucco and

gypsum wallboard are given by Pardoen et al. (2003).

Figures 4.35 and 4.36 show the NSF material effects on the performance of the

wall with the large garage door opening for various assumed seismic weights. As with

the solid wall, NSF materials are seen to contribute significantly to shearwall

performance. The effects of NSF materials on peak displacement of the wall with the

pedestrian door opening are shown in Figures 4.37 and 4.38. As was seen in the wall

with the large garage door opening, NSF materials serve to reduce the peak wall

displacement.

63

0.9

0.8

0.7

0.6

0.5U-

0.4

0.3

0.2

0.1

U

/*/

OSB /x i/i/

OSB + Stucco

OSB + GWB + Stucco // ,/

I

H H8ft

OW(i36X8adq/122W = 281 lbs/ft (20 kN total),LS (10/50)

0 0.2 0.4 0.6 0.8 1

max (in.)


0.9

0.8

0.7

0.6

0.5

0.4

0.3

0.2

0.1

(1

OSB

OSB + Stucco /(OSB + GWB + Stucco

A' /H16 ft.

H8 ft.

- OW (16 x 8'), 8d@3"/12",=2%,LS(10I50)

OSB (/8"), GWB (1/2)

Stucco (/8),


0 0.5 1 1.5 2 2.5 3

max (in.)


0.9

0.8

0.7

0.6

0.5U-

0.4

0.3

0.2

0.1

('I

OSB+Stucco / OSB,'QSB + GWB + Stucc

/+

/I' OW (16' x 8'), 8d6"/12", ç = 2%,

/ OSS (/"), GWB (/2), Stucco (/8"),

W = 703 lbs/ft (50 kN total),LS (10/50)

0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.66max (in.)


0.9

0.8

0.7

0.6

0.5U-

0.4

0.3

0.2

0.1

n

OSB + Stucco/bSB + GWB

OSB

H H H8ft.

16ff.

OW (16' x 8'), 8d6"/12", ç = 2%,

OSB (/8"), GWB (/2"), Stucco (/"),W = 984 lbs/ft (70 kN total),LS (10/50)

0 0.5 1 1.5 2 2.5 3 3.5 4

ömax (in.)

Figure 4.38 Effect of nonstructural finish materials on peak displacement (W 984lbs/ft)

65

4.2.4 Construction quality

While construction quality issues (specifically, missing fasteners) were

addressed in a very cursory way in the sensitivity studies in Section 4.2.1.7, there are a

number of other construction quality issues which could significantly influence overall

shearwall behavior. Among these are missing or misplaced fasteners and anchors,

deterioration of structural and nonstructural finish materials, improper selection of

fasteners, under-driven or over-driven fasteners, missing blocking, the use of smaller

panel segments, cutouts in framing members (e.g., for installation of conduit), and so

forth. Wood structures may deteriorate with time. In addition to natural aging, walls in

woodframe structures may be subject to severe environmental conditions such as

moisture absorption and fungus attack. A number of such durability issues could

significantly impact the dynamic behavior of fasteners and woodframe assemblies.

Isoda et al. (2002) developed numerical models for deterministic nonlinear

time-history analyses of four index woodframe buildings (small house, large house,

small town house, and apartment building). The required input parameters for the

shearwalls in the index buildings were developed using the CASHEW program and

available experimental test data [Folz and Filiatrault, 2000; SEAOSC, 2001]. The

walls in the four woodframe buildings included nonstructural finish materials such as

gypsum wallboard and stucco. Three categories of construction quality were

considered for each of the index woodframe buildings: superior quality, typical

quality, and poor quality. These are described in Table 4.5. A nonlinear dynamic time

history analysis was conducted to investigate the effects of construction quality using

the global hysteretic parameters and the three construction quality categories.

Superior Quality Typical Quality Poor QualityGood nailing of diaphragms. Good nailing of diaphragms. Poor nailing of100% of stiffness and strength 90% of stiffness and strength diaphragms.from high-quality laboratory from high-quality laboratory 80% of stiffness andtests. tests, strength from high-quality

laboratory tests.Good nailing of shearwalls. Average nailing of shearwalls. Poor nailing of shearwalls.100% of stiffness and strength 5% greater nail spacing. 20% greater nail spacing.from high-quality laboratory 5% reduction stiffness andtests. strength due to water

damage.Good connections between Typical connections between Poor connections betweenstructural elements, structural elements. structural elements.100% of stiffness and strength 10% reduction of stiffness and 20% reduction of stiffnessfrom high-quality laboratory strength in shearwalls from high- and strength in shearwallstests. quality laboratory tests. from high-quality

laboratory tests.Good quality stucco. Average quality stucco. Poor quality stucco.100% of stiffness and strength 90% of stiffness and strength 70% of stiffness andfrom high-quality laboratory from high-quality laboratory strength from high-qualitytests. tests, laboratory tests.Superior nailing of interior Good nailing of interior gypsum Poor nailing of interiorgypsum wallboard. wallboard, gypsum wallboard.100% of stiffness and strength 85% of stiffness and strength 75% of stiffness andfrom high-quality laboratory from high-quality laboratory strength from high-qualitytests. tests, laboratory tests.

Table 4.5 Definitions of three construction quality categories (from: Isoda et al., 2002)

The global hysteretic parameters developed to correspond to each of the cases

are similar in form to those considered in the nonlinear time history analyses

performed in the CCWP Task 1.5.3 and described earlier in this dissertation. A

sensitivity study was performed to investigate the construction quality on performance

of shearwalls. Deterministic modification factors (relating back to superior quality)

were then developed for each hysteretic parameter. Table 4.6 shows the deterministic

modification factors for the different construction quality levels. These factors are

67

used to modify the original hysteretic parameters, assumed to correspond to superior

quality, to obtain the global hysteretic parameters of shearwalls of poor or typical

construction quality. Complete results of this sensitivity study are presented in

Appendix B.

Quality Sheathing K0 r1 r2 r3]

F0 F F aTYP. OSB 0.86 0.99 0.99 1.00 0.85 0.85 0.85 1.00 1.00

OSB+NSF 0.87 0.99 0.99 1.00 0.86 0.86 0.87 1.00 1.00OSB 0.61 0.99 0.97 1.01 0.63 0.63 0.61 1.00 1.00

POOR OSB+NSF 0.66 1.00 0.98 1.00 0.67 0.66 0.66 1.00 1.00OSB + NSF

(GWB)0.69 1.00 0.98 1.00 0.69 0.69 0.69 1.00 1.00

lable 4.6 Developed deterministic modification factor for construction quality

(1)r4 is assumed to be 1.00 for all cases.

The peak displacement curves were constructed using the modified global

hysteretic parameters (i.e, global hysteretic parameters obtained from experimental

test x modification factor) and the CCWP Task 1.3.1 shearwall test results. Various

assumed seismic weights were considered to construct peak displacement distributions

for shearwalls with different sheathing combinations (i.e, OSB only, OSB + GWB,

and OSB + Stucco). These results can be used to develop performance curves and

design charts (discussed later in Chapter 4.3.2.2) considering different construction

quality. Figures 4.39 through 4.41 show comparisons of peak displacements for

different combinations of sheathing materials and nonstructural finish materials. The

seismic weights in these figures were selected such that the majority of peak

displacements of the poor quality shearwalls were below a practical drift limit of four

inches. As another comparison, Figure 4.42 shows the peak displacements for the

L!I

different combinations of sheathing materials with one value of seismic weight (W

840 lbs/ft) for all shearwalls with or without nonstructural finish materials. The NSF

materials significantly improve the shearwall performance at all quality levels. Similar

analyses were performed for a shearwall with a large garage door opening using the

modified hysteretic parameters and the CCWP Task 1.4.4 test results. The results are

shown in Figures 4.43 through 4.46.

0.9

0.8

0.7

0.6

0.5

0.4

0.3

0.2

0.1

ni

TYP.

SUP.

/*

//

>--

IPOOR

8ft.0?

8ft

BW (8' x 8'), 8d@4"/12, OSB (/"), ç = 2%,


0 0.5 1 1.5 2 2.5 3 3.5 4

6max (in.)

Figure 4.39 Peak displacement distributions for construction qualities (BW1, OSBonly)

0.9

0.8

0.7

0.6

0.5U-

0.4

0.3

0.2

0.1

n

-

/ -. POOR

/ /

// ,H

8ft

8ft.1/ BW (8 x 8'), 8d@4"/1 2", = 2%,

7' - 9 OSB (I') GWB (1/) LS (10/50),- W = 1350 lbs/ft (48 kN total)

0 0.5 1 1.5 2 2.5 3 3.5 4

max (in.)

Figure 4.40 Peak displacement distributions for construction qualities (BW1, OSB +GWB)

0.9

0.8

0.7

0.6

0.4

0.3

0.2

0.1

SUPIYP.TT/ POOR//

// H-://

8L.

8ft.

/ 8W (8' x 8'), 8d4"/12', = 2%,/ OSB (/8") + Stucco (7/), LS (10/50),W = 1690 lbs/ft (60 kN total)

0 0.5 1 1.5 2 2.5 3 3.5 4

6max (in.)

Figure 4.41 Peak displacement distributions for construction qualities (BW1, OSB +Stucco)

70

0.9

0.8

0.7

0.6

0.5U-

0.4

0.3

0.2

0.1

n

I:! //

II,'! I

j: :/

/:7/

/

"-.-- OSB only

(SUP., TYP., POOR)

OSB+GWB(SUP., TYP., POOR)

OSB + Stucco(SUP., TYP., POOR)

8L.

8ft

BW (8' x 8'), 8d@4"/12", 2%,OSS (/"). GWB (1/2), Stucco (7/),


0 0.5 1 1.5 2 2.5 3 3.5 4

ömax (in.)

Figure 4.42 Peak displacement distributions for BW1 considering differentconstruction qualities

0.9

0.8

0.7

0.6

0.5

0.4

0.3

0.2

0.1

n

/ 1

POOR

H H8ft/ / 16ft.

/ / OW (16' x 8'), 8d©3"/12", = 2%,/ OSB (3/), LS (10/50),P W = 562 lbs/ft (40 kN total)

(,J

0 0.5 1 1.5 2 2.5 3 3.5 4

ömax (in.)

Figure 4.43 Peak displacement distributions for OWl (OSB only) consideringdifferent construction qualities

71

0.9

0.8

0.7

0.6

0.5LL

0.4

0.3

0.2

0.1

//v7TYP./iSUP

POOR

if H/e/1 :"// " H/ l6ft.

/ / OW (16' x 8'), 8d3"I12', = 2%,/ OSB (3/), GWB (/2), LS (10/50),-


0 0.5 1 1.5 2 2.5 3 3.5 4

max (in.)

Figure 4.44 Peak displacement distributions for OWl (OSB + GWB) consideringdifferent construction qualities

0.9

0.8

0.7

0.6

0.5

0.4

0.3

0.2

0.1

ft

TYP. / -

sup.,,-'

POOR

/ I

I /

I

/ / ,/ H H8ftI /I / OW (16'>< 8'), 8d3"I12", ç = 2%,/ - 0 OSB (3/5), Stucco (7/), LS (10/50),


0 0.5 1 1.5 2 2.5 3 3.5 4

max (in.)

Figure 4.45 Peak displacement distributions for OWl (OSB + Stucco) consideringdifferent construction qualities

72

1

0.9

0.8

0.7

LL

0.4

0.3

0.2

0.1

0

0

7ITYP.

SUP. POOR

7/1/

//

H8ft

/ I 16f1.

/ / OW (16'>< 8'), 8dc3"/12", = 2%,

I 0 OSB (3/), GWB (/2"), Stucco (p18").


0.5 1 1.5 2 2.5 3 3.5 4

6max (in.)

Figure 4.46 Peak displacement distributions for OWl (OSB + GWB + Stucco)considering different construction qualities

73

4.2.5 Effects of different seismic hazard regions

Previous studies (see Section 4.2.1.1) have shown that the greatest source of

uncertainty arises from the characterization of the seismic hazard. Thus, the inherent

variability in the ordinary ground motion records contribute significantly to the

variability in performance (peak displacement) of a shearwall. The CUREE-Caltech

Woodframe Project (CCWP) Task 1.5.3 was conducted using a suite of ordinary

ground motion records selected to characterize the seismic hazard in California

(seismic zone IV). These ground motions were recorded far enough from the rupture

to be free of typical near-fault pulse characteristics, and therefore near-fault ground

motions were not included. Furthermore, seismic zone IV and soil type D were

assumed when scaling to match the design spectrum. As such, it is unlikely the

performance curves and design charts developed in the CCWP Task 1.5.3 are

applicable to other seismic regions. The procedure developed in CCWP Task 1.5.3 is

sufficiently modular to allow a different suite of ground motion records (and hence

regions of seismic hazard) to be considered. Seismic zone II and III were considered to

investigate this issue. Much of the Pacific Northwest, including parts of Washington

and Oregon, has been designated seismic zone III [ICBO, 1997; FEMA, 2000 a,b].

Many woodframe buildings were damaged in the recent Nisqually earthquake, which

occurred in February, 2001 [Filiatrault et al., 2001]. The Northwest (seismic zone III)

earthquakes are characterized by long duration, subduction zone or interplate seismic

source, long hypocentral depth, and little aftershock activity. Earthquakes in Southern

74

California (seismic zone IV), on the other hand, are characterized by short duration,

shallow crustal seismic source, short hypocentral depth, and many aftershocks.

Three suites of ordinary ground motion records (20 records from LA, 20

records from Seattle, and 20 records from Boston) obtained from SAC Joint Venture

Project [Somerville et al., 1997] were used to generalize the methodology for

shearwall design (selection) for different seismic hazard regions (seismic zone II, III

and IV) and soil profile types (soil type B and D). Each ordinary ground motion record

was scaled independently for the appropriate performance level (e.g., life safety, 10%

probability of exceedance in 50 years) over the period range of interest. This was done

according to the procedures given by UBC '97 and the NEHRP guidelines [ICBO,

1997; FEMA, 2000a, b], as described previously. Figure 4.47 shows the target

response spectra for different seismic hazard regions according to UBC '97.

Information about the target response spectra for the different seismic regions is

shown in Table 4.7. The analysis of peak displacements was conducted as described

previously (see Section 3.3). The scaled peak ground accelerations for the 20 records

in the three different seismic zones are shown in Tables 4.8 through 4.10.

The method used for scaling earthquake records also was investigated. Three

cases are examined: scaling over the plateau region (presumed to be the period range

into which most woodframe structures fall), scaling to match at a period of 0.2 secs,

and scaling to match at a period of 0.5 secs. Figure 4.48 presents the peak

displacement distributions for the three different scaling methods for one given wall

configuration (BW 1), assuming the Durham nail hysteretic parameters, a 3 "/12"

75

nailing schedule, and the life safety (LS, 10/50) hazard level. When earthquake

records are scaled to target periods (0.2 sec or 0.5 sec), the displacement distribution

exhibits greater variability than when the earthquake records are scaled over the

plateau region. However, the median displacement values are similar. This is similar

to the results obtained by Shome (1999).

The performance levels and drifts limits are adopted from FEMA 356 [FEMA,

2000a, b], as was done previously. Details about the earthquake records used in this

study are provided in Appendix D.

1''

0

0)

(I)

0

0

Typical period range of interest (0.1 sec 0.6 sec)I I I for woodframe structure-

SI5 Zone fl (SeaU Region), SD

6 j Seismic Zone II (Boston Region), SD

\ Seismic Zone II (Boston Region), SB

4 1:

S___5___

2 _SS___ --

00 0.5 1 1.5 2 2.5 3 3.5 4

Period (sec)

Figure 4.47 Target response spectra for different seismic hazard regions

Seismic zone Soil profile type]

Period of interest1(plateau region) j

Spectral accelerationat plateau region, Sa

IV (LA) D 0.12 sec 0.58 sec 1.lgIII (Seattle) D 0.12 sec 0.60 sec 0.9g

II (Boston) B 0.08 sec 0.40 sec 0.375gD 0.12 sec 0.58 sec 0.55g

lable 4.7 Target response spectra for different seismic hazard regions

76

0.9

0.8

0.7

0.6

0.4

0.3

0.2

0.1

(1

/ Plateau Region

0.2sec0.5sec

8ft._8ft.

/ SW (8' x 8'), @3"/12", OSB (/8"),

> / ED /8, G = 184 ksi, ç = 2%, 0,', W = 843 lbs/ft (30 kN total), LS (10/50)

0 0.5 1 1.5 2

max (in.)

Figure 4.48 Comparison of earthquake record scaling to target response spectra

EQ Event &Year File Station

Peak Ground Acceleration (g)Unsealed Scaled

0GM LS (10/50)Imperial Valley LAOI Imperial Valley, El Centro 0.229 0.498

(1940) LAO2 Imperial Valley, El Centro 0.336 0.527LAO3 Imperial Valley, Array #5 0.390 0.460

Imperial Valley LAO4 Imperial Valley, Array #5 0.483 0.509(1979) LAOS Imperial Valley, Array #6 0.359 0.603

LAO6 Imperial Valley, Array #6 0.279 0.440LAO7 Landers, Barstow 0.132 0.677

Landers (1992) LAO8 Landers Barstow 0.133 0.638LAO9 Landers, Yermo 0.240 0.597LA1O Landers, Yermo 0.166 0.429

Loma Prieta LA1 1 Loma Prieta, Gilroy 0.372 0.464(1989) LA12 LomaPrieta, Gilroy 0.542 0,430

LA13 Loma Prieta, Newhall 0.658 0.430LA14 Loma Prieta, Newhall 0.63 8 0.479

Northridge LA15 Northridge, Rinaldi RS 0.675 0.5 14(1994) LA16 Northridge, Rinaldi RS 0.734 0.555

LA17 Northridge, Sylmar 0.575 0.691LA18 Northridge, Sylmar 0.825 0.480

N. Palm Springs LA19 North Palm Springs 0.343 0.497(1986) LA2O North Palm Springs 0.332 0.43 1

I able 4. 20 Ordinary ground motion records and PGA values (seismic zone IV, LA)

77

While using the SAC-Joint Venture earthquake records to consider other

seismic regions in this study, it was decided to compare the results obtained using the

CCWP and the SAC earthquake records for seismic zone IV (LA region). It was

presumed that the results would be similar. The nonlinear dynamic time history

analysis described previously was performed to investigate this issue. Both sets of

earthquake records were scaled to the same target response spectra and assumed soil

types D (SD). Figure 4.49 shows a comparison of peak displacements for both sets of

earthquake records with various assumed seismic weights. As expected, no significant

difference was observed. Therefore, it was decided that the 20 earthquake records

developed by CCWP would be used for all further analysis considering the LA

(California) hazard in this study. Table 4.11 summarizes the ground motions used to

evaluate shearwall response in different seismic hazard regions.


Peak Ground Acceleration (g)Unsealed Scaled

0GM LS (10/50)Imperial Valley SEO1 Long Beach, Vermon CMD Bldg 0.3 55 0.492

(1979) SEO2 Long Beach, Vermon CMD Bldg 0.276 0.362Morgan Hill SEO3 Morgan Hill, Gilroy 0.136 0.371

(1984) SEO4 Morgan Hill, Gilroy 0.233 0.495SEO5 West Washington, Olympia 0.206 0.361SEO6 West Washington, Olympia 0.189 0.345

Olympia(1949) SEO7

West Washington,Seattle Army B

0.055 0.360

SEO8West Washington,Seattle Army B

0.073 0.474

N. Palm Springs SEO9 North Palm Springs 0.344 0.4 12(1986) SE1O North Palm Springs 0.333 0.3 56

SEll Puget Sound, WA, Olympia, 0.175 0.423SE12 Puget Sound, WA, Olympia, 0.139 0.382

SeattleSE13

Puget Sound, WA,Federal OFC B

0.070 0.308

(1949)SE 14

Puget Sound, WA,Federal OFC B

0.057 0.343

5E15 Eastern WA, Tacoma County 0.033 0.392SE16 Eastern WA, Tacoma County 0.066 0.49 1SE17 Llolleo, Chile 0.563 0.385

Valparaiso SE18 Llolleo, Chile 0.541 0.401(1985) SE19 Vinadel Mar, Chile 0.320 0.432

SE2O Vinadel Mar, Chile 0.227 0.430

Table 4.9 20 Ordinary ground motion records and PGA values (seismic zone III,Seattle)

79


Peak Ground Acceleration (g)Unscaled Scaled

0GM LS(l0/50)

Reverse 1 BOO! Simulation, Hanging Wall 0.319 0.279B002 Simulation, Hanging Wall 0.191 0.217

Reverse 2B003 Simulation Foot Wall 0.267 0.282B004 Simulation, Foot Wall 0.207 0.270

New Hampshire B005 New Hampshire 0.054 0.4 14(1982) B006 New Hampshire 0.029 0.306

B007 Nahanni 0.978 0.329B008 Nahanni 0.920 0.417

Nahanni BOO9 Nahanni 0.303 0.385(1985) BOlO ianni 0.368 0.371

B011 Nahanni 0.145 0.493B012 Nahanni 0.148 0.535B013 Saguenay 0.128 0.252B014 Saguenay 0.174 0.261B015 Saguenay 0.163 0.319

Saguenay B016 Saguenay 0.077 0.432(1988) B017 Saguenay 0.056 0.246

B018 Saguenay 0.070 0.220B019 Saguenay 0.053 0.266B020 Saguenay 0.082 0.288

Table 4.10 20 Ordinary ground motion records and PGA values (seismic zone II,Boston)

Seismic Hazard Soil PerformanceRegion Type Level

Records Source

Seismic Zone IV 20 0GM from CUREE-Caltechjon)D LA region Woodframe Project

Seismic Zone III 20 0GM from(Seattle Region) LS (10/50) Seattle region

SAC Joint VentureSeismic Zone jj B 20 0GM from

D(Boston Region)[

Boston region

Table 4.11 Analysis matrix for effects of different seismic hazard regions

0.9

0.8

0.7

0.6

0.5

0.4

0.3

0.2

0.1

n

cCCWP (30 kN) /1SAC (30 kN)' // CCWP (40 kN)

/1 / SAC (40 kN)

/ /

ii /iV CCWP(5OkN)

/ SAC (50 kN)

I _8ft.Bft.

/ I // BW (8 x 8), @4"/12", OSB (3/)

ED /8, G = 180 ksi, ç = 2%, LS (10/50)

0 0.5 1 1.5 2 2.5 3 3.5 4

ömax (in.)

Figure 4.49 Comparison of peak displacement between CCWP and SAC earthquakerecords

The two sets of 20 SAC earthquake records for seismic zone III (Seattle

region) and seismic zone II (Boston region) each contain 10 fault-normal and 10 fault-

parallel earthquake records. Figure 4.50 shows the peak displacement distributions for

the 10 fault-normal earthquake records and the 10 fault-parallel earthquake records for

each seismic region and for a given seismic weight. This figure also shows the peak

shearwall response for the 20 combined earthquake records (i.e., 10 fault-normal and

10 fault-parallel earthquake records). No significant difference was observed between

the fault-normal and fault-parallel earthquake records. Therefore, it was decided to use

the combined of 20 earthquake records (i.e., 10 fault-normal plus 10 fault-parallel

earthquake records) for the purpose of investigating the effects of different seismic

hazards on shearwall performance.

1

U-

0.4

0.3

0.2

0.1

n

-

/7'(Boston) //'iii (Seattle) (LA)

1/:-Fault_Nor'l FaultjNormaj.<' - Fault_Normal

/- Combineçt/' CombineØ"/' - Combined

Ii - Fault_Pllel - Fault Pp.1l'el - Fault_Parallel

/ //'

/1 /A

/ 1' i"/ //' BW (8' x 8'), @6/12', OSB (/8"),J_y 9 ED = G = 180 ksi, = 2%,

) W = 1400 lbs/ft (50 kN total), LS (10/50)

0 0.5 1 1.5 2 2.5 3 3.5 46max (in.)

E:JI

Figure 4.50 Comparison of peak displacement between fault-normal and fault-parallelearthquake records

The effect of soil profile types also was considered. Soil profile type D (SD)

would be a relatively common (and conservative) design assumption for seismic zone

IV (LA) and III (Seattle). However, soil profile types for seismic zone II (Boston) are

difficult to determine (widely varying, highly localized conditions). Based on

consultation with a geotechnical engineer [Home, 20021, the representative soil profile

types for Boston region were determined. Thus, soil profile type D (SD) was assumed

for seismic zone IV (LA) and III (Seattle), and soil profile type B (SB) and D (SD)

were assumed for seismic zone II (Boston). Figures 4.51 and 4.52 show a comparison

of peak displacement distributions for these seismic regions considering two typical

nailing schedules and one value of seismic weight.

0.9

0.8

0.7

0.6

0.5U-

0.4

0.3

0.2

0.1

C)

-C7 --

iiostonI eattie D

(LA)D/ / ,,

/C+

I "' /I /

H

88.

I / fl

1/ BW (8' x 8'), @4/12", OSB (/8"),

I I -'

ED = J8", G = 180 ksi, = 2%,/ W = 1400 lbs/ft (50 kN total), [S (10/50)

0 0.5 1 1.5 2 2.5 3 3.5 4

max (in.)

Figure 4.51 Comparison of peak displacement for different seismic hazard regions(@4"/12", W 1400 lbs/fl)

0.9

0.8

0.7

0.6

0.5LL

0.4

0.3

0.2

0.1

n

/ \)7' II (Bostn)_D -

/ 'II (Boston)_B -1// /

III (SeatUey6 iv (LA)_D

II L-

I / ,-H /C

II

1/

88

/ BW (8'x 8'), @6/12", OSB (3/),

I / ED = /8", G = 180 ksi, - 2%,JJ - - - ' W = 1400 lbs/ft (50 kN total), LS (10/50)

0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5

6max (in.)

Figure 4.52 Comparison of peak displacement for different seismic hazard regions(@6"/12", W = 1400 lbs/ft)

Clearly, the shearwalls in seismic zone IV (LA) perform the worst because of

the higher peak ground accelerations. These figures illustrate the need to specifically

consider seismic hazard when specifying design requirements, i.e., selection of dense

nailing schedules, use of thicker sheathing materials, and so forth. In the case of

seismic zone II (Boston), shearwalls built in soil profile type B (SB, rock) perform

better than those in soil profile type D (SD, stiff soil).

4.3 Additional studies

4.3.1 Development of modification factors

4.3.1.1 Sheathing-to-framing connection hysteretic parameter variability

It is important to understand the extent to which sheathing-to-framing

connection hysteretic parameter variability influences the predicted response (peak

displacement) of the shearwall. Significant variability was observed in the fastener

data obtained under Task 1.4.8.1 of CUREE-Caltech Woodframe Project [see Figures

4.23 through 4.27]. Careful consideration must be given to how best utilize the data

obtained in CCWP Task 1.4.8.1. This is a potentially valuable database; however

some additional post-processing and evaluation of the data still may be needed to

develop design recommendations using a model-based procedure. Once done, it

should be possible to evaluate an appropriate modification factor to account for

fastener parameter variability.

The sheathing-to-framing connection hysteretic parameter variability can be

handled similar to the treatment of model uncertainty in Section 4.2.1.8, i.e., in the

form of modification factors (with parameters Xy, y) applied to the peak displacement

distribution. The lognormal parameters (Xz, z) for the worst-case (target) peak

displacement distribution and lognormal parameters (2x, Ex) for the median peak

displacement distribution can be obtained by the method of maximum likelihood.

Figure 4.53 illustrates the example for selection of median and target peak

displacement distributions for a particular set of wall parameters. The parameters of

these distributions are given by:

(4.3)

z =k + (4.4)

where, Xz and z = lognormal parameters for target peak displacement distribution, Xx

and x = lognormal parameters for median peak displacement distribution, Xy =

logarithmic mean of modification factor, and y = logarithmic standard deviation of

modification factor, respectively. The lognormal parameters Xy and can be obtained

by solving eqs. 4.3 and 4.4. Once they are determined, the statistical moments (mean t

and standard deviation ) for the modification factors (Xy, y) can be determined by:

(Y+)= e (4.5)

a e') (4.6)

where, ty and Jy are the mean and standard deviation of the modification factor,

respectively.

To illustrate the effect of choice of moments for the modification factors, the

mean was varied while holding the COy constant. This is shown Figure 4.54. Next,

the COy was varied while holding the mean constant. This is shown in Figure 4.55.

0.9

0.8

0.7

0.6

0.5U-

0.4

0.3

0.2

0.1

n

Target (worst-case)

Median

/

'I,,, ,// /// // / 8ft._8ft.

,' / / / BW (8' x 8'), 8d©3"112", OSE (3/3),

,/ / ED G = 180 ksi, = 2%,


0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9ömax (in.)

Figure 4.53 Selection of median and target peak displacement distributions

0.9

0.8

0.7

0.6

0.5U-

0.4

0.3

0.2

0.1

n

Median

Targetw=l.00,Vy=0.20yl.2O, Vy0.2O/ j.tyl.4O, Vy0.20

/ / / 1=1.50,V=0.20/ Ly=1.60,Vy=0.20

/jiyl.7O, Vy0.20

//////1 BW (8'

ED=3/5",G=180ksil2%,W = 1400 lbs/ft (50 kN total), [S (10/50)

0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5

ömax (in.)

Figure 4.54 Change of peak displacement considering various mean values ofmodification factor

0.9

0.8

0.7

0.6

0.5

0.4

0.3

0.2

0.1

n

Median

ty=1.00,Vy=0.20iy=l.00, Vy0.40jy=l.00, Vy=0.50ty=l.00, Vy0.60

j.ty=l.00, Vy0.70iy=l.00, Vy0.80

Target

/S 8ft.

I

I I I

I I I

8ft.II I

BW (8' x 8'), 8d@3"/12", OSB(3/),

ED=3/6",G=l8Oksi,ç2%,W = 1400 lbs/ft (50 kN total), LS (10/50)

0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5

6max (in.)

Figure 4.55 Change of peak displacement considering various COy values ofmodification factor

The baseline solid wall, BW1 (8 ft. x 8 ft.) with two sheathing panels oriented

vertically and a 3"/12" fastener schedule was considered when developing the

modification factors for sheathing-to-framing connection hysteretic parameter

variability. Only the life safety (10/50) limit state was considered. Five sets of nail

parameters (Durham, Folz and 3 BYU) were used. Also, the range of seismic weight

considered was 560 lbs/ft to 2530 lbs/ft. The average values for the mean and COV of

the modification factor (see Figure 4.56) for sheathing-to-framing connection

hysteretic parameter variability were 1.295 and 0.2 13, respectively.

[SIb]

1.6

1.4

1.2

>-

>0

0.6

0.4

0.2

01

500

Y,Avg .295

coy

-

1000 1500 2000 2500 3000

Seismic Weight (Ibs/ft)

Figure 4.56 Modification factors for sheathing-to-framing connection hystereticparameter variability

4.3.1.2 Construction quality

Significant differences were observed in shearwall performance (peak

displacement) considering different levels of construction quality (see Section 4.2.4).

Modification factors to account for construction quality can be obtained using the

procedure described in the previous section. The modification factor is intended here

to adjust the peak displacement distribution for a wall having superior quality to walls

having either typical or poor qaulity. Therefore, in eqs 4.3 and 4.4, Xz and z =

lognormal parameters for peak displacement distribution of typical or poor quality

walls, ?x and E = lognormal parameters for a peak displacement distribution of

superior quality wall, ?y and y = logarithmic mean and logarithmic standard

deviation of the construction quality modification factor, respectively.

Figure 4.57 shows the peak displacement distributions for BW1 assuming

superior, typical and poor quality. Modification factors are sought to predict the peak

displacement of shearwalls built with typical or poor quality, in relation to the superior

quality case. While these factors apply only to the wall being considered here, the

approach can be generalized for other wall configurations and/or definitions of

construction qaulity. As another example, Figure 4.58 shows the peak displacement

distribution for the wall with a large opening including nonstructural finish materials.

Two types of shearwalls, the baseline solid wall, BW1 (8 ft. x 8 ft.) with

sheathing panels oriented vertically, and the wall with a garage door opening, OWl

(16 ft. x 8 ft.), are considered to develop modification factors to account for effects of

construction quality. The life safety (10/50) limit state and various assumed seismic

weights are considered. Using the same procedure described in Section 4.3.1.1,

modification factors were developed for different values of seismic weight, and

different sheathing materials and shearwall types. Figures 4.59 through 4.64 show the

statistical moments of the modification factors for baseline wall BW1 as a function of

seismic weight. The mean modification factor remains relatively consistent,

particularly when the effects of nonstructural finish materials are not considered.

However, the COV tends to increase at high demands (large seismic weights). Figures

4.65 through 4.72 show the same results for the wall with a garage door opening.

Similar trends (consistent mean modification factor and increasing COV as a function

of seismic weight) were observed. Table 4.12 summarizes the moments of the

modification factors developed in this section.

Wall type Baseline solid wall (BW1) Wall with a garage door opening (OW 1)

Sheathing OSBOSB+ OSB+

OSBOSB+ OSB+ OSB +

GWB+GWB Stucco GWB Stucco

Stucco1.251 1.238 1.243 1.268 1.253 1.281 1.242

COVTYP 0.136 0.178 0.134 0.125 0.152 0.093 0.133

k'OOR 2.081 1.827 1.775 2.173 1.750 2.074 1.840COVPOOR 0.242 0.287 0.238 0.226 0.207 0.269 0.254

Table 4.12 Summary of modification factors considering construction quality

91

0.9

0.8

0.7

0.6

0.5U-

0.4

0.3

0.2

0.1

n

- 0 BW (8' x 8'), 8d@4"/12", OSB (/8"), = 2%,


0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5

ömax (in.)

Figure 4.57 Graphical method for determination of modification factors inconstruction quality (BW1)

0.9

0.8

0.7

0.6

0.5

0.4

0.3

0.2

0.1

n

SUP

TYP.

6 ft

118ft

OW (16' x 8'), 8d3"/12", 2%,

OSB (/"), GWB (/2"), Stucco (/8")

W = 422 lbs/ft (3) kN total),LS (10/50)

0 0.5 1 1.5 2 2.5

6max (in.)

Figure 4.58 Graphical method for determination of modification factors inconstruction quality (OWl)

92

2.5

2

1.5

>-

1

0.5

Y,AVQ 2.081

POOR

0' I

200 300 400 500 600 700 800 900 1000Seismic Weight (Ibs/ft)

Figure 4.59 Mean of modification factor for BW1 (OSB only)

0.4

0.3

>000.2

0.1

n

POOR

COVAvg = 0.242

/

TYP.

200 300 400 500 600 700 800 900 1000Seismic Weight (ibs/ft)

Figure 4.60 COV of modification factor for BW1 (OSB only)

93

2.5

2

1.5

>-

1

0.5

Y,Avg 1.827

.tY,Avg 1.238

TYP.

POOR

0'I

200 300 400 500 600 700 800 900 1000 1100 1200 1300 1400Seismic Weight (Ibs/ft)

Figure 4.61 Mean of modification factor for BW1 (OSB + GWB)

0.5

0.4

0.3

>00

0.2

0.1

COVAvg = 0.287

COVAvg = 0.178

TYP.

200 300 400 500 600 700 800 900 1000 1100 1200 1300 1400Seismic Weight (Ibs/ft)

Figure 4.62 COV of modification factor for BW1 (OSB +GWB)

2.5

2

1.5

>-:1

I

0.5

,Avg 1.775

lY,Avg = 1.243

POOR

TYP.

0'I


Figure 4.63 Mean of modification factor for BW1 (OSB + Stucco)

0.5

0.4

0.3

>000.2

0.1

POOR

COVAvg0238,'

COVAVg = 0.134

TYP.

0'I


Figure 4.64 COV of modification factor for BW1 (OSB + Stucco)

2.5

2

1.5

>-

I

0.5

A2.173 - _!

POOR

_±'_= 1 .

,iii:;1- -..

TYP.

0' I

100 200 300 400 500 600Seismic Weight (Ibs/ft)

Figure 4.65 Mean of modification factor for OWl (OSB only)

>00

0100 200 300 400


Figure 4.66 COY of modification factor for OWl (OSB only

500 600

2.5

2

1.5

>-

1

0.5

POOR

-Y,=

LY,Avg = 1.253 -

TYP.

0'100 200 300 400 500


Figure 4.67 Mean of modification factor for OWl (OSB + GWB)

C

0.4

0.3

>000.2

0.1

COVAvg = 0.207

COVAVg = 0.152

TYP.

POOR

0'100 200 300 400 500

Seismic Weight (lbs/ft)

Figure 4.68 COV of modification factor for OWl (OSB + GWB)

2.5

tY,Avg = 2.074

2

1.5JiY,Avg 1.281

0.5

TYP.

UI

400 500 600Seismic Weight (Ibs/ft)

Figure 4.69 Mean of modification factor for OWl (OSB + Stucco)

[II

0.3

>000.2

0.1

rir400

POOR

COV=ft26:

CoVAv90oN500 600Seismic Weight (Ibs/ft)

Figure 4.70 coy of modification factor for OWl (OSB + Stucco)

700

700

97

2.5

2.tY,Avg = 1.840

U

1.5

> l-Y,Avg = 1.242

0.5

POOR

0'100 200 300 400 500 600 700 800


Figure 4.71 Mean of modification factor for OWl (OSB + GWB + Stucco)

>00

0100 200 300 400 500 600 700 800

Seismic Weight (tbs/ft)

Figure 4.72 COY of modification factor for OWl (OSB + GWB + Stucco)

4.3.1.3 Contribution of nonstructural finish materials

In Section 4.2.3, it was shown that the presence of stucco and/or gypsum

wallboard would result in significantly reduced peak drifts, particularly at high

demands. Many of the walls considered previously (particularly those without

nonstructural finish materials) were generally well behaved (i.e., low COV's in peak

displacement distribution) at lower values of seismic weight, but frequently exhibited

very large (and highly variable) drifts at larger seismic weights. (This also relates to

the geometric instability concept described in the Incremental Dynamic Analysis

section, see Section 4.4.1.) The result is often a poorer fit to the peak displacement

cumulative distribution function (CDF), in particular over the upper tail, which forms

the basis for the performance curves, design charts, and fragility curves. Simply put,

the more well behaved the shearwall response (i.e., the lower the variability in peak

displacement, by avoiding geometric instabilities), the more robust the procedure

developed in this study becomes. By taking proper account of the finish materials, not

only will the peak drifts be reduced, but also it is likely that the variability in peak

displacements can be maintained at relatively low levels over a wider range of seismic

weights.

The theory of products of statistically independent lognormal variables was

used to develop the modification factors for sheathing-to-framing connection

hysteretic parameter variability (Section 4.3.1.1) and construction quality (Section

4.3.1.2). However, that procedure cannot be used when developing modification

factors to account for nonstructural finish material effects because the COV's in peak

100

displacement of walls with nonstructural finish materials generally are lower than

those without nonstructural finish materials (i.e., OSB sheathing only). Since

= + , then the logarithmic standard deviation (,y) for the modification factor is

obtained from the following equation;

Y ='J (4.7)

where, z logarithmic standard deviation for the target peak displacement

distribution (wall with NSF materials), x = logarithmic standard deviation for the

median peak displacement distribution (wall built with OSB only), and y

logarithmic standard deviation of the modification factor. The expression under the

square root must be positive for a real solution. However, in many cases, this value is

negative because the value of z is smaller than Ex (i.e., the value of COV in peak

displacement distribution considering OSB and NSF materials is lower than the COV

considering OSB only). Therefore, it was decided to develop a deterministic

modification factor to match the 90tlpercentile values of peak displacement for

shearwalls built with nonstructural finish materials. The deterministic modification

factor (i.e., logarithmic mean, ?y) can be obtained by varying the logarithmic mean

(?y) to visually match the 90thpercentile value of target peak displacement

distribution (OSB + NSF materials). Figures 4.73 and 4.74 graphically illustrate this

approach for the baseline solid shearwall and the large wall with a pedestrian door

opening, respectively. Alternatively, the deterministic modification factor can be

obtained directly using the lognormal distribution:

101

F(z) (4.8)

where cI?(.) is CDF of the standard normal distribution, is the logarithmic mean,

and z is the logarithmic standard deviation. The distribution parameters (X, ) are

obtained using a maximum likelihood procedure. Once the lognormal parameters (?z

z) for the target displacement distribution are determined (i.e., peak displacement

distribution including the effects of nonstructural finish materials), the 90thpercentile

value can be estimated knowing the lognormal parameters by solving the following

equation for lnZ.

(lnZ= 0.90 = (1.28) (4.9)

in which, Xz and are the lognormal parameters of peak displacement considering

the wall with NSF materials. Once lnZ is obtained from eqn. 4.9, it can be used in the

following equation:

jlnz(2 +%)'( +) ,J=0.9o=(1.28) (4.10)

where, ?x and x are the lognormal parameters of peak displacement considering the

wall with OSB only, 2y is the logarithmic mean of modification factor, and y is the

logarithmic standard deviation of modification factor. It is assumed that the

logarithmic standard deviation (y) is nearly zero because the value of x is generally

larger than z. Finally, the deterministic logarithmic mean (Xy) can be obtained

solving eqn. 4.10.

102

The resulting mean modification factor to account for the effects of

nonstructural finish materials, considering the baseline solid wall (BW1) sheathed

with OSB, are shown in Figure 4.75. The result in Figure 4.75 suggest that adding

stucco results in a greater reduction in peak displacement than only adding gypsum

wallboard, and this effect increases with increasing seismic weight. This is also seen in

Figure 4.76, which considered the shearwall sheathed with plywood. A shearwall with

two-sided gypsum wallboard (GWB) also was considered (Figure 4.77). The mean

modification factor for the two-sided GWB wall was about 1.5 times greater than that

wall with one-sided gypsum wallboard (see Table 4.13).

In the case of the wall with a large garage door opening, attaching stucco to the

outside and gypsum wallboard to the inside significantly decreases the peak

displacements. The mean modification factors as a function of seismic weight for

OWl are shown in Figure 4.78. Unlike the wall with the garage door opening, the use

of gypsum wallboard (in addition to stucco) does not significantly improve the

performance of the large wall with a pedestrian door opening. This can be seen in

Figure 4.79 which shows that the mean modification factors as a function of seismic

weight are nearly the same. Table 4.13 summarizes the deterministic modification

factors to account for the contribution of nonstructural finish materials.

103

Wall type Baseline wallWall with Qpening

Garage door Pedestrian doorOSB+GWB 0.548 N/A N/AOSB + Stucco 0.470 0.904 0.450

OSB + GWB + Stucco N/A 0.770 0.450PWD + GWB 0.827 N/A N/A

PWD + GWB (2 sidL 0.42 1 N/A N/APWD + Stucco 0.632 N/A N/A

Table 4.13 Developed deterministic modification factor (JLy) for contribution ofnonstructural finish materials effects

0.9

0.8

0.7

1x

0.5

0.4

0.3

0.2

0.1

OL

0

-90t5 Percenthe

11 1/ OSB only

/ /1/OSB + Stucco

'1/'I I

I I

/ 11/I I 18ff.IIIII

8ff

/ /1/ BW(8 x 8'), 8d@4"/12", OSB (3/),

0 GWB (/2"), Stucco (/8"), = 2%,I = 1400 lbs/ft (50 kN total), LS (10/50)

0.5 1 1.5 2 2.5 3 3.5 4

6max (in.)

Figure 4.73 Graphical method to develop deterministic modification factors innonstructural finish materials effects (BW1)

104

0.9

0.8

0.7

0.6

0.5

0.4

0.3

0.2

0.1

9ot0percenthe

OSB + StuccoOSB only

kg OSB+GWB

OSB + Stuc

I l/I'I/Il/I

li/i

/

:,' ii __8ft.I'll,'

i/i I,

OW (16' x 8'), 8d@6"/12", OSB (3/),

/ GWB (1/2), Stucco (7/), = 2%,


0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6

max (in.)

Figure 4.74 Graphical method to develop deterministic modification factors innonstructural finish materials effects (0W2)

0.9

0.8

0.7

0.6

>-

0.4

0,3

0.2

0.1

n

OSB + Stucco

---- - - - l.tg

OSB + GWB

200 400 600 800 1000 1200 1400 1600 1800 2000


Figure 4.75 Mean of deterministic modification factor for BW1 (OSB sheathing)

0.9

0.8

0.7

0.6

>-:1

0.4

0.3

0.2

0.1

n

PWD + GWB

- - vg0.8vg.87__

PWD + Stucco

105

200 400 600 800 1000 1200 1400 1600 1800 2000Seismic Weight (lbs/ft)

Figure 4.76 Mean of deterministic modification factor for BW1 (Plywood sheathing)

09

0.8

0.7

0.6

>-:1

0.4

0.3

0.2

0.1

C)

PWD + GWB

200 400 600 800 1000 1200 1400 1600 1800 2000Seismic Weight (lbs/fl)

Figure 4.77 Mean of deterministic modification factor for BW1 (Plywood sheathing)

106

0.9

0.8

0.7

0.6

>-

0.4

0.3

0.2

0.1

(1

= 0.904

OSB + Stucco

100 200 300 400 500 600 700 800 900 1000Seismic Weight (Ibs/ft)

Figure 4.78 Mean of deterministic modification factor for OWl (OSB sheathing)

1

0.9

0.8

OSB + Stucco + GWB0.7

OSB + Stucco

0.4

0.3

0.2

0.1

n

100 200 300 400 500 600 700 800 900 1000 1100 1200 1300Seismic Weight (Ibslft)

Figure 4.79 Mean of deterministic modification factor for 0W2 (OSB sheathing)

107

4.3.2 Construction of performance curves and design charts

4.3.2.1 Baseline walls

The sensitivity studies described previously (Section 4.2.1) were used to

establish appropriate ranges of system parameters to consider in developing

performance curves and design charts. Two baseline wall types were considered: an 8

ft. x 8 ft. solid wall with two sheathing panels oriented vertically, and a 8 ft. x 16 ft.

long wall with a large garage door opening, vertically oriented sheathing at the ends,

and a solid header over the opening. Two sheathing types were considered, OSB and

plywood, both 3/s-in. thick. The assumed deterministic fastener hysteretic parameters

(power-driven spiral nail for the OSB and 8d common nail for the plywood; i.e., the

Durham and Dolan parameters) are shown in Table 4.14. Fastener schedules

(edge/field) ranging from 3"/3" to 6"/12" were considered. The suite of 20 ordinary

ground motion records developed for the CUREE-Caltech Woodframe Project

(CCWP) was used, scaled for the LS (10/50) and JO (50/50) limit states, with 2%

damping ratio assumed. No additional account was taken of model uncertainty or

fastener hysteretic parameter variability. Full overturning anchorage and all fasteners

were assumed to be properly installed.

Values obtained by Durham (1998). Fasteners were 2 in. (50mm) long, power-driven spiral nailsattaching 3/8-in. (9.5mm) OSB to SPF framing members.

K0 r1 r2 r3 r4W F0 F1 A J_us

3.2030.061 -0.078 1.40 0.143

0.169 0.032 0.492_ç

0.8 1.1kip/in kips kips in0.561

0.061 -0.078 1.40 0.1430.751 0.141 12.5

0.8 1.1kN/mm kN kN mmValues otr4 changed per note 3 in Table 4.1.

Fastener hysteretic parameters (average values shown) determined using results obtained by Dolan(1989). Fasteners were 8d common nails attaching 318-in. (9.5mm) Plywood to SPF framing member,3/8-in, edge distance, loading perpendicular to grain.

K0 r1 r2 r32 r4 F0 F1 a

US 0.050 -0060 1 1.40 0.0270.227 0.041 0.315

0.8 1.1kip/m kips kips in0.907

0.050 -0.060 1.40 0.0271.010 0.182 8.0

0.8 1.1kN!mm kN kN mmValue of r3 was assumed.

Table 4.14 Fastener parameters used to develop performance curves and design chartsfor baseline walls

4.3.2.L1 Construction of performance curves

The information presented in the peak displacement distributions can be post-

processed into a more useful form for engineering designlassessment, using seismic

weight as the dependent variable (design parameter). This is not dissimilar from the

approach taken by researchers in New Zealand in which walls are rated using

sustainable seismic mass as the primary design parameter [Deam, 1997, 2000]. For the

present study, the range of seismic weight was determined from engineering design

calculations for selected woodframe structural configurations. For example, using

0.138g, unit shears were found to range from 23 lbs/ft to 200 lbs/ft in the second floor,

and 47 lbs/ft to 492 lbs/ft in the first floor of a two-story house. This corresponds to a

maximum seismic weight of 3565 lbs/ft. (A practical limit on mid-range demand

commonly seen in one- and two-family detached dwellings was suggested [Cobeen,

2000] to be about 3000 lbs/ft.) Performance curves can be constructed, using the

lognormal parameters for the appropriate peak displacement distributions, for

increasing values of seismic weight and for a set of structural parameters. Each

performance curve therefore corresponds to a particular limit state (JO, 50/50 or LS,

10/50) and non-exceedence probability (i.e., 50%, 84%, 90%, 95%, and 99%). Design

charts (described in next section) can then be constructed using the information in the

performance curves. These charts allow selection of a particular sheathing type and

fastener spacing (e.g.), for a given seismic weight, at a particular performance level or

non-exceedence probability. The quantities shown on the axes in both cases

(performance curves and design charts) are peak displacement and seismic weight.

The performance curves, as defined above, for baseline wall (BW1) with 3/8-in.

OSB, assuming the Durham fastener parameters and considering four different

fastener spacings, are shown in Figures 4.80 through 4.83. Each figure shows the 99%,

95%, 90%, 84%, and 50% non-exceedence curves, for both the life safety (LS, 10/50)

and immediate occupancy (JO, 50/50) limit states. Also shown are the FEMA 356 drift

limits, 2% for life safety (LS, 10/50) and 1% for immediate occupancy (JO, 50/50). A

slightly more restrictive JO (50/50) drift limit of 0.75% also is shown. Figure 4.83

presents the same performance curve information as shown in Figure 4.80, but with

the axes switched. Performance curves presented using either format (6max vs. seismic

weight, or seismic weight vs. ömax) can be used as design aids for shearwall selection.

For example, considering Figure 4.80 or Figure 4.84, and assuming a target peak drift

110

non-exceedence probability of 95%, the wall having parameters shown in the figure

can sustain about 1450 lbs/ft , limited by drift limit of 2% for LS (10/50).

Figures 4.85 and 4.86 present performance curves for 3/8-in. plywood,

assuming the Dolan fastener parameters and two different spacings. Note that in

addition to the different assumed fastener hysteretic parameters, the plywood has a

significantly lower shear modulus than the OSB. While not a complete range of

fastener spacings, these cases are presented for comparative purposes. Notice that this

wall can sustain significantly higher seismic weights than the wall with 3/8-in. OSB

due to differences in the fasteners and the sheathing material.

The performance curves for baseline wall (OWl) are shown in Figures 4.87

through 4.89 (OSB, three different fastener spacings) and Figures 4.90 and 4.91

(plywood, two different fastener spacings). As with the BW1 performance curves, each

figure shows the 99%, 95%, 90%, 84% and 50% non-exceedence curves, for both the

life safety (LS, 10/50) and immediate occupancy (JO, 50/50) limit states, and the

corresponding drift limits. Notice again the increased allowable seismic weight, due to

the use of larger nails and plywood (vs. OSB), evident when comparing Figure 4.87

and 4.89.

Figure 4.92 shows the effect of including model uncertainty on the LS (10/50)

performance curves for BW1 (cf. Figure 4.79). Values of COV in the model error term

of 0%, 15%, and 30% are considered. The effect is seen to be relatively small at low

seismic weights, but can become significant at higher weights. Note that model

111

uncertainty is not explicitly considered in developing the performance curves and

design charts in this dissertation.

4.3.2.1.2 Design charts

Design charts are constructed using the information in the performance curves.

Specifically, one design chart (i.e., set of selection curves) is developed for each of the

two performance limit states (LS, 10/50 and JO, 50/50) at a given percentile value, or

non-exceedence probability. The design chart can thus be used to select a particular

sheathing type and fastener spacing for a given seismic weight to ensure that the wall

performs within the specified drift limit. The performance curves presented previously

(see Figures 4.80 through 4.91) are used to construct examples of these design charts.

Design charts for baseline wall BW1 (8 ft. x 8 ft. solid wall) are shown in Figures 4.92

and 4.93 for the JO (50/50) and LS (10/50) limit states, respectively. Design charts for

baseline wall OWl (16 ft. x 8 ft. wall with a large garage door opening) are shown in

Figures 4.95 and 4.96, for the same two performance limit states.

112

3.5

3.0

2.5

2.0

><

Ero 1.5

1.0

0.5

0.0

/ 99% (LS)8ff /

8ff / / 95% (LS)BW (8 x 8'), @3/6", OSB (/"),

/ 90% (LS)ED '8, G = 180 ksi, = 2%

,' ./._' 84% (LS)/ , a,2%Drift=1.92in. 777.-- 50% (LS)

1% Drift= 096 in.

99% (10)0.75%Drift=072ui.95%(l0)

400 600 800 1000 1200 1400 1600 1800 2000


Figure 4.80 Performance curve for BW1, OSB (3/8-in.), @3"/6"

5

4

3

E02

1

99% (LS)88.

88.

95% (LS)BW (8' x 8'), @4"/12", OSB (/8"),

ED3/8", G180ks1, =2%90% (LS)

84% (LS)

2%Drift=1.92jn. /1% Drift = 0.96 in.

9s 0

400 600 800 1000 1200 1400 1600 1800 2000


Figure 481 Performance curve for BW1, OSB (3/8-in.), @4"/12"

113

6.5

6.0

5.5

5.0

4.5

4.0

. 3.5

J3.0

2.5

2.0

1.5

1.0

0.5

0.0

/ 99% (LS)8 ft.

8 ft.

SW (8' x 8'), @6"/6", os ç3i"), 95% (LS)ED = /8", G = 180 ksi, = 2% /

90% (LS)

84% (LS)

2% Drift = 1.92 in 50% (LS) 99% (tO)

400 600 800 1000 1200 1400 1600 1800 2000



3.5

3.0

2.5

2.0

E1.5

1.0

0.5

I 2%Drift=1

99% (LS)

95% (LS)

90% (LS)84% (LS)

/,2' 50% (LS)

1% Drift = 0.96 in.

/ 0 75% DrIft2_

BW (8'>< 8'), @6/12",

8ft. OSB(3/8"), ED=3/8",G = 180 ksi, = 2%

8 ft.

, 99% (10)

95% (10)90% (10)

V ::-- 84% (10)

0.0

400 600 800 1000 1200 1400 1600 1800 2000



114

2000

1800

1600

E(I,

. 1400

0)1200

0E 1000C/)

a)

800

:iiii

400

50, 84, 90, 95, 99% (tO) 50% 84% 90% 95% 99% (LS)

i ii:' .

I j / ,/ 8 ft.

!! :'.

I /,?,' 8fti'l aBW (8' x 8'), @3/6", OSB (3/),

ED=3/8",G=l8oksi,ç=2%

0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5

max (in.)

Figure 4.84 Performance curve for BW1, OSB (3/8-in.), @3"/6", axes switched

8 ft. 99% (LS)3.0

8ft. / 95% (LS)2.5 BW (8' x 8'), 8d@3"/6", PWD (/") I'

G=60 ksi 2/o

1 0 1/ Drift = 096 in

0.5 j90%(IO)84% (tO)

50% (tO)

0.0

400 600 800 1000 1200 1400 1600 1800 2000


Figure 4.85 Performance curve for BW1, PWD (3/8-in.), 8d@3"/6"

115

6

5

4

c

E

2

1

0

8ft.

/99%(LS)

8ft. /BW (8' x 8'), 8d@4"/12", PWD (3/J

ED /8, G = 60 ksi, ç = 2% I" / 95% (LS)/ // /

,. 90%(LS)7 // /, 84%(LS)

;

400 600 800 1000 1200 1400 1600 1800 2000


Figure 4.86 Performance curve for BW1, PWD (3/8-in.), 8d@4"/12"

9

8

7

6

5

'8

E4

3

99% (LS)

184 in.

H H96mn

/OW (16' x 8), @3/3", OSB (/8"), 95% (LS)ED=3/8",G=l8Oksi,ç=2%,

90% (LS)

7/ 84% (LS)

100 300 500 700 900 1100 1300 1500


Figure 4.87 Performance curve for OWl, OSB (3/8-in.), @3"/3"

116

8

7

6

5

Ee.o

3

2

0

199%(LS)I I

96 in.

184 in.

95% (LS) OW (16' 8'), @4"14", OSB (/'),

ED=3/8", G= 180 ksi,ç=2%,

i;

100 300 500 700 900 1100 1300 1500



9

8

7

6

x

E4

3

2

0

L__J

[J96in

184 in.

OW (16'x 8'), @6/6", OSB (/8").

95% (LS) ED = G

90% (LS)

84% (LS)

99% (10)1

95% (10)

90% (10)

84% (10)

100 300 500 700 900 1100 1300 1500

Seismic Weight (lbs/fl)


117

6

5

4

C8<3E

2

1

0

L

H96 in.

/99% (LS)

184 in. /OW (16 x 8'), 8d©4"/4", PWD (/a")/ 95% (LS)ED = '8, 0 = 180 ksi, = 2%, / j

/::::

90% (LS)

84% (LS)99% (10)

95% (10)

90% (10)

84% (10)

3.

100 300 500 700 900 1100 1300 1500


Figure 4.90 Performance curve for OWl, PWD (3/8-in.), 8d@4"/4"

6

I I I99%(l0)

96 in. OW (16' >< 8'), 8d@6"16", PWD (/8"),5 ED=3/<",G=l8Oksi,ç=2%,

184 in.

,,

99/o(LS)

4 I1 950/ '10

3 /95%(LS)

0

/ 90% (LS) 90% (10)

/,//84%(LS)

84% (10)

TI!' Tr;O;I

100 300 500 700 900 1100 1300 1500


Figure 4.91 Performance curve for OWl, PWD (3/8-in.), 8d@6"/6"

3.5

3.0

2.5

><

Ec 1.5

1.0

0.5

8 ft.

8ft.

8W (8 x 8'), ©3"/6", OSB (/8"),ED3/8",G180ksi,?=2%

2% Drift= 1.92 in.

/ 95% (COV=30%)/, 95% (COV=15%)/ 95% (COV=0%)

784% (COV=30%)84% (COV=15%)84% (COV=0%)

118

0.0-200 400 600 800 1000 1200 1400 1600 1800 2000 2200


Figure 4.92 Effect of model uncertainty on performance curve for BW1, OSB (3/8-in.),@3"/6"

2.0

1.8

1.6

1.4

1.2C

1.0E

0.8

0.6

0.4

0.2

0.0

PWD (4/12")j OSB (6/12")

8ft. / OSB (6/6")

BW (8 x 8), ED /" = 2%, , ./G08 = 180 ksi, GPWD = 60 ksi, 10(50/50) f

1% Drift = 0.96 in. ,vOSB(4"i12")

0.75% Drift = 0.72 in. ..V

OSB (3/6")PWD(3"/6')

400 600 800 1000 1200 1400 1600 1800 2000

Seismic Weight (Ibs!ft)

Figure 4.93 95t11-Percentile design chart for BW1, TO (50/50)

119

6

5

4

E

2

8 ft.

8 ft.

BW(8'x8'), ED3/5",=2%,G055 = 180 ksi, GFWD = 60 ksi,

LS (10/50)

2% Drift

OSB (6/6")

PWD (4/12")OSB (4/12")

PWD (4/12")

PWD (3/6')

0-!

400 600 800 1000 1200 1400 1600 1800 2000


Figure 4.94 95t''-Percenti1e design chart for BW1, LS (10/50)

5

4

3C

E

2

1

L__j

Li 96 in.

184 in.

BW(16' x 8'), ED = '8, = 2%,

G0s8 = 180 ksi, GPWD 60 ksi,

10 (50/50)

OSB (6/6"V

)ND (6/6"

0DB (4"/4"

B (3/3")

'D (4/4"

0-!

100 300 500 700 900 1100 1300 1500

Seismic Weight (!bs/ft)

Figure 495 95thPercenti1e design chart for OWl, JO (50/50)

7

6

5

9.c3

2

1

L_i

L_J 96 in.

184 in.

8W (16 x 8), ED 3/a", = 2%,

GOSB = 180 ksi, GPWD = 60 ksi,

LS (10/50)

1 2% Drift= 1.92 i

PWD

OSB (6/6")

OSB (4"/4")

/PWD (4"/4")

0!

100 200 300 400 500 600 700 800


Figure 4.96 95t1'-Percentile design chart for OW!, LS (50/50)

120

121

4.3.2.2 Construction quality

4.3.2.2.1 Construction of performance curves

The peak displacement distributions for the three different construction

qualities (superior, typical, and poor) were compared in Section 4.2.4. Significant

differences were observed among walls built with different construction quality levels.

The post-processing procedure described in Section 4.3.2.1 also was used to construct

performance curves considering different construction quality levels.

Performance curves for BW1, assuming 3/8-in. OSB, 4"/12" nailing schedule

with two rows of nails along the sides of the wall (as was tested in CUREE Task

1.3.1), were developed considering the three different levels of construction quality

defined in Table 4.5. These are shown in Figures 4.97 through 4.99. Each figure shows

the 95% and 84% non-exceedence curves, for the life safety (LS, 10/50) limit states.

Also shown is the FEMA 356 drift limit of 2% for life safety (LS, 10/50). Considering

Figure 4.99, and assuming a target peak drift non-exceedence probability of 95% and a

drift limit of 2% for LS (10/50), the wall built with superior quality and having

parameters shown in the figure can sustain about 1700 lbs/fl, the wall built with

typical quality can sustain about 1450 lbs/ft, and the wall built with poor quality can

sustain about 1030 lbs/ft.

Performance curves for OWl, assuming 3/8-in. OSB and a 3"/12" nailing

schedule, are shown in Figures 4.100 through 4.103. As with the BW1 performance

curves, each figure shows the 95% and 84% non-exceedence curves, for the life safety

(LS, 10/50) limit state and the corresponding drift limit. Note that model uncertainty

122

and sheathing-toframing hysteretic parameter variability are not explicitly considered

in developing the performance curves and design charts in this section since the

shearwall global hysteretic parameters were obtained directly the experimental test

results (see Section 4.2.4).

4.3.2.2.2 Design charts

Design charts are constructed using the information in the performance curves.

As an example, design charts (i.e., sets of selection curves) are developed for the LS

(10/50) performance limit state at a given percentile value, or non-exceedence

probability. The design chart can thus be used to select a particular sheathing type and

nonstructural finish material combination (with consideration of construction quality)

for a given seismic weight. The performance curves presented previously (see Figures

4.97 through 4.103) are used to construct examples of these design charts. Design

charts for BW1 (8 ft. x 8 ft. solid wall) are shown in Figures 4.104 and 4.105 for

typical and poor quality, respectively. Design charts for OWl (16 ft. wall with a large

garage door opening) are shown in Figures 4.106 and 4.107, again for the two

construction quality levels. Figures 4.108 and 4.109 present design charts for the two

different walls (BW1 with OSB + Stucco, and OWl with OSB + GWB + Stucco)

considering the three different levels of quality. Note that CUREE Task 1.3.1 did not

consider walls sheathed with OSB and both GWB and stucco.

4.5

4.08ff

88

BW (8' x 8), 8d@4"/12" (2 rows),3.0

OSB (/8"), = 2%, [S (1 0/50),

OSB only

2.5

j2.0 2% Drift = 1.92 in.

1.5

1.0

200 400 600 800


Figure 4.97 Performance curve for BW1, OSB only

4.0

3.58ff.

3.0BW (8' x 8'), 8d©4"/12" (2 rows),

OSB (3/), ç = 2%, LS (10/50),

2.5OSB+GWB

29

1.0

0.5

123

95% (POOR)

84% (POOR)

95%(TYP.)

84%(TYP.)95% (SUP.)

84% (SUP.)

1000 1200

95% (POOR)

84% (POOR)

95% (TYP.)

84% (TYP.)

95% (SUP.)84% (SUP.)

200 400 600 800 1000 1200 1400 1600 1800


Figure 4.98 Performance curve for BW1, OSB + GWB

5.0

4.58 ft.

4.08 ft.

3.5 BW (8' x 8'), 8d@4"/12" (2 rows),OSB (/"), = 2%, LS (10/50),

3.0 OSB+Stucco

2.5E

2.0 2

1.5

1.0

0.5

0.0

/ 95% (POOR)

/ ,.. 84% (POOR)

95%(TYP.)

84% (TYP.)95% (SUP.)84% (SUP.)

200 700 1200 1700


Figure 4.99 Performance curve for BW1, OSB + Stucco

4.0

I I 95% (POOR)

L_i3.0 16ft.

OW (16' x 8'), 8d@3"/12", OSB (/"),ç = 2%, LS (10/50), OSB only

C

2.0 2% Drift= 1.92 in.

1.5

1.0 -

,./_, ..-...___.,_'

0.5

o.oi

84% (POOR)

95% (TYP.)84% (TYP.)95% (SUP.)84% (SUP.)

100 200 300 400 500 600


Figure 4.100 Performance curve for OWl, OSB only

700

2200

800

124

H 1186

3.0 16ft.

OW (16' < 8'), 8d@3"/12", OSB (/8"),

2 5= 2%, LS(10/50), OSB + GWB

C2°A

1.0:

0.5

0.0 mrr

/ 95% (POOR)

84% (POOR)

95% (TYP.)

84% (TYP.)

95% (SUP.)

84% (SUP.)

100 200 300 400 500 600 700


Figure 4.101 Performance curve for OWl, OSB + GWB

:

H 1186.

3.0 l6ft.

OW (16' < 8'), 8d©3"/12", OSB (/8"),

ç = 2%, LS (1 0/50), OSB + Stucco

2.5

C2%Drjftl.921n.

0.0

95% (POOR)

84% (POOR)

95% (TYP.)

84% (TYP.)

95% (SUP.)84% (SUP.)

100 200 300 400 500 600 700


Figure 4.102 Performance curve for OWl, OSB + Stucco

800

H800

125

126

4.0

95% (POOR)

H H8ft3.0 16 ft.

./ 84% (POOR)OW (16 x 8), 8d©3/12", OSB (/"),

2 5= 2%, LS (10/50), / /

OSB + GWB + Stucco ./ // / / 95% (TYP.)

2.0 2%Drrft=1.92in. /.. / 84% (TYP.)

E ./ / / ,, 95%(SUP.)

1.5 / V / ,./ 84% (SUP.)

1.0

::

100 300 500 700 900


Figure 4.103 Performance curve for Owl, OSB + GWB + Stucco

5.0

4.5

4.0

3.5

3.0

2.5

E

2.0 2% Drift = 1.92 in.

//

OSB + Stucco

8 ft.

2" (2 rows),(10/50),

Poor Quality

200 700 1200 1700 2200


Figure 4.104 95thPercentile design chart for BW1, poor quality

127

3.0

OSB + Stucco2.5

OSB+GWB

OSB only /220

1921fl1Typical Quality

0.0200 700 1200 1700 2200


Figure 4.105 95thPercentile design chart for BW1, typical quality

4.0

OSB + GWB + Stucco / i,1 / /3.5 .. /-// / IOSB + Stucco ,)& / /

3.0 OSB+GWBIII1uI'I?"2.5 OSB only / / /

22.0 2%Drift=1.921n. ,/

:: H16ff.

118ff

0.5 -' OW (16 >< 8'), 8d@3"/12",OSB (I8"), = 2%, LS (10/50).

Poor Quality

0.0100 300 500 700 900

Seismic Weight (lbslft)

Figure 4.106 95thPercentile design chart for OWl, poor quality

2.5

2.0 2% Drift 1.92 in. /

1.5

OSB + GWB + Stucco

OSB + Stucco

OSB only

H H8tt166.

OW (16 x 8'), 8d@3"/12",OSB (/8), = 2%, LS (10/50),

0.0Typical Quality

100 300 500 700 900


Figure 4.107 95th1Percenti1e design chart for OWl, typical quality

5.0

4.5 POOR8ft. /

401 /8 ft.

3.5 BW (8'>< 8'), 8d@4"/12" (2 rows),OSB (I"), = 2%, LS (10/50),

3.0OSB+Stucco

10

200 700 1200 1700


Figure 4.108 95thPercentile design chart for BW1, (OSB + Stucco)

2200

:,H

3.0 l6ft

OW (16 x 8'), 8d@3"/12', OSB (/8"),

ç = 2%, LS (10/50),2.5 OSB + GWB + Stucco

C

1.0

0.5

POOR

TYP.

sup.

0.0

100 300 500 700

Seismic Weight (lbs/It)

Figure 4.109 95thPercenti1e design chart for OWl, (OSB + GWB + Stucco)

900

129

130

4.3.2.3 Effects of different seismic hazard regions

In Section 4.2.5, shearwall performance was compared for three different

hazard regions (LA, Seattle, and Boston). Different suites of earthquake records,

characterizing each of the three regions, were used to develop peak displacement

distributions. Again using the post-processing procedure described in Section 4.3.2.1,

performance curves were developed for one wall configuration (BW1) for each of the

three different seismic hazard regions.

Performance curves for BW1, assuming 3/8-in. OSB, Durham fastener

parameters, and three different fastener spacings are shown in Figures 4.110 through

4.117. Considering Figures 4.115 through 4.117, assuming a drift limit of 2% for LS

(10/50) and a target peak drift non-exceedence probability of 95%, the wall having the

parameters shown in the figure can sustain about 1900 lbs/ft in seismic zone II

(Boston), 920 lbs/ft in seismic zone III (Seattle), and 700 lbs/ft in seismic zone IV

(LA). Design charts (see Section 4.3.1) were constructed using the information in the

performance curves. The design charts for the baseline solid wall BW1 (8 ft. x 8 ft.),

considering the life safety (10/50) limit state, is shown in Figures 4.118. Since Boston

(seismic zone II) is a relatively low seismic hazard region, the 3"/12" nailing schedule

was not considered. Note also that only OSB sheathing materials were considered.

131

4.0

/99%3.5 8ft.

3Q 8ft.95%

BW (8 x 8'), @3"/12", OSB (/8"), /ED = '8, G = 180 ksi, ç 2%, ,' 9Q%

2.5 Seismic Zone Ill (Seattle), LS (10/50) // 84%

2.0 2% Drift = 1.92 in. ," ///

-,. 1

," ..-." 50%1 .5 " .-'

1.0 _-.9

0.5 _-.._.--

0.0I

200 400 600 800 1000 1200 1400 1600 1800 2000 2200


Figure 4.110 Performance curve for BW1, seismic zone III (Seattle), @3"/12"

7.0

/99%6.0 8ft. /

8ft.5.0

BW(8'x8'),@3"/12",OSB(3/8"), / . 95%ED = G = 180 ksi, ç = 2%, / /

4.0

Seismic Zone IV (LA), LS (10/50)

:::

3.0

50%2% Drift = 1.92 in.

0.0 r-rrTJ-r--r---r--rJ--r---1I !!IrrIrI!!rrIr

200 400 600 800 1000 1200 1400 1600 1800 2000 2200

Seismic Weight (bs!ft)

Figure 4.111 Performance curve for BW1, seismic zone IV (LA), @3"/12"

132

7.0

6.0 8ft.

8 ft.

5.0 BW(8 x 8), @4"112", OSB (I"),ED = /8. G = 180 ksi, ç= 2%,Seismic Zone II (Boston), LS (10/50)

4.0 95%

2% Drift= 1.92 in.

50%1.0

I0.0 C

II

0 500 1000 1500 2000 2500 3000 3500 4000

Seismic Weight (lbs/fl)

Figure 4.112 Performance curve for BW1, seismic zone II (Boston), @4"/12"

4.0

99%3.5 8ff.

3.0 8ft.

BW (8' x 8'), @4/12", OSB (/"). / 95%ED G = 180 ksi, = 2%, / /

2.5 Seismic Zone III (Seattle), LS (10/50) 90%

2.0 2%Driftl.92jn. 84%

1.5

50%

0.0

200 400 600 800 1000 1200 1400 1600 1800 2000



133

L8L8ft.

3.5 BW (8' x 8'), @4/12", OSB (/8"),ED=3/8",G=l8oksi,ç=2%,Seismic Zone IV (LA), LS (1 0/50)

2.5

0

2.0 2% Drift = 1.92 in.

0.0

200 400 600 800 1000 1200 1400



99%

95%

90%

84%

50%

6.0

8ft.5.0

8 ft.

BW (8' x 8'), @6/12" OSB (/8"),4.0 ED-3/8",G- l8Oksi,ç2%,

Seismic Zone II (Boston), LS (10/50)

3.0

E

2.0 - 2% Drift = 1.92 in.

1.0

1600

/ 99% 1

95%

90%

84%

50%

0.0

0 500 1000 1500 2000 2500 3000


Figure 4.115 Performance curve for BW1, seismic zone II (Boston), @6"/12"

134

_18ft /L%

8W (8' x 8'), @6/12", OSB (/"),40 ED3/8",G180ksi,1=2%,Seismic Zone Ill (Seattle), LS (10/50) 95%

:.z

3.0 90%

84%!

2 0 2°/ Drift = 1 92 in ." -- - -50%

1.0

0 J -] ..............T---r----r-- !I!!--------r--------rF

200 400 600 800 1000 1200

Seismic Weight (Ibs!ft)


3.5

3.0 8ft.

8 ft.

2.5 BW (8' x 8'), @6/12", OSB (/8"),

ED3/e",Gl8Oksi,/=2%,Seismic Zone IV (LA), LS (10150)

2.0 2% Drift= 1.92 in.

><

E.c 1.5

0.0

99%

200 300 400 500 600 700 800 900



135

5.0

4.5

4.0

3.5

3.0

LA (6/1 2")

//SEA(

LA (3/12")

LA (4/12")

/ SEA(3"(4/112")

(6/12")

BOS (4/12")

2.0 D=2

0 500 1000 1500 2000 2500 3000 3500 4000


Figure 4.118 95th..percentile design chart for BW1, LS (10/50)

136

4.4 Performance-based design

4.4.1 Incremental dynamic analysis

Development and implementation of performance-based design requires

translating performance requirements into structural checking equations for use by

design engineers. This is not always straightforward and may present considerable

challenges to codes and standards committees. Even among CCWP Element 1

researchers, and the broader community of earthquake engineers in general, the three

performance levels and corresponding drift limits specified in FEMA 356 (e.g.) have

raised questions. The limit state associated with structural collapse is well understood

by structural engineers. Performance limit states associated with life safety (LS, 10/50)

(access/egress) and immediate occupancy (JO, 50/50) are less well defined and less

well understood. The FEMA 356 drift limits for woodframe shearwalls were adopted

in this study, and no further statement is made about their suitability as a basis for

performance-based design.

It may be possible to use nonlinear analysis models (such as CASHEW and

SASH1) to evaluate appropriate definitions for the collapse prevention (e.g., CP, 2/50)

limit state. This approach, called an Incremental Dynamic Analysis (IDA), has been

applied to nonlinear MDOF systems [Cornell, 2000]. Incremental dynamic analysis

(IDA) is a new analysis method that involves performing nonlinear dynamic analyses

of the structural model under a suite of earthquake ground motion records, each scaled

to several intensity levels designed to force the structure all the way from elasticity to

final global dynamic instability [Vamvatsikos, 2002].

137

In an IDA, given record is scaled incrementally and the nonlinear response

(peak displacement) is evaluated. Thus, for each record considered, a curve of spectral

acceleration (Sa) vs. peak displacement is obtained. Among the characteristics of these

curves is often a point at which the "slope" reduces dramatically, indicating a Sa value

above which the displacement increases very quickly. This is analogous to the

buckling response one sees in an imperfect column. If such a point can be evaluated

for a range of structural configurations and ground motion records, for example, it

may be possible to suggest a physically-based drift limit associated with impending

collapse. This concept is briefly explored here using the CASHEW modeling

procedure, BW1, and the ordinary ground motion records used in this study. The suite

of 20 ordinary ground motion records was divided into three groups because it was too

crowded to show all results on one figure. Two types of baseline shearwalls were

considered, BW1 and OWl. Figure 4.119 shows a typical IDA curve obtained from

nonlinear dynamic time history analysis with increasing spectral acceleration. The

IDA curve usually starts linearly in the elastic range, however, it becomes highly

nonlinear after it reaches the break point with a dramatic change in slope. (A

phenomenon, termed structural resurrection, has been observed in which a system is

pushed all the way to global collapse at some intensity measure, only to "reappear" as

non-collapsing at a higher intensity level [Vamvatsikos, 2002J.) One method for

estimating the collapse prevention point is shown in Figure 4.120. For comparison, the

collapse prevention (CP) 3% drift limit provided by FEMA 356 also is shown.

138

Figures 4.121 through 4.123 show Sa vs. peak displacement for baseline wall

BW1. Also shown are the tangents defining the apparent break points for those points.

A characteristic value could be selected as the design drift limit for collapse

prevention (CP, 2/50). For this particular example, the mean value corresponds to a

peak displacement of 3.04 in., or about 3.17% of the wall height. (Note that the FEMA

356 drift limit for collapse prevention (CP, 2/50) is 3% for wood shearwalls.) Figures

4.124 through 4.126 show Sa vs. peak displacement for the baseline wall with the large

opening. Also shown are the tangents defining the apparent break points for those

points. A characteristic value could be selected as the design drift limit for collapse

prevention (CP, 2/50). In this example, the mean value corresponds to a peak

displacement of 3.70 in., or about 3.86% of the wall height. This approach to limit

state identification appears to have merit and may be worth further study. The IDA

approach is expected to be more appropriate, however, for the analysis of entire

buildings rather than individual subassemblies.

1.2

1.0

0.8

0.6C,)

0.4

0.2

0.0

0.0 10.0 20.0 30.0 40.0 50.0 60.0 70.0

6max (in.)

Figure 4.119 Typical IDA curve

0.6

0.5

0.4

0.3

0.2

0.1

0.0

0.0 2.0 4.0 6.0

max (in.)

Figure 4.120 Estimated collapse points by tangent slope

8.0 10.0

139

1.0

0.9

0.8

0.7

0.6

a,0.5

(I)

0.4

0.3

0.2

0.1

0.0

0 1 2 3 4 5 6 7 8 9 10

max (in.)

Figure 4.12 1 Set of IDA curves (BW1, group 1)

1.5

1.2

0.9

a)

(I)

0.6

0.3

0.0

0 1 2 3 4 5 6

6max (in1)

Figure 4.122 Set of IDA curves (BW1, group 2)

7 8 9 10

140

1.2

1.0

0)0.6

C,)

0.4

0.2

0.0

0 1 2 3 4 5 6 7 8 9 10

ömax (im)

Figure 4.123 Set of IDA curves (BW1, group 3)

1.2

1.0

0.8

0.6

0.4

0.2

0.0

0 1 2 3 4 5 6 7

ömax (in.)

Figure 4.124 Set of IDA curves (Owl, group 1)

8 9 10

141

142

2.0

1.8

1.6

1.4

1.2

C)1.0

(I)

0.8

0.6

0.4

0.2

0.0

0 1 2 3 4 5 6 7 8 9 10

max (in.)

Figure 4.125 Set of IDA curves (OWl, group 2)

1.0

0.8

C)0.6

0)

0.4

0.2

0.0

0 1 2 3 4 5 6 7 8 9 10

max 011)

Figure 4.126. Set of IDA curves (Owl, group 3)

143

4.4.2 Fragility curves

While performance curves are constructed as a function of seismic weight,

fragility curves can be develop as a function of hazard level (e.g., spectral

acceleration, Sa). This fragility approach has a number of potential advantages,

particularly when considering multiple damage states. Such an approach also may be

useful for performing loss estimation studies. A fragility methodology may have

applications to design and post-earthquake condition assessment {Rosowsky and

Ellingwood, 2001].

The fragility of a structural system commonly is modeled by a lognormal

distribution function (CDF). The lognormal CDF is given by:

FR(y)=[ln(Y/mR)

(4.11)

where 1(.) = standard normal distribution function, mR = median capacity and R =

logarithmic standard deviation of capacity, approximately equal to the coefficient of

variation (COy), VR, when VR<O.3.

4.4.2.1 Fragility curve based on peak displacement

The baseline solid shearwall BW1 (8 ft. x 8 ft.), with two OSB sheathing

panels oriented vertically, was used to develop illustrative fragility curves. The Folz

nail parameters were used to develop the global hysteretic parameters of the shearwall

using CASHEW. Various nailing schedules (2"112", 3"/12", 4"/12", and 6"/12") and

the corresponding allowable seismic weights (back-calculated from the UBC '97

144

allowable unit shear values) were considered. Since the UBC '97 allowable unit shear

values include nonstructural finish material contribution and assume the wall is acting

as part of an entire woodframe building, it is necessary to consider the appropriate

overstrength factor (R) when converting allowable unit shears to seismic weights.

Since the assembly considered here is an isolated shearwall, values of R factor ranging

from 2.5 to 5.5 were considered when developing the fragility curves.

The total horizontal base shear, V, can be derived using the following equation:

F=ma=[Ja=[JW (4.12)

in which, F = force, m mass, a = acceleration, and g = acceleration due to gravity.

The UBC '97 form of this expression is somewhat modified. The (a/g) term is

replaced by a "seismic base shear coefficient". The UBC '97 base shear formula for a

the main lateral force resisting system is given by:

CI 2.5C1W

RT R

= 25"wo.l1c 1W (4.13)R

v2.5CI

w O.8ZCI w (for seismic zone 4

where V = base shear, W = weight of structure, CVI/RT = velocity-based seismic base

shear coefficient, 2.5CaI/R = acceleration-based seismic base shear coefficient, Z

seismic zone factor, I = occupancy importance factor, Ca and C == seismic (response

spectrum) coefficients, and T = C1(h)213 = structure period [ICBO, 1997; Breyer et.

al., 1998]. The acceleration-based seismic base shear coefficient (2.5CaI/R) usually

governs for buildings with short fundamental periods, and most woodframe structures

145

fall into this category. Therefore, using eqs. 4.12 and 4.13, and the allowable unit

shear values from UBC '97 Table IT-I-i, the seismic weight can be obtained. These

values are shown in Table 4.15 for the cases considered here (importance factor of 1.0,

seismic zone 4, soil profile type D, and various overstrength factors). Only unit shear

values considering 3/8-in. thick sheathing panel and 8d nails are used to calculate

seismic weights from allowable unit shear values in Table 23-IT-I-i in the UBC '97.

Panelgrade

Minimumnominal

panelthickness

Minimumnail

penetrationin framing

Panel applied directly to framingNail spacing at panel edges

Nail size (in).6 4 3 2

Structurall/8 l'/2 8d 230 360 460 610

Overstrength factor(R)

Seismic baseshear coefficient

Weight (lb.)

2.5 0.440 5854.5 I 9163.6 I 11709.1 I 15527.33.5 0.314 8203.8 12840.8 16407.6 21758.04.5 0.244 10557.4 16524.6 21114.8 28000.05.5 0.200 12880.0 20160.0 25760.0 34160.0

Table 4.15 Seismic weights calculated based on UBC '97 allowable unit shear values(Table 23-TI-I-i)

The peak displacement curves were next developed for the different nailing

schedules and R factors. Figures 4.127 through 4.129 show the sample CDF's

assuming a 3"/12" nailing schedule, various R factors, and the three different hazard

levels (JO, LS, CP). As expected, as the R factor increases, the allowable seismic

weight increases, and hence shearwall peak displacement increases. Peak displacement

curves for the other nailing schedules are provided in Appendix E.

0.9

0.8

0.7

0.6

0.5U-

0.4

0.3

0.2

0.1

n

V/ R=2.5, W=1464 lbs/ft

R=3.5, W=2051 lbs/ft

W=2639 lbs/ft

/ R5.5, W=3229 lbs/ft

// 88.

;1 ___/

88.

/ 0 BW (8' x 8), 8d@3/12", OSB (/8"),

ED=3/5',G=200ksi,ç2%, 10 (50/50)

0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2

ömax (in.)

Figure 4.127 Peak displacement distributions for different R factors (3"/12", JO)

0.9

0.8

0.7

0.6

0.5U-

0.4

0.3

0.2

0.1

n

---

/ ---

/

/ / ,' R=2.5, W1464 lbs/ft

/ VR=3.5, W=2051 lbs/ft

// R=4.5, W=2639 lbs/ft

I / / R5.5, W=3229 lbs/ft

/ /

8ft.

1/:' 88.

0 8W (8' x 8'), 8d©3 /12 , OSB (I8"),

- - ED G = 200 ksi,ç = 2%, LS (10/50)

0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 5.5 6

ömax (in.)

Figure 4.128 Peak displacement distributions for different R factors (3"/12", LS)

147

0.9

0.8

0.7

0.6

0.5U-

0.4

0.3

0.2

0.1

n

R=2.5, W=1464 lbs/ft

R=3.5, W=2051 lbs/ft

R=4.5, W=2639 lbs/ft

R=5.5, W=3229 lbs/ft

,,

fl8ft.

/,/- BW (8 < 8), 8d@3"/12", OSB (/8'),

-. ED = G = 200 ksi, ç = 2%, CP (2/50)

0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 5.5 6

max (in.)

Figure 4.129 Peak displacement distributions for different R factors (3"/12", CP)

By changing the spectral acceleration for the 20 earthquake records, a peak

displacement CDF can be developed for each level of scaling. The probability of

failure can be determined non-parametrically as the relative frequency of the peak

displacement exceeding the specified drift limits. If this probability of failure is

conditioned on a given value of spectral acceleration, this becomes one point on the

fragility curve. This has the advantage of not requiring that a particular distribution be

fit to the peak displacements. The records were scaled to five different hazard levels:

50% in 50 years (72-year mean return period or MRI), 20% in 50 years (225-year

MRI), 10% in 50 years (474-year MRI), 5% in 50 years (975-year MRI) and 2% in 50

years (2475-year Mifi).

Figures 4.130 through 4.133 present fragility curves for a baseline solid wall (8

ft. x 8 ft.) with two full sizes OSB sheathing panels oriented vertically, and four

different nailing schedules. The wall is assumed to be fully anchored. The effective

seismic weight acting on the wall was determined based on the allowable unit shear

values in the UBC '97, as described previously. An overstrength factor (R) of 5.5 was

assumed and drift limits of 1%, 2% and 3% were considered. The seismic demand

(interface) variable is spectral acceleration, Sa. Fragility curves of this type can be

used either as design aids or to assess risk consistency in current design provisions.

The fragility curves for other overstrength factors (2.5, 3.5 and 4.5) are provided in

Appendix F.

1

0.9

0.8

0.7

0.6

0.5

0.4

0.3

0.2

0.1

Ii

7-8ft. /

8ft. / /,,BW (8' x 8'), 8d@2"112", OSB (/"), /"-

ED=318",G= 185ksi,=2%, R=.5W = 4271 lbs/ft (152.0 kN total)

/

/0 (50/50)

/ /

/ / ,'

//

0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2

Sa(g)

Figure 4.130 Fragility curves for three different hazard levels (2"/12")

149

1

0.9

0.8

0.7

0.6

0.4

0.3

0.2

0.1

C)

- -

/10 (50/50)

/ /

/ /,'

/ /,",

//,,,,

/ / ,/ 8ft.

/ 1/8W (8' x 8'), 8d@3"/12", OSB (I")ED=3/8",G=200ksi,ç=2%, R=5.5

- - - W = 3220 lbs/ft (114.6 kN total)

0 0.5 1 1.5 2 2.5

S5(g)


1

0.9

0.8

0.7

0.6

0.4

0.3

0.2

0.1

(1

- - -

I0 (50/50)

/' CP (2/50)

//,,,,,//,,,,

/ / 8ft.

/ / ' BW (8 x 8'), 8d@4"/l 2", OSB/ / ED = /8, G = 180 ksi, = 2%, R = 5.5

- - W = 2521 lbs/ft (89.7 kN total)

0 0.5 1 1.5 2 2.5

Sa(g)


150

0.9

0.8

0.7

0.6

a 0.5

0.4

0.3

0.2

0.1

QL

0

/ 0 (50/50)

/ /,,,

//,,,,,

//:"

/,,'/' BW (8 x 8), 8d@6112", OSB (/8"),

ED=3/5",G=185ksi,=2%,R=5.5W = 1610 lbs/ft (57.3 kN total)

0.5 1 1.5 2 2.5

Sa(g)


Figures 4.134 through 4.137 show the resulting fragility curves for the baseline

solid shearwall (8 ft. x 8 ft.) with various nailing schedules (2"/12", 3"/12", 4"/12",

and 6"/12") considering life safety (LS, 10/50), for R factors ranging from 2.5 to 5.5.

The UBC walls provided relatively consistent levels of safety, as evidenced by the fact

that the resulting fragility curves were quite close for all nailing schedules. That is, the

allowable seismic weights provided in the UBC '97 for the different nailing schedules

resulted in comparable levels of performance. This permit the results for the different

nailing schedules to be combined to construct a single fragility curve for a given R-

factor (see Figure 4.138). Complete fragility curves for various R factors and

considering different seismic hazard levels (TO and CP) are provided in Appendix G.

Figure 4.139 shows the fragility curves for different assumed R factors considering the

151

baseline solid shearwall (8 ft. x 8 ft.), a 3"/12" nailing schedule, and the life safety

(10/50) hazard level. As expected, the failure probability of the shearwall increases as

the overstrength factor (R) increases.

I

0.9

0.8

0.7

0.6

0.5

0.4

0.3

0.2

0.1

0

0

8d@3"/12", W=1464 lbs/ft / /8d@4"/12", W=1147 lbs/ft

/ /

8d@2"/12", W=1940 lbs/ft //

8d@6"/12", W=730 lbs/ft

/1///

BW (8 x 8'), oSB (3/5), ED =

- = 2%, R = 2.5, LS (10/50)

0.5 1 1.5 2 2.5

Sa(g)

Figure 4.134 Fragility curves considering R = 2.5 (LS, 10/50 hazard level)

152

1

0.9

0.8

0.7

0.6

0.5

0.4

0.3

0.2

0.1

0

_8ft. /BW (8 x 8'), OSB (/8"), ED = /8", /'!'ç = 2%, R = 3.5, LS (10/50)

/8d@6"/12" W1026 lbs/ft

/'// 8d@2"/12", W=2720 lbs/ft

///8d@3"/12", W=2051 lbs/ft

/ // 8d@4"/12", W=1605 lbs/ft

/,//I

0 0.5 1 1.5 2 2.5Sa(g)


1

0.9

0.8

0.7

0.6

o 0.5

0.4

0.3

0.2

0.1

0

7/8 ft.

BW (8' x 8'), OSB (I") ED = /8,

= 2%, R = 4.5, LS (10/50)

8d©6"/12", W=1321 lbs/ft

/ 8d©2"/12", W3500 lbs/ft

/' 8d(4"/12", W=2065 lbs/ft'I,' / 8d@3"/12", W=2639 lbs/ft

0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2

Sa(g)


153

1

0.9

0.8

0.7

0.6

0.5

0.4

0.3

0.2

0.1

0

L8fl.

88.

BW (8' x 8'), OSB (/8"), ED =

ç = 2%, R = 5.5, LS(1O/50)

/rd©I W=4270 lbs/ft1/ 8d©3"/12", W=3220 lbs/ft

// 8d)4"/12", W=2521 lbs/ft

/// / 8d@6"/12", W=1610 lbs/ft//

0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2

Sa(g)


1

0.9

0.8

0.7

0.6

0.5

0.4

0.3

0.2

0.1

0

88.

88.

BW (8' x 8'), OSB ( /"), ED = '8,' /

= 2%, R = 4.5, LS (10/50)

Single Average Fragility/ 18d©6"/12", W=1321 lbs/ft

8d@2"/12", W=3500 lbs/ft

: ::::

0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2

Sa(g)

Figure 4.138 Single fragility curve considering R = 4.5 (LS, 10/50 hazard level)

154

1

0.7

0.6

0.4

0.3

0.2

0.1

Ill

-_-

8ft. 7 78ft. /,/

BW (8' x 8), 8d©3"/12", OSB (/") /

ED G = 200 ksi, = 2% //

R=2.5

/ R=3.5

/;' / R=4.5

/,,' /R=5.5

/,'/0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2

Sa(g)

Figure 4.139 Fragility curves considering different assumed R factors (LS, 3"/12")

4.4.2.2 Fragility curve based on ultimate force

In this section, the fragility curve concept is extended to the issue of shearwall

anchorage. Fragility curves can provide a useful tool for the selection of seismic hold-

downs making some assumptions about the amount of force being transferred from the

top of the wall down to the wall corner being anchored. By statics, the horizontal force

acting on top of the wall is equal to uplift load in the bottom plate if the dimension of

shearwall is square. Three types of hold-down anchors (Simpson HTT 22, LTT 20B,

and PHD2-SDS3) were considered. The hold-down ultimate tension capacities were

assumed to be Normally distributed with mean values taken then average ultimate

tension capacities obtained from the Simpson Strong-Tie catalogue and assumed

COV's of 0.2. With this information, the Sthpercentile value for capacity of each hold-

155

down was determined and was treated as a capacity limit. (This is similar to the drift

limit used when considering peak displacements.) These are shown in Table 4.16.

1

Hold-down type Average ultimate tension capacity(from Simpson Catalogue)

Design value(5thpercentile)

HTT22(16) 13150 lbs. 8824 lbs.LTT 20B 8733 lbs. 5860 lbs.

PHD2-SDS3 12520 lbs. 8401 lbs.

Table 4.16 Capacities of hold-downs considered in this study

Baseline wall BW1 (solid shearwall, 8 ft. x 8 ft., two OSB sheathing panels

oriented vertically) was used to develop illustrative fragility curves for anchorage

selection considering uplift capacity. The CASHEW program and the Folz nail

parameters were used to develop the global shearwall hysteretic parameters. Different

nailing schedules (2"/12", 3"/12", 4"/12", and 6"/12") were considered along with the

corresponding allowable seismic weights back-figured from the UBC '97 allowable

unit shear values (described in Section 4.4.2.1). Various R factors (2.5, 3.5, 4.5, and

5.5) also were considered.

The peak displacement distributions were then developed for the different

nailing schedules and R factors. Figures 4.140 through 4.142 show the sample CDF's

assuming a 3"/12" nailing schedule, various R factors, and the three different hazard

levels (JO, LS, and CP). As expected, as the R factor increases, the allowable seismic

weight increases, and hence shearwall peak displacement increases. Peak displacement

curves for the other nailing schedules are provided in Appendix H.

156

0.9

0.8

0.7

0.6

0.5U-

0.4

0.3

0.2

0.1

n

FT't ::::

I ;tLLR=5.5

I :1 _____

8ft.

Ii; Bft.

cI II

ii/ BW (8' x 8), 8d@3"/12",

/II OSB(318"),ED-3/8',

, / G = 200 ksi, = 2%,

0 (50/50)0_I

0 2000 4000 6000 8000 10000 12000 14000

Fmax (lbs.)

Figure 4.140 CDF for ultimate force with various R factors (3"/12", TO)

0.9

0.8

0.7

0.6

0.5LL

0.4

0.3

0.2

0.1

n

.4' ,'/ _______8ft.

8,,BW(8 x8)ED /8 G

8d@3/12 OSB(/8)200 ksi ç 2% LS (10/50)

4000 6000 8000 10000 12000 14000

Fmax (lbs.)

Figure 4.141 CDF for ultimate force with various R factors (3"/12", LS)

157

0.9

0.8

0.7

0.6

0.5

0.4

0.3

0.2

0.1

n

/8ff.

BW (8 x 8), 8d@3"/1 2", OS ('/8"), /ED = /8, G = 200 ksl, ç = 2°h, CP (2/50)

E /1 R=5.5

c!1C10

6000 8000 10000 12000 14000

Fmax (lbs.)

Figure 4.142 CDF for ultimate force with various R factors (3"/12", CP)

The same method used to develop the fragility curves for peak displacement

was used to develop fragilities for uplift force of shearwall, i.e., time-history analysis

using a suite of 20 ordinary ground motions scaled to different hazard levels. Using

the same baseline solid shearwall (8 ft. x 8 ft.), and considering three different hold-

downs (Simpson HTT 22, PHD2-SDS3, and LTT 20B), fragility curves for ultimate

uplift force were constructed for differing nailing schedules and assuming different R

factors from 2.5 to 5.5. Figures 4.143 through 4.145 present the fragility curves for the

baseline solid shearwall (8 ft. x 8 ft.) for each of the three different hold-downs

(considering one particular nailing schedule for each case). As the overstrength factor

(R) increases, the failure probability of each of the shearwalls increase.

1

0.9

0.8

0.7

0.6

( 0.5

0.4

0.3

0.2

0.1

n

8ft. 1/8 ft.

BW (8 x 8'), 8d©3"112", OSB (/8"),

ED /8, G200ks1, ç=2%, HTT22(16),' /

,"

/ R=4.5

/ R=3.5

/

158

0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2

Sa(g)

Figure 4.143 Fragility curve for ultimate uplift force with various R factors (3 "/12",HTT 22)

0.9

0.8

0.7

0.6

0.4

0.3

0.2

0.1

n

/7/8ft. //

BW (8' x 8'), 8d@3"112", OSB el8'), /

ED318", G=2OOksi,2%,PHD2 / /I/I /

/

::::

/ R=2.5

0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2

Sa(g)

Figure 4.144 Fragility curve for ultimate uplift force with various R factors (3"/12",PHD2-SDS3)

0.9

0.8

0.7

0.6

a o.

0.4

0.3

0.2

0.1

/

1

/

//

'I /

'I/

1/

R = 5.5

R = 3.5

R = 2.5

8 ft.

8 ft.

BW (8' >< 8), 8d©4"1l2", OSB (I8'),

ED=3/8", G = 180 ksi, =2%, LTT2OB

159

0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2Sa(g)

Figure 4.145 Fragility curve for ultimate uplift force with various R factors (4"/12",LTT 20B)

The fragilities for each of the three hold-downs, considering a single nailing

schedule and overstrength factor, also can be shown on the same figure. One example

is shown in Figure 4.146 for a 3"/12" nailing schedule and assuming R = 3.5. Again,

the effective seismic weight acting on the wall was determined based on the allowable

unit shear values in the UBC '97. The seismic demand (interface) variable is the

spectral acceleration, Sa. Fragility curves of this type can be used either as design aids

(selection of hold-downs) or to assess risk consistency in current design provision.

160

0.9

0.8

0.7

0.6

ci 0.5

0.4

0.3

0.2

0.1

C)

88. //BW (8' x 8'), 8d©3"/12",

OSBeI8"), ED =18,

G=2OOksi,=2%,R=3.5 LTT2OB

/

/

/

/

0 0.4 0.8 1.2 1.6 2

Sa(g)

Figure 4.146 Hold-down fragility curve considering ultimate uplift capacity

161

5. ANALYSIS OF SHEARWALLS IN COMPLETE STRUCTURES

5.1 Model configuration

This chapter presents the results from studies of shearwalls acting as part of

complete woodframe structures. As with the isolated shearwalls analyzed in the

previous chapter, the shearwalls are subjected to a suite of earthquake records, scaled

appropriately to specified hazard levels.

For more than 40 years, researchers have been conducting full-scale

experimental tests to investigate the performance of woodframe structures and

assemblies under wind, snow and earthquake loading [e.g., Dorey and Schriever,

1957; Hurst, 1965; Yokel et al., 1973; Tuomi and McCutcheon, 1974; Stewart et al.,

1988; Sugiyama et al., 1988; Phillips et al., 1993; Ohashi et al., 1998; Seo et al., 1999;

Paevere et al., 2003]. As part of CUREE-Caltech Woodframe Project (CCWP), the

seismic response of two and three-story woodframe structures were tested on large

shake tables [Fischer et al., 2001; Mosalam et al., 2002]. 'While tested at full-scale, the

overall size of these test structures was slightly scaled-down due to size (footprint)

limitations of the shake tables. The results were used to validate a numerical model

used to predict the seismic performance of complete woodframe structures [Folz and

Filiatrault, 2002]. The model can then be used to evaluate performance of structures

having other configurations.

The numerical model, SAWS (Seismic Analysis of Woodframe Structures),

developed as part of the CCWP, was used in this study to investigate the performance

of shearwalls in complete structures. The SAWS program was described in Chapter 3.

162

Nonlinear dynamic time-history analysis was conducted using the SAWS program and

the suite of 20 ordinary ground motion records, as was used previously. Two structural

configurations were considered, a one-story and a two-story structure. These two

structures are described in the following sections.

5.1.1 Model configuration of one-story residential structure

The model of the one-story single-family residential structure was developed

to be representative of typical southern California construction (i.e., a "bungalow"

style house). The plan of this structure was 32 ft. x 20 ft. and the structure had

openings for pedestrian doors and windows. The shearwalls in the structure were built

using 3/8-in. OSB, attached to the framing using the Durham spiral nail (2 in. long x

0.105 in. diameter). In most cases, a 6"/12" nailing schedules was used. The top-plate

and end studs were double members, while the sole-plate and the interior studs were

single members. The framing members were nominal 2 in. x 4 in. spaced (in most

cases) at 24 in. on-center. Properly installed hold-downs were assumed to be present.

Nonstructural finish materials (1/2-in. gypsum wallboard and 7/8-in. stucco) were

assumed to be properly attached. The plan and section views are shown in Figure 5.1.

Figure 5.2 presents elevation views of each exterior wall in the one-story structure.

The information in this figure was used to develop the global hysteretic parameters for

each shearwall using CASHEW and the Durham nail parameters (see Table 4.1).

4 ft.

20ft. 4ft.

4 ft.

4 ft.

H'

ft 6ft. 4ft. 4ft. 4ft. 8ft. 4ft.

rhroomBedroom

Kitchen

Living Room

I BedroomLi IIJ4ft. 8ft. 4ft. 3ft. 3ft. 8ft. ft4 4 444

32 ft.4

124L

/8 Stucco/8" OSB

2x4 @24in o.c1/2" Gypsum Wallboard

32ft.

4ft

3 ft.

2 ft.

3ft.

3 ft.

20ft.

163

Roofing (3-ply with gravel)/8" Plywood

2x6 ©l6in o.c51/ Fiberglass Loose Insulation1/2" Gypsum Wallboard

Figure 5.1 Plan view and section view for the one-story house model

8ft.

-ft,

4ft. 8ft. 4ft. 3ft. 3ft. 8ft. 2ft.

*-4---ø

32ft.

East Wall (EW)

ii

1.3 ft

4 ft.

11.1

2ft 6ft. 4fL 4ft. 4ft 8ft. 4ft.

3ft.%AI.._LlAI....II/AflAI\VVOL VV4II VVVVJ

1.311

4ft

2.7 ft.

4ft. 4ft. 4ft.

-44ft. 4ft.

I

20ft.

C'..... ..L AI._.II I('lAI'...J'.JL.1 LI I V V II V VJ

1.3ft

4ft

2.7 ft.

===

2ft 3ft. 4ft. 3ft. 2ft 3ft. 3ft.'I - *-- -4-pI -p *-*

20ft.

KI.-....4L AI....II 11.IIAI\I'l'.JILII VVQIIkINVV1

ft.

ft.

Figure 5.2 Detailed wall configurations for the one-story house model

ft.

ft.

164

165

5.1.2 Model configuration of two-story residential structure

The model of the two-story single-family residential structure was based on the

model in Task 1.1.1 of the CCWP [Fischer et al., 2001]. The structure has plan

dimensions of 20 ft. x 16 ft. and has various openings for a garage door, a pedestrian

door, and windows. The shearwalls in the structure were built using 3/8-in. OSB,

attached to the framing using 8d box nails (2.5 in. long x 0.113 in. diameter) at 6"/12"

(edge/field). A 3"/12" nailing schedule was used for the shearwalls on either side of

the garage door opening. Hold-downs were assumed to be properly installed.

Nonstructural finish materials (1/2-in. gypsum wallboard and 7/8-in. stucco) also were

assumed to be properly attached. The elevation and plan views are shown in Figure

5.3. This structure was design according to UBC '94 following typical construction

practices in California. A more detailed description of this structure can be found

elsewhere [Fischer et al., 2001].

166

'I 1F 9.5 mm ( in)OSB sheathing

I/ with8db@

L 4

lSOmm(6in)

'4.9m1

North -a------- HTT22 holdowns typ. at / North\XTZ+ Ai,ii ni+;,, shearwall element ends rat ui,it

ISNW

6.lm (20 F)

North and South Wall Elevations

80 man x 305 n

(3 in x 12 in)

Interior bearintwall

9.5 mm(in)OSB sheathingwith 8db @150mm (6 in)

9.5 mm (l in)OSB sheathingwith 8db@75 mm (3 in)

12.7 mm ( in) CDX plywood roofsheathing with 8db @ 150mm (6 in)

9.5 mm in) OSB sheathingwith 8db @150mm (6 in)

4.9 m (16 ft)

Roof trusses@ 610mm(24 in) o.c.

0

H(2 x 10) floor joists / \ Floor openingii

@406mm (16 in) / i belowo.c. lapped over \_/ _ / \

38 mm x 235 mm \_ ) / \_/11

_____ GLB & bearing wall _ \_/Floor sheathed with 19.0mm ( in) T & G Interior partition walls, typ.plywood with 10db @ 150mm (6 in)

0.9 m (3 ft) pedestrian doorNorth North

Plan View 2nd Story Plan View

Figure 5.3 Elevation and plan view for two-story house model (from: Fischer et al.,2001)

167

5.2 Shearwall performance in complete structures

5.2.1 One-story structure

5.2.1.1 Performance of shearwalls with OSB only

The performance of shearwalls in a one-story structure was investigated using

the model described in Section 5.1.1. Based on assumed weight tributary to the roof

diaphragm and wall dead load, a calculated total seismic weight of 15040 lb. was

assigned at the roof level and an equivalent viscous damping of 1% of critical was

assumed.

The CASHEW program was used to develop global hysteretic parameters for

each shearwall assuming the Durham nail parameters and considering each specific

wall configuration. Table 5.1 presents the resulting hysteretic parameters for each

shearwall (OSB only) in the one-story structure. As done previously, seismic zone IV

(LA) and soil profile type D (SD) were assumed. Figure 5.4 presents the SAWS model

for this one-story structure (OSB only) composed of four zero-height nonlinear

shearwall spring elements.

Spring K0F0 F1

Element II

(kips/ r1 r2 r3 r4(kips) (kips) (in.)

aii

in.)

East_Wall18.52 0.079 -0.031 1.298 0.066 7.35 1.51 2.32 0.76 1.10

West_Wall21.52 0.080 -0.046 1.297 0.066 8.36 1.74 2.32 0.75 1.10

sy'15.58 0.083 -0.079 1.285 0.068 5.61 1.19 2.32 0.73 1.09

South_Wall

10.63 0.051 -0.046 1.383 0.068 5.56 1.00 2.92 1.00 1.12North_Wall

Table 5.1 Hysteretic parameters for the shearwall spring elements in one-Storystructure, OSB only

It,rsl

Figure 5.4 SAWS model of the one-story structure, OSB only

The peak displacement distributions for each shearwall are shown in Figures

5.5 through 5.7 for the three hazard levels (JO, LS, and CP). The performance of all

shearwalls is well below the drift limit at the low hazard level (JO, 50/50), however

the South wall (SW) and North wall (NW) performed less well inthe high hazard level

(CP, 2/50) because of the many openings and the relatively small number of fasteners

(6"/12" nailing schedule). The East wall (EW) and West wall (WW) performed well at

all hazard levels.

1

0.9

0.8

0.7

0.6

0.5

0.4

0.3

0.2

0.1

n

WW//EW SW

J+ ' NW

!/

Ii'

01

// J"'o o Structure Type: One Story (32 >< 20')

// <Nailing Schedule: 8d©6"-12"

/ / Sheathing: OSB (/8")

[IShearwall HP: Durham + CASHEW

II ( NSF: None/o Damping: 1%

/ Hazard Level: 10 (50%I5Oyrs)

- - W15040 lb. total

169

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8

max (in.)

Figure 5.5 Peak displacement distributions for shearwalls in one-story structure, OSBonly (10, 50/50 hazard level)

0.9

0.8

0.7

0.6

0.5

0.4

0.3

0.2

0.1

n

--;--

EWSW

/NW/°

IStructure Type: One Story (32' x 20')I Nailing Schedule: 8d@6"-12"

/ Sheathing: OSB (I")Shearwall HP: Durham + CASHEW

0 0/ NSF: None4/ : / Damping: 1%

4// Hazard Level: LS (10%I5Oyrs)

- W = 15040 lb. total

0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2

ömax (in.)

Figure 5.6 Peak displacement distributions for shearwalls in one-story structure, OSBonly (LS, 10/50 hazard level)

1

0.9

0.8

0.7

0.6

0.5U-

0.4

0.3

0.2

0.1

0

0

EW

ww

//

4/Structure Type: One Story (32 >< 20)Nailing Schedule: 8d@6"-12"

/ Sheathing: OSB (/8')

i/ / Shearwall HP: Durham + CASHEW

p /NSF: NOneDamping: 1%

I / 9 Hazard Level: CP (2%I5Oyrs)W= 15040 lb. total

1 2 3 4 5 6

6max (in.)

170

Figure 5.7 Peak displacement distributions for shearwalls in one-story structure, OSBonly (CP, 2/50 hazard level)

5.2.1.2 Performance of shearwalls with NSF materials

The performance of shearwalls with nonstructural finish materials (stucco and

gypsum wallboard) in the one-story structure also was investigated. Two cases were

considered: (1) walls with OSB and gypsum wallboard, and (2) walls with OSB,

gypsum wallboard and stucco. Based on calculation, and assuming weight tributary to

the roof diaphragm and wall dead load, a total seismic weight of 16793 lb. was

estimated for the structure with OSB and gypsum walls, while a total seismic weight

of 20952 lb. was estimated for the structure with OSB, gypsum wallboard, and stucco.

(Stucco was assumed to have a weight of 10 psf.) Equivalent viscous damping of 1%

of critical was assumed in both cases.

171

The hysteretic parameters for the stucco and gypsum wallboard were based on

available experimental test data and were adjusted for the length of the wall. In the

case of the partition walls, gypsum wallboard was attached on both sides, and it was

assumed that the stiffness and strength was twice that of a single side of gypsum

wallboard [Folz and Filiatrault, 2002]. The resulting hysteretic parameters for each

shearwall (with NSF materials) in the one-story structure are shown in Table 5.2. The

hysteretic parameters for the OSB-only shearwall (Table 5.1) can be used for the

OSB-only walls use in Table 5.2. Each subscript number corresponds to a layer in the

shearwall and subscripts x and y indicate a direction. If only gypsum wallboard is

considered (i.e., OSB + GWB), the hysteretic parameters for stucco (Sx1, SX8, Sy, and

Sy9) in Table 5.2 are eliminated. The SAWS model for the one-story structure (with

NSF materials) is shown in Figure 5.8. This structure is composed of 17 zero-height

nonlinear shear spring elements, one each for: four OSB only layers, four stucco

layers, and nine gypsum wallboard layers. If only the gypsum NSF materials are

considered, the four stucco layers are removed. Figure 5.9 shows the SAWS model

considering the gypsum NSF materials only (note fewer springs).

172

Spring Wall Type K0 r

AUElement & Location (ks/ r1 r2 r3 r4

(kips)(kip

(in.)a 13

SXIStucco

46.39 0.058 -0.050 1.000 0.020 2.92 0.44 0.96 0.60 1.10East Wall

S>2OSB

18.52 0.079 -0.031 1.298 0.066 7.35 1.51 2.32 0.76 1.10East Wall

GWB (I Side)24.13 0.029 -0.017 1.000 0.005 1.30 0.29 1.54 0.80 1.10

East_WallGWB (2 Sides)

44.54 0.029 -0.017 1.000 0.005 2.40 0.54 2.83 0.80 1.10Partition_Wall

GWB (2 Sides)29.69 0.029 -0.017 1.000 0.005 1.60 0.36 1.89 0.80 1.10

Partition_WallGWB (I Side)

25.98 0.029 -0.017 1.000 0.005 1.40 0.31 1.65 0.80 1.10West_Wall

SX7OSB

21.52 0.080 -0.046 1.297 0.066 8.36 1.74 2.32 0.75 1.10West Wall

SX8Stucco

49.96 0.058 -0.050 1.000 0.020 3.15 0.47 1.03 0.60 1.10West Wall

5Y1Stucco

42.83 0.058 -0.050 1.000 0.020 2.70 0.40 0.89 0.60 1.10South Wall

SY2OSB

15.58 0.083 -0.079 1.285 0.068 5.61 1.19 2.32 0.73 1.09South WallGWB (1 Side)

22.27 0.029 -0.017 1.000 0.005 1.20 0.27 1.42 0.80 1.10South_WallSy4 GWB (2 Sides)

& Sy Partition Wall 25.98 0.029 -0.017 1.000 0.005 1.40 0.31 1.65 0.80 1.10

GWB (2 Sides)37.12 0.029 -0.017 1.000 0.005 2.00 0.45 2.36 0.80 1.10Partition_Wall

GWB (1 Side)20.41 0.029 -0.017 1.000 0.005 1.10 0.25 1.30 0.80 1.10North_Wall

S\8OSB

North Wall10.63 0.051 -0.046 1.383 0.068 5.56 1.00 2.92 1.00 1.12

StuccoNorth_Wall

39.26 0.058 -0.050 1.000 0.020 2.47 0.37 0.81 0.60 1.10

Table 5.2 Hysteretic parameters for the shearwall spring elements in one-storystructure, OSB and NSF materials

173

'X6i I

sx7 I

VI I

__________

Exterior sheanivall (OSB, GWB, stucco)

Window or doqr

Interior partitio wall (GWB on both sides)

s3

VVVS1sY2 SY4 SY5 S6 Sy8

$y3 Sy9

Figure 5.8 SAWS model of the one-story structure, OSB and NSF materials (GWBand Stucco)

Sx6

VI I

------

Exterior shear'vall (OSB, GWB)

Window or dock

Interior partitio1 wall (GWB on both sides)

4.sY1 sY6

sY2 sY3 sY4 sY5 sY7

Figure 5.9 SAWS model of the one-story structure, OSB and GWB

174

Figures 5.10 through 5.12 present the peak displacement distributions for each

shearwall (OSB + gypsum wallboard) for the three different hazard levels (TO, LS, and

CO). The distributions for the wall with NSF materials (i.e., OSB + gypsum wallboard

+ stucco) are shown in Figures 5.13 through 5.15. As expected, the performance of the

shearwalls with NSF materials is better than OSB-only walls at all hazard levels. The

addition of stucco dramatically improves the shearwall performance. This also was

noted in Section 4.2.3, considering isolated shearwalls. This is especially evident at

the highest seismic hazard level.

1

0.9

0.8

0.7

0.6

05U-

0.4

0.3

0.2

0.1

n

-

/ SW

.1 NW

i/i!

j I Structure Type: One Story (32' x 20')

// / Nailing Schedule: 8d@6"-12"

/ / Sheathing: OSB (3/..)

/ / Shearwall HP: Durham + CASHEWrI / NSF: GWB1/ / Damping 1%J Hazard Level: 10 (50%/5Oyrs)

W = 16793 lb. total

0 0.05 0.1 0.15 0.2 0.25 0.3

6max (in.)

Figure 5.10 Peak displacement distributions for shearwalls in one-story structure, OSB+ GWB (TO, 50/50 hazard level)

0.9

0.8

0.7

0.6

0.5U-

0.4

0.3

0.2

0.1

n

:

/SW

NW

Structure Type. One Story (32 x 20)Nailing Schedule: 8d@6"-12"

Li.. / Sheathing: OSB (/")

I ? / Shearwall HP: Durham + CASHEW

fJ NSF:GWB

1 / Damping: 1%hi Hazard Level: LS (10%/5oyrs)

W = 16793 lb. total

175

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8

6max (in.)

Figure 5.11 Peak displacement distributions for shearwalls in one-story structure, OSB+ GWB (LS, 10/50 hazard level)

0.9

0.8

0.7

0.6

0.5

0.4

0.3

0.2

0.1

n

SW

I/ NW

IIJ1

ft Structure Type: One Story (32' x 20)1/ / Nailing Schedule: 8d@6"-12"II / Sheathing: OSB (/8")

/1Shearwall HP: Durham + CASHEWNSF: GWB/ Damping: 1%Hazard Level: CP (2%I5Oyrs)

- - W= 16793 lb. total

0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2

max (in.)

Figure 5.12 Peak displacement distributions for shearwalls in one-story structure, OSB+ GWB (CP, 2/50 hazard level)

0.9

0.8

0.7

0.6

U-

0.4

0.3

0.2

0.1

n

1/ /

i/i'NW

I,/ Structure Type. One Story (32 x 20)If / Nailing Schedule: 8d©6-1 2

/ Sheathing: OSB ('")Shearwall HP: Durham + CASHEW

/ NSF: GWB + Stucco/ Damping: 1%

Hazard Level: 10 (50%/5oyrs)W = 20952 lb. total

176

0 0.04 0.08 0.12 0.16 0.2

6max (in.)

Figure 5.13 Peak displacement distributions for shearwalls in one-story structure, OSB+ GWB + Stucco (JO, 50/50 hazard level)

0.9

0.8

0.7

0.6

0.5U-

0.4

0.3

0.2

0.1

()

:/NW

r:/

q: ///

x Structure Type: One Story (32 x 20)

A.Nailing Schedule: 8d@6"-12"Sheathing: OSB

/ ; /

7/, /

(/8")

Shearwall HP: Durham + CASHEW

/1 /

/

NSF: GWB + StuccoDamping: 1%

/ Hazard Level: LS (10%/S0yrs)W = 20952 lb. total

0 0.1 0.2 0.3 0.4 0.5 0.6

6max (in.)

Figure 5.14 Peak displacement distributions for shearwalls in one-story structure, OSB+ GWB + Stucco (LS, 10/50 hazard level)

0.9

0.8

0.7

0.6

0.5

0.4

0.3

0.2

0.1

C)

H

:/ NW

/4

II

Ii .' /// Structure Type: One Stery (32 x 20)

Nailing Schedule: 8d@6-12"I! / Sheathing: OSB (I")

*1I Shearwall HP: Durham + CASHEW/ NSF: GWB + Stucco

f x' Damping: 1%

f Hazard Level. CP (2 /ol50yrs)W 20952 lb. total

177

0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6

max (in.)

Figure 5.15 Peak displacement distributions for shearwalls in one-story structure, OSB+ GWB + Stucco (CP, 2/50 hazard level)

5.2.2 Two-story structure

5.2.2.1 Performance of shearwalls with OSB only

The SAWS program was used to investigate the performance (peak

displacement) of shearwalls in a two-story structure under actual earthquake loading,

using a suite of 20 ordinary ground motions to characterize the seismic hazard. A total

seismic weight acting on this structure of 24730 lbs. was estimated, with 13938 lbs.

applied to the second floor diaphragm and 10792 lbs. applied to the roof diaphragm.

Equivalent viscous damping of 1% of critical in the first and second modes of

vibration was assumed. This value of viscous damping is consistent with other studies

[Foliente, 1995; Folz and Filiatrault, 2002].

178

As described in Chapter 3, the SAWS program requires several input

parameters including the global hysteretic parameters for each shearwall, seismic

weights, viscous damping parameters, integration time-step, and input ground

acceleration parameters. The CASHEW program was used to determine the global

hysteretic parameters for each shearwall, using the Durham nail parameters. The

resulting sets of hysteretic parameters for each shearwall in the two-story structure are

shown in Table 5.3 [from: Folz and Filiatrault, 2002].

The peak displacements of each shearwall in complete structure were obtained

using SAWS program. In the discussion of the results in this section, peak

displacements are measured from the bottom of the first-story wall either to the top of

the first-story wall or to the top the second-story wall. Three hazard levels (JO, LS,

and CP) were considered, and seismic zone IV (LA) and soil profile type D (SD) were

assumed. Figure 5.16 illustrated the SAWS model for this structure (with OSB only).

This structure is composed of eight zero-height nonlinear shearwall spring elements

and two rigid diaphragms: one for the second floor and one at the roof level.

The resulting peak displacement distributions (peak displacement measured

relative to ground) for each shearwall are shown in Figures 5.17 through 5.19 for the

three different hazard levels (JO, LS and CP). Since the South and North shearwalls

(both 1SNW and 2SNW) have the same configuration for both stories, their

performance was identical. The performance (peak drift relative to ground level) of the

East shearwall (2EW) in the complete structure, located above the wall with the

garage door opening, performs the worst. However, as seismic demand increases, the

179

peak displacement distribution for the East wall (2EW) and West wall (2WW) in the

complete structure show more similar performance (see Figure 5.19).

K0 iSpring(kips/ r1 r2 r3 r4

F0 I F aElement (kips) (kips) (in.)I

J in.) I

S1 Level 116.73 0.083 -0.088 1.00 0.030 8.23 1.88 3.44 0.79 1.07

East_WallSx2 Level 1

22.21 0.064 -0.056 1.07 0.030 8.25 1.98 2.28 0.87 1.11West_WallSy1 Level 1

32.49 0.065 -0.074 1.10 0.030 10.88 2.43 2.39 0.81 1.09South_WallSy Level 1

32.49 0.065 -0.074 1.10 0.030 10.88 2.43 2.39 0.81 1.09NorthWallSx3Level2 11.99 0.069 -0.038 1.16 0.020 4.41 1.07 3.02 0.77 1.10East_Wall

S4 Level 211.99 0.069 -0.038 1.16 0.020 4.41 1.07 3.02 0.77 1.10

West_WallSy3 Level 2

19.13 0.054 -0.060 1.10 0.030 7.94 2.90 2.91 0.84 1.09South_WallSy4 Level 2

19.13 0.054 -0.060 1.10 0.030 7.94 2.90 2.91 0.84 1.09North_Wall

Table 5.3 Hysteretic parameters for the shearwall spring elements, OSB sheathingonly (from: Folz and Filiatrault, 2002)

secono uoor oiapnragm

Exterior shearwall (OSB)

Window or door

Exterior shearwall (OSB)

Window or door

Figure 5.16 SAWS model of the two-story structure, OSB only (from: Folz andFiliatrault, 2002)

0.9

0.8

0.7

0.6

0.5LL

0.4

0.3

0.2

0.1

(1

/I.L

/1SNW :x

1WW

1 / 1EW/,2WII 2SNW 2EWii I

a

III,

I/x'

S / //

Structure Type: Two Story (20 x 16)Nailing Schedule: 8d@6"/2" and 3/12"

/

/Sheathing: OSB (/")

/ / /Shearwall HP: Durham + ASHEW

J // I,'

NSF:None

/Damping: 1%

II Hazard Level: 10 (50%/50 rs)W = 24730 lb. total

0 0.2 0.4 0.6 0.8 1 1.2

max (in.)

181

Figure 5.17 Peak displacement (relative to ground) distributions for shearwalls in two-story structure (TO, 50/50 hazard level)

0.9

0.8

0.7

0.6

LL

0.4

0.3

0.2

0.1

1SNW

:1

1EW

,1WW

/, 2SN!'

I; //1'

II)//:x

ii

//I'j< Structure Type: Two Story (20 x 16)

//I . Nailing Schedule: 8d@6"/12" and 3/12'

/ Sheathing: OSB (/8")

I' Shearwall HP: Durham + CASHEW

/i NSF: NoneDamping: 1%/' Hazard Level: [S (10%/5oyrs)y} W = 24730 lb. total

0 0.5 1 1.5 2 2.5 3 3.5 4

6max (in.)

Figure 5.18 Peak displacement (relative to ground) distributions for shearwalls in two-story structure (LS, 10/50 hazard level)

1

0.9

0.8

0.7

0.6

0.5U-

0.4

0.3

0.2

0.1

0

0

I SNW

1 WW2 EW

1EW /"/

/'2WW

IA:

[ JR /ff q

// Structure Type: Two Story (20' x 16)

// / ./ Nailing Schedule: 8d@6Y12" end 3/12"


7/ / Shearwall HP: Durham CASHEW

7/ / 7 NSF: NoneI/ / 1 Damping: 1%

Hazard Level: CP (2%/5pyrs)J' _/_. - W = 24730 lb. total

1 2 3 4 5 6 7 8

max (in.)

182

Figure 5.19 Peak displacement (relative to ground) distributions for shearwalls in two-story structure (CP, 2/50 hazard level)

5.2.2.2 Performance of shearwalls with NSF materials

The performance of shearwalls with nonstructural finish (NSF) materials such

as stucco and gypsum wallboard also was investigated using the SAWS program.

Hysteretic parameters for stucco and gypsum wallboard developed by Folz and

Filiatrault (2002) were used; these are shown in Table 5.4. These parameters were

developed from experimental test data obtained as part of the CCWP and CoLA

project [Gatto and Uang, 2002; Pardoen, 2001], and have been adjusted for the length

of the walls and presence of door and window openings in the two-story structure

considered in this study. Gypsum wallboard is assumed to be attached to the interior of

the wall, while stucco was assumed to be applied to the exterior. The case of an

interior partition wall with gypsum wallboard on both sides also was considered. In

183

that case, it was assumed that the two layers of gypsum wallboard have twice the

strength and stiffness of one layer of gypsum wallboard. The hysteretic parameters for

the OSB sheathing are the same as those shown in Table 5.3 [Folz and Filiatrault,

2002].

HystereticParameters

K0

(kips/in.) r1 r2 r3 r4F0

.

(kips)F1

(kips)A

(in.)a

Stucco 5.00 0.058 -0.050 1.00 0.020 8.00 1.20 15.0 0.60 1.10GWB 2.60 0.029 -0.017 1.00 0.005 3.56 0.80 24.0 0.80 1.10

Table 5.4 Fitted hysteretic parameters for the SDOF shear element model of an 8 ft. x8 ft. shearwall with stucco and gypsum wallboard (from: Folz and Filiatrault, 2002)

The hysteretic parameter sets for each of the shearwalls having NSF materials

(for use in the analysis of the complete structure) are shown in Table 5.5. Again, in

this Table, subscripts x and y indicate the direction while the subscript number

indicates a given layer in the shearwall. Figure 5.20 presents the SAWS model

composed of 27 zero-height nonlinear shear spring elements corresponding to the

eight OSB layers, eight stucco layers, and eleven gypsum wallboard layers.

Figures 5.21 through 5.23 present the peak displacement distributions for each

of the shearwalls with the NSF materials for the three different hazard levels (10, LS,

and CP). Again, the performance of South and North shearwalls (both stories) is

identical since they have the same configuration. The performance of the East

shearwall second story (2EW), which is located above the wall with the large garage

door, exhibits the worst performance. This result was also observed in the previous

section (considering walls with OSB only). As expected, the shearwalls with the NSF

materials acting as part of the complete structure perform very well relative to the bare

shearwall. Comparing Figures 5.19 and 5.23 (CP, 2/50 hazard level), one sees that the

performance of walls with NSF materials is much better than that of walls with OSB

only.

Spring Wall Type K0 IF0 F1

r

A,,r

I

Element & Location (kips r1 r2 r3 r4(kips) (kips) (in.) I

Stucco21.41 0.058 -0.050 1.00 0.030 1.35 0.22 0.44 0.60 1.10Level_I_(EW)

SX2OSB

Level I (EW) 16.73 0.083 -0.088 1.00 0.030 8.23 1.88 3.44 0.79 1.07

GWB (1 Side)11.13 0.029 -0.017 1.00 0.005 0.60 0.13 0.72 0.80 1.10Level_l(EW)

SX4GWB (2 Sides)

22.27 0.029 -0.017 1.00 0.005 1.20 0.27 1.45 0.80 1.10Levell(PW)

S5Stucco

46.42 0.058 -0.050 1.00 0.030 2.92 0.44 0.96 0.60 1.10Level 1 (WW)

5X6OSB

Level I (WW) 22.21 0.064 -0.056 1.07 0.030 8.25 1.98 2.28 0.87 1.11

SX7GWB (1 Side)

24.15 0.029 -0.017 1.00 0.005 1.30 0.29 1.54 0.80 1.10Level 1 (WW)Sy1 Stucco

& Sv Level I (SNW) 49.96 0.058 -0.050 1.00 0.030 3.15 0.47 1.04 0.60 1.10

SY2 OSB& SY5 Level 1 (SW) 32.49 0.065 -0.074 1.10 0.030 10.88 2.43 2.39 0.81 1.09

Sy3 GWB (1 Side)& Svo Level 1 (SNW) 25.98 0.029 -0.017 1.00 0.005 1.40 0.31 1.65 0.80 1.10

Sxs Stucco& SXI3 Level 2 (EVoW)

25.01 0.058 -0.050 1.00 0.030 1.57 0.24 0.52 0.60 1.10

SX9 OSB& SX14 Level 2 (EW) 11.99 0.069 -0038 1.16 0.020 4.41 1.07 3.02 0.77 1.10

SXIO GWB (I Side)& SX15 Level 2 (E\VW)

13.02 0.029 -0.017 1.00 0.005 0.70 0.16 0.83 0.80 1.10

Sxi GWB (2 Sides)& SXI2 Level 2 (PW) 49.39 0.029 -0.017 1.00 0.005 2.65 0.60 3.15 0.80 1.10

Stucco& Sy Level 2 (SNW) 42.83 0.058 -0.050 1.00 0.030 2.70 0.40 0.89 0.60 1.10

Sy OSB& Sy11 Level 2 (SW) 19.13 0.054 -0.060 1.10 0.030 7.94 2.90 2.91 0.84 1.09

Sy GWB (I Side)&Syu Level 2 (SNW) 22.27 0.029 -0.017 1.00 0.005 1.20 0.27 1.42 0.80 1.10

Table 5.5 Hysteretic parameters for the shearwall spring elements, OSB and NSFmaterials (from: Folz and Filiatrault, 2002)

sx13I

sx14I

sx15

sx12r

-

I Partition Wall

L

Roof Diaphragmsx8

sx9

sx10

sY7 sylosy8 syllsY9 sY12

Connection to second __,,/floor diaphragm

sx6

sx7

S4 Partition Wall

I4Second Floor Diaphragm

sx1

sx2

sx3

syl sY4

sY2 sY5

sY3 sY6

185

Exterior shearwall (OSB, GWB, stucco)

Window or door

Interior partition wall (GWB on both sides)

L...:.:.:.:> N

Exterior shearwall (OSB, GWB, stucco)

Window or door

Interior partition wall (GWB on both sides)

Fixed support -/

Figure 5.20 SAWS model of the two-story Structure, OSB and NSF materials (from:Folz and Filiatrault, 2002)

0.9

0.8

0.7

0.6

0.5U-

0.4

0.3

0.2

0.1

0

- -/

/

/7/ 1 WW

/ '' 2SNW/

/IEW

I/ / /7 2WW

/ J //' "

2EW

I " i /

' Structure Type: Two Story (20 x 16')

/ / / /' Nailing Schedule: 8d@6"/12" and 3/12"

/ /11/

<

Sheathing: OSB (/8")

Shearwall HP: Durham + CASHEW// / /7 ' NSF: GWB + Stucco

4 /1L

/ Damping: lobHazard Level: 10 (50%/5oyrs)

>

-W = 24730 lb. total

0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4

max (in.)

I,1

Figure 5.21 Peak displacement (relative to ground) distributions for shearwalls in two-story structure (JO, 50/50 hazard level)

0.9

0.8

0.7

0.6

0.5U-

0.4

0.3

0.2

0.1

n

/

1SNW

/

//x'///- 2SNW

//1WW

/ / /

1 EW

I / / ;2WWti // ,' 2EW

/ Structure Type: Two Story (20 x 16')

/ /.I j1 Nailing Schedule: 8d@6"/12" and 3/12"

I L ' Sheathing: OSB (3/..)

I' / / /' Shearwall HP: Durham + CASHEW

/// / ' NSF: GWB + Stucco

/ , Damping: 1%

'j / / Hazard Level: LS (10%/5oyrs)} _/_____' W=24730 lb. total

0 0.2 0.4 0.6 0.8 1 1.2 1.4

max (in.)

Figure 5.22 Peak displacement (relative to ground) distributions for shearwalls in two-story structure (LS, 10/50 hazard level)

0.9

0.8

0.7

0.6

40.5

0.4

0.3

0.2

0.1

0

0

-

/ 1SNW

/i

/

1ww://

/

!7 1 EW

/ /

/ /1/ ,

//

//

/ Structure Type: Two Story (20 < 16)Nailing Schedule: 8d@6"/12" and 3/12"

/ / Sheathing: OSB (3/)

/

/ t

Shearwall HP: Durham + CASHEWNSF: GWB + Stucco

/

I / / Damping: 1%

/ Hazard Level. CP (2/o/5oyrs),/ _:::i-__-' - W = 24730 lb. total

0.5 1 1.5 2 2.5 3

6max (in.)

187

Figure 5.23 Peak displacement (relative to ground) distributions for shearwalls in two-story structure (CP, 2/50 hazard level)

5.2.3 Additional studies

5.2.3.1 Interstory displacement

The performance of shearwalls acting as assemblies in complete structures was

investigated in Section 5.2. In addition to peak displacement (e.g., total drift at the top

of a multistory structure), interstory drift also is a criterion used to evaluate structural

performance under lateral loading. In this section, the interstory drift in the two-story

structure was separated out from the total drift at the first and second stories. Because

all displacements are functions of time, and the structure may deflect according to its

first or second mode at any given time, the peak interstory drift does not necessary

correspond to the difference between the peak drifts at the first and second stories. The

two-story structure is the same as the structure described in Section 5.1.2, and effects

188

of NSF materials also were considered. Only the wall having the worst performance at

each story (i.e., both 1EW and 2EW) was considered in the comparisons made in this

section.

Figures 5.24 through 5.26 present the distributions for peak displacement at the

top of the first-story wall, peak displacement at the top of the second-story wall, and

the interstory displacement, for the three different hazard levels (JO, LS, and CP),

respectively. Since the interstory drift is the absolute value of displacement difference

between the second-story and the first-story at any given time in the displacement

time-history, similar relative displacement behavior is observed at all hazard levels.

Figures 5.27 through 5.29 present the different peak displacement distributions

for the three different hazard levels (JO, LS, and CP) for a structure built with NSF

materials (stucco and gypsum wallboard). The peak displacement is significantly

reduced by adding NSF materials to the OSB-only walls, and the effect is more

pronounced at the higher hazard levels. (This also was seen in Section 5.2.2.2.)

However, the relative magnitude of interstoiy drift is different from that seen with the

OSB-only walls. (Specifically, the interstory displacement distribution is lower than

that of the first-story shearwall.) This might be the effect of the gypsum wallboard and

stucco, providing additional stiffness as well as connection between the first and

second stories. (The stucco layer also serves to restrain sheathing nail head rotation

under cyclic (dynamic) loading.) The application of NSF materials to shearwalls

improves the overall performance of shearwall, as shown previously.

0.9

0.8

0.7

0.6

0.5

0.4

0.3

0.2

0.1

max first story

Interstory drift

ömax t second story

x'

Structure Type: Two Story (20 x 16')Nailing Sche1uIe: 8d@6"/12" and 3/12"

L Sheathing: OSB (I8")Shearwall HF?: Durham + CASHEWNSF: NoneDamping: 1%Hazard Level: 0 (50%/50yrs)} - - W = 24730 l. total

189

0 0.2 0.4 0.6 0.8 1 1.2 1.4

max (in.)

Figure 5.24 Comparison of peak displacements at first and second stories, OSB (JO,50/50 hazard level)

0.9

0.8

0.7

0.6

0.5

0.4

0.3

0.2

0.1

n

/ ' 6rnax at first story

A J--T/ I Interstory driftI -I

/I x,' max at second story

//

/t

i Structure Type. Two Story (20 x 16)

I I. Nailing Schedule: 8d@6"112" and 3/12'

/ I ' Sheathing: OSB (3I")

/ ,Shearwall HP: Durham + CASHEW

I ,' II NSF: None

/ / Damping: 1%

/ / Hazard Level: LS (10%/50yrs)

- - W = 24730 lb. total

0 0.5 1 1.5 2 2.5 3 3.5 4

6max (in.)

Figure 5.25 Comparison of peak displacements at first and second stories, OSB (LS,10/50 hazard level)

0.9

0.8

0.7

0.6

40.5U-

0.4

0.3

0.2

0.1

C)

// Hmax at first stow

/ Interstry drift

/ max at second story

// ': Structure Type: Two tory (20 x 16')

/ Nailing Schedule: 6d6"/12" and 3/12"

/ Sheathing: OSB (/")X Shearwall HP: Durha + CASHEW

NSF: NoneDamping: 1%

- Hazard Level: CP (2%I5Oyrs)

- - W = 24730 lb. total

190

0 1 2 3 4 5 6 7 8

5max (in.)

Figure 5.26 Comparison of peak displacements at first and second stories, OSB (CP,2/50 hazard level)

0.9

0.8

0.7

0.6

40.50.4

0.3

0.2

0.1

(1

I /-/ Interstory drift:

/::: :

/Structure Type: Two Story (20' x 16')

L Nailing Schedule: 8d@6"/12" and 3/12"


/ / Shearwall HP: Durham + CASHEW

/ / NSF: GWB + Stucco/ ' Damping: 1%

-Hazard Level: 10 (50%/50yrs)

J / - W = 24730 lb. total

0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4

max (in.)

Figure 5.27 Comparison of peak displacements at first and second stories, OSB +GWB + Stucco (JO, 50/50 hazard level)

191

0.9

0.8

0.7

0.6

0.5

0.4

0.3

0.2

0.1

n

Interstory drift

6max at first story

/ xmax at second story

Structure Type: Two Story (20 x 16)/ >

Nailing Schedule: 8d©6"/12" and 3/12"

/ Sheathing: OSB (/8")f Shearwall HP: Durham + CASHEW

/ NSF: GWB + Stucco/ , Damping: 1%I >

Hazard Level: LS (10%I5Oyrs)

- - W=247301b. total

0 0.2 0.4 0.6 0.8 1 1.2 1.4

6max (in.)

Figure 5.28 Comparison of peak displacements at first and second stories, OSB +GWB + Stucco (LS, 10/50 hazard level)

0.9

0.8

0.7

0.6

0.5U-

0.4

0.3

0.2

0.1

n

7 /_* /+ ,'<

t

Interstory drift

6max at first story

/: 6max at second story

/ ,< Structure Type: Two Story (20 x 16)

/ Nailing Schedule: 8d@6"/12" and 3/12"

/ ' Sheathing: OSB (/8")

,Shearwall HP: Durham + CASHEW

/ " NSF: GWB + StuccoDamping: 1%

+Hazard Level: CF (2%/5Oyrs)

- - W 24730 lb. total

0 0.5 1 1.5 2 2.5 3

ömax (in.)

Figure 5.29. Comparison of peak displacements at first and second stories, OSB +GWB + Stucco (CP, 2/50 hazard level)

192

5.2.3.2 Effect of partition walls

Gypsum wallboard is a common material used to sheath partition walls that

divide space in a building. Typical partition walls are constructed with gypsum

wallboard attached to both sides of the wall framing using mechanical fasteners

(drywall screws). The partition walls usually are treated as nonstructural elements in a

building (i.e., they are excluded in a structural analysis or in the design of the primary

shearwalls), however they may contribute to the overall structural performance. This

was investigated using the SAWS model of the one-story structure described in

Section 5.2.1.

The SAWS model of the one-story structure without partition walls is shown in

Figure 5.30. The dimensions of this structure are the same as shown in Figure 5.1.

This structure is composed of eight zero-height nonlinear shearwall spring elements

(shown), one for each wall layer. Gypsum wallboard was assumed to be used an all

interior walls.

Nonlinear dynamic time-history analysis was performed using the SAWS

model to investigate the contribution of the partition walls to peak displacement. The

peak displacement distributions of the structure without partition walls are shown in

Figures 5.31 through 5.33 for the three hazard levels (TO, LS, and CP), respectively.

Comparing these results with these shown in Figures 5.10 through 5.12 (with partition

walls), one sees that the partition walls significantly influence the shearwall

performance at all hazard levels. The results in Figures 5.10 through 5.12 are also

shown as light lines in Figures 5.31 through 5.33 to allow for easy comparison. While

193

the displacements of the North wall (NW) in the one-story structure with partition

walls are well below the drift limit, those for the same wall without partition walls are

above the drift limit. This also can be seen in Figures 5.34 through 5.36 which show

comparisons of peak displacement distributions considering different NSF materials

and the effect of partition walls for the North wall (NW). As expected, the OSB-only

shearwalls (without partition walls) exhibited the worst performance, while the

shearwalls with NSF materials (stucco and gypsum wallboard, and with partition

walls) performed considerably better. The worst-case wall performance was improved

even further when the partition walls were considered. All shearwalls (with or without

NSF materials) analyzed with consideration of partition walls perform well below the

drift limit at JO (50/50) and LS (10/50) hazard levels. Also, the variability in peak wall

displacement is reduced when the effect of partition walls is considered in the

analysis. This was observed at all hazard levels.

194

Figure 5.30 SAWS model of one-story structure without partition walls, (OSB +GWB)

1

I.

0.5

0.4

0.3

0.2

0.1

ru

-

/// /7//

/

sw

/

NW

with itonWL /

Ii!!

/ /

I! // Structure Type: One Stow (32 20)I /

/

// / Nailing Schedule: 8d@6'-12"1/ / Sheathing: OSB (3//)

/ // Shearwall HP: Durham + CASHEW

/ //

' /7

/ / ,//7 / NSF: GWB (without Partition Wall)

/ Damping: 1%/ Hazard Level: 10 (50%ISOyrs)

/ 1/-, -..-

' / W16793 lb. total--.---- .-___ ____0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4

max (in.)

Figure 5.31 Peak displacement distributions for one-story structure, OSB + GWB(without partition walls), JO (50/50 hazard level)

195

0.9

0.8

0.7

0.6

0.5

0.4

0.3

0.2

0.1

n

/7

/ Sw

/ WW /NW

/iI /

with t'arti'tiondias / Structure Type: One Story (32 x 20)ii / // / Nailing Schedule: 8d@6"-12"

/I / Sheathing: OSB (/")

// I Shearwall HP: Durham + CASHEW

/ / // / NSF: GWB (without Partition Wall)

/ / / Damping: 1%/7 / II / Hazard Level: LS (10%/5oyrs)

/// - W = 16793 lb. total

0 0.2 0.4 0.6 0.8 1 1.2 1.4

max (in.)

Figure 5.32 Peak displacement distributions for one-story structure, OSB + GWB(without partition walls), LS (10/50 hazard level)

0.9

0.8

0.7

0.6

0.5U-

0.4

0.3

0.2

0.1

n

/7/7

// //Ew/1

: // . NW1:11 WW,''

,"1

with aitiQ WaHs ," / Structure Type: One Story (32' x 20')

// // / Nailing Schedule: 8d@6"-12"

/ / Sheathing: OSB (/8")

// / Shearwall HP: Durham + CASHEW

// / /1/ NSF: GWB (without Partition Wall)

I / / Damping: 1%

/ Hazard Level: CP (2%I5Oyrs)J /J W=167931b.total

0 0.5 1 1.5 2 2.5 3 3.5 4

max (in.)

Figure 5.33 Peak displacement distributions for one-story structure, OSB + GWB(without partition walls), CP (2/50 hazard level)

I

0.9

0.7

0.6

0.5U-

0.4

0.3

0.2

0.1

ii

I OSB,'+ GWB + StØ'cco(with Partition Walls)

OSB + GVB(with Partition Walls)

I:/ OSBG\)/ (without P

/ /

SB only(without Partition Walls)

on Walls)

Structure Type: One Story (32' x 20')Nailing Schedule: 8d@6'-12"Shearwall HP: Durham + CASHEWDamping: 1%Hazard Level: 10 (50%I5Oyrs)Wall: North Wall (NW)

196

0 0.1 0.2 0.3 0.4 0.5 0.6

max (in.)

Figure 5.34 Comparison of peak displacement distributions for the effect of partitionwalls and NSF materials, (JO, 50/50 hazard level)

I

S.

I;0.7

0.6

0.5

0.4

0.3

0.2

0.1

00

OB+GWB+Stuc o(with Partition Wall)

II

OSB+GWB/(with Partiti n Walls)I

/ OSB+GWB(without Partition Walls

OSB only(without Partition Wa Is

Structure Type: One StoW (32' x 20')Nailing Schedule: 8d@6"-12"Shearwall HP: Durham + CASHEWDamping: 1%Hazard Level: LS (1 0%I50rs)Wall: North Wall (NW)

0.4 0.8 1.2 1.6 2

6max (in.)

Figure 5.35 Comparison of peak displacement distributions for the effect of partitionwalls and NSF materials, (LS, 50/50 hazard level)

1

0.9

0.8

0.7

0.6

0.5Li

0.4

0.3

0.2

0.1

00

+ GWB + S/cco1(with Partition Walls)

//

/OSB+GW /(with Parti(on Wa,}(s)

I: / /I: ///: //

OSB + GWB(without Partition WaIIs

OSB only(without Partition Walls)

Structure Type: One Story (32' x 20')Nailing Schedule: 8d@6'-12"Shearwall HP: Durham + CASHEWDamping: 1%Hazard Level: CF (2%/5oyrs)Wall: North Wall

1 2 3 4 5 6

max (in.)

197

Figure 5.36 Comparison of peak displacement distributions for the effect of partitionwalls and NSF materials, (CP, 2/50 hazard level)

5.2.3.3 Performance comparison for isolated wall and wall in one-story structure

Most experimental tests of wood shearwalls are performed on isolated

shearwall assemblies (with or without NSF materials), with solid walls (no openings)

being the most common assembly tested. Although some shake table tests of full-scale

structures have been performed recently, isolated shearwall assemblies remain the

most common test configuration used to evaluate the performance of wood shearwalls

under seismic loading.

Using the north wall (NW) in the model of the one-story structure (see Figure

5.1), the difference between performance of an isolated shearwall and the same

shearwall acting as part of a one-story structure was investigated. The seismic weight

acting on the isolated shearwall was assumed to be one-half of that on the full-scale

structure used in the Section 5.1.1. As described previously in Section 5.2.1.1, a set of

ten hysteretic parameters for the north wall in a one-story structure was obtained using

the CASHEW program and assuming the Durham nail parameters. The peak

displacement distributions were obtained using SASH! for the isolated wall and using

SAWS for the wall in the complete one-story structure.

Figures 5.37 and 5.38 present comparisons of the peak displacement

distributions for the isolated wall and the wall in the complete one-story structure for

the 10 and LS hazard levels, respectively. Only the wall having the worst performance

(North wall, NW) is considered here. The difference in peak displacement

distributions is relatively small at the JO hazard level, however increases as the hazard

level increases to the LS hazard level. (At the CP hazard level, most of the peak

displacements exceeded the drift limit of 3%, and so that figure is not included here.)

This suggests that consideration of the performance of the complete structure system

should be included in the design of wood shearwall assemblies, particularly at high

hazard level events. This might be able to be accomplished using a modification factor

(applied to peak drift, e.g.), however this factor may be very structure-dependent.

0.9

0.8

0.7

0.6

0.5

0.4

0.3

0.2

0.1

n

2;

Wall in System

Isolated Wall

Structure Type: One Story House (32 x 20)

Nailing Schedule: 8d@6"-12°Shearwall HP: Durham + CASHEWDamping: 1%Hazard Level: 10 (50%/50yrs)

199

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8

6max (in.)

Figure 5.37 Comparison of peak displacement distributions for isolated shearwall andshearwall in complete one-story structure (10, 50/50 hazard level)

0.9

0.8

0.7

0.6

0.5

0.4

0.3

0.2

0.1

n

Structure Type: One Story (32' x 20) /Nailing Schedule: 8d@6"-12"Shearwall HP: Durham + CASHEWDamping: 1%Hazard Level: LS (10%I5Oyrs)

Wall in system

/Isolated wall

/

/

/

/

0 0.5 1 1.5 2 2.5 3

ömax (in.)

Figure 5.38 Comparison of peak displacement distributions for isolated shearwall andshearwall in complete one-story structure (LS, 10/50 hazard level)

200

5.3 Performance-based design

5.3.1 Incremental dynamic analysis

Incremental dynamic analysis (IDA) was performed on isolated shearwalls

with and without openings in Section 4.4.1. The results could be used to help define

appropriate collapse limit state definitions. The same methodology is used in this

section to develop IDA curves for shearwalls in a two-story structure. The same suite

of 20 ordinary ground motion records was used as input to the nonlinear dynamic

time-history analysis. Seismic zone IV (LA) and soil profile type D (SD) were

assumed.

IDA curves were developed specifically for the East wall 2EW in the two-story

structure. This was the shearwall exhibiting the largest displacements (see Section

5.2.2). The suite of 20 ordinary ground motion records was divided into three groups

to make it easier to show the resulting IDA curves on a single figure. Only one set

(seven records) is shown here, however the results are representative.

Figure 5.39 shows Sa vs. peak displacement for shearwall 2EW in the two-

story structure. Also shown are the tangents defining the apparent break points and an

estimated mean value for those break points. A characteristic value could be used to

define the design drift limit for collapse prevention (CP, 2/50). The mean value of this

break point corresponds to a peak displacement of 5.56 in. or about 2.7% of the total

wall height. (Note that the FEMA 356 drift limit for collapse prevention (CP, 2/50) is

3% of the total wall height for wood shearwalls.) If NSF materials are considered

201

(Figure 5.40), the mean value of the IDA break point decreases to 4.01 in., or about

1.9% of the total wall height.

Similar analyses were performed for the wall with a pedestrian door opening

(2WW). There was no significant difference in the estimated collapse limit (IDA break

point) between wall 2EW with the garage door opening and wall 2WW with the

pedestrian door opening. The 3% drift limit suggested by FEMA 356 appears to

correlate well with the collapse limit determined by IDA for the wall without NSF

materials. However, if walls are built with NSF materials, the estimated collapse limit

decreases. Table 5.6 summarizes the estimated collapse limit (mean value) for the

shearwalls considered in this study.

IDA Set Sheathing Shearwall Mean, t1 OSB 2EW 5.56 in.2 OSB 2WW 5.63 in.3 OSB + Stucco + GWB 2EW 4.05 in.4 OSB + Stucco + GWB 2WW 4.15 in.

lable 5.6 Estimated collapse limit (from IDA) for shearwall in the complete two-storystructure

2.0

1.5

c,)

1.0

C')

0.5

202

0.0

0 1 2 3 4 5 6 7 8 9 10

Smax (h.)

Figure 5.39 Set of IDA curves for selected OSB-only walls with garage door opening(2EW)

2.0-

1.5

0)1.0

C')

0.5

0.0 r

0 1 2 3 4 5 6 7 8 9 10

max (1)

Figure 5.40 Set of IDA curves for selected OSB + NSF walls with pedestrian dooropening (2WW)

203

5.3.2 Fragility curves

Fragility curves which could be used for design as well as for post-disaster

condition assessments were developed for an isolated shearwall (BW1) in Section

4.4.2. In this section, fragility curves for shearwalls in representative one and two-

story residential structures are developed. The seismic demand (interface) variable is

the spectral acceleration, 5a Fragility curves of this type can be used either as design

aids or to assess risk consistency in current design provision.

5.3.2.1 Fragility curve for one-story structure

Fragility curves were developed for the North wall (NW) of the one-story

structure, which has a pedestrian door and windows as shown in Figure 5.1. The North

wall (NW) exhibited the worst displacement performance in the one-story structure

(see Figures 5.5 through 5.7). As in the previous section, the records were scaled to six

different hazard levels: 50% in 50 years (72-year MRI), 20% in 50 years (225-year

MRT), 10% in 50 years (474-year MRI), 5% in 50 years (975-year Mifi), 2% in 50

years (2475-year MRI), and 1% in 50 years (4795-year MRI). The procedure for

constructing the fragility curves is the same as was described in the previous section

(Section 4.4.2).

The fragility curves for the North wall (NW) sheathed with OSB only, are

shown in Figure 5.41. Drift limits of 1%, 2% and 3% of the total wall height were

considered. Figure 5.42 presents a comparison of fragility curves for the North wall

constructed with gypsum wallboard, but without consideration of interior partition

204

walls. Figure 5.43 presents the fragility curves for the North wall (NW) with different

combinations of finish materials, and with and without consideration of the partition

walls. As before, only the JO (50/50, 1% drift limit) performance level is shown here

since the other performance levels (LS, CP) result in very low failure probabilities for

the walls built with NSF materials. Figure 5.43 confirms that NSF materials (stucco

and gypsum wallboard) contribute significantly to the performance of shearwalls

acting as part of complete structures under earthquake loading. It also shows that

partition walls significantly influence the performance of shearwalls acting as part of a

complete structure and subject to earthquake loading.

0.9

0.8

0.7

0.6

d 0.5

0.4

0.3

0.2

0.1

n

/ /,"3%/ /

/ /

1/:"/ ' Structure Type: One Story (32' >< 20')

/ Nailing Schedule: 8d@6"-12"Sheathing: OSB (/8")

/ ' Shearwall HP: Durham + CASHEW/ ' NSF: NoneDamping: 1%

- - - W=150401b.total

0 0.5 1 1.5 2 2.5 3

Sa(g)

Figure 5.41 Fragility curves for the North wall (OSB only) in the one-story structure(without partition walls)

205

1

0.9

0.8

0.7

0.6

Q: 0.5

0.4

0.3

0.2

0.1

Structure Type: One Story (32 x 20)[Nailing Schedule: 8d@6-12" /Sheathing: OSB (I)Shearwall HP: Durham + CASHEW / /NSF: GWB (without Partition Walls)/ I /0 / L /0

Damping: 1% / "...W = 16793 lb. total

/ / 3%

I

/,,,,I /"I /

0 0.5 1 1.5 2 2.5 3

Sa(g)

Figure 5.42 Fragility curves for the North wall (OSB + GWB) in the one-storystructure (without partition walls)

I

0.9

0.8

0.7

0.6

0 o.s

0.4

0.3

0.2

0.1

n

OSB only(without Partition Wall

OSB + GWB(without Partition Wa ls)

/ B + GW + Stucco1(th Partition Walls)

I OSB+GWB/ / (with Partition Walls)

/

/ Stvucture Type: One Story (32 x 20)

/ Niling Schedule: 8d@d"-12"sheathing: OSB (3/5)

Shearwall HP: Durham + CASHEW

/ NSF: GWB + Stucco/ Damping: 1%/ / Hazard Level: 10 (50%I0yrs)

0 0.5 1 1.5 2 2.5 3

Sa(g)

Figure 5.43 Comparison of fragility curves for the North wali in the one-storystructure (JO, 50/50, 1% drift limit)

206

5.3.2.2 Fragility curve for two-story structure

Fragility curves were developed for the East wall (2EW) and West wall

(2WW) with a garage door and pedestrian door opening, respectively, in the two-story

structure described in Figure 5.3. Peak displacement distributions were obtained for

each hazard level and the probability of failure was determined non-parametrically as

the relative frequency of the peak displacement exceeding specified drift limits. The

records were scaled to six different hazard levels: 50% in 50 years (72-year mean

return period, or MRI), 20% in 50 years (225-year MRI), 10% in 50 years (474-year

Mifi), 5% in 50 years (975-year MRI), 2% in 50 years (2475-year MRI), and 1% in 50

years (4975-year MRT).

Figures 5.44 through 5.46 show the fragility curves for the East wall (2EW)

considering three different peak displacements: 3max at first story (relative to ground),

interstory drift, and max at second story (relative to ground). Drift limits of 1%, 2%

and 3% of the relevant wall height were considered. The first story has a height of 8 ft.

1 in., and the second story has a (total) height of 17 ft. 2 in. Considering the life safety

drift limit (2% of total wall height), the limit for the first story is 1.94 in., the limit for

the second story is 4.12 in., and the drift limit considering interstory drift is 1.94 in.

Figures 5.47 and 5.48 show the fragility curves for the West wall (2WW) in

the two-story structure considering two different peak displacements @max at the first

story and max at the second story). Drift limits of 1%, 2% and 3% of the wall height

were considered.

207

All of the fragility curves for the two walls with openings (garage door and

pedestrian door) in the two-story structure are shown for comparison in Figures 5.49

and 5.50 assuming the FEMA 356 drift limit (1%, and 2% for JO and LS,

respectively). In the plateau region of the response spectrum considering seismic zone

IV (LA) and SD soil profile type, the spectral acceleration Sa = 0.633g for immediate

occupancy (TO, 50/50) and Sa = 1.lg for life safety (LS, 10/50). In this example, the

fragility curves indicate very low probabilities of failure for these performance levels,

with the exception of interstory drift. Thus, interstory drift might be the most

appropriate (conservative) displacement criteria to consider in design.

1

0.9

0.8

0.7

0.6

0.5

0.4

0.3

0.2

0.1

n

1%

2%'1/ 3%

/ /

/ /

//

I I Structure Type: Two Story (20 x 16)

I / Nailing Schedule: 8d@6Il:2" and 3/12"

I / Sheathing: OSB (/8")

/ / Shearwall HP: Durham + CASHEW

/ I ' NSF: None

/ / - Damping: 1%./ .../- W = 24730 lb. total

0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5

Sa(g)

Figure 5.44 Fragility curve for wall with garage door opening, max (relative toground) at first story

1

0.9

0.8

0.7

0.6

o 0.5

0.4

0.3

0.2

0.1

ft

------

/ / ,,3%

/ /

/,,,,

/ :'

/ ' Structure Type: Two Story(20' x 16)

/ ,' Nailing Schedule: 8d©6"/12" and 3/12'

/ , Sheathing: OSB (/8")

/ ' Shearwall HP: Durham + CASHEW/ ' NSF: None

/ - Damping: 1%

- - W = 24730 lb. total

0 0.5 1 1.5 2 2.5 3

Sa(g)

Figure 5.45 Fragility curve for wall with garage door opening, interstory drift

1

0.9

0.8

0.7

0.6

ci 0.5

0.4

0.3

0.2

0.1

1/o/ /2%

3%

/ /,','

/ /

/ /,"/ / Structure Type: Two Story (20' x 16')

/ / Na/lIng Schedule: 8d@6"/12' and 3/12"

/ / Sheathing: OSB (/8')

/ / Shearwall HP: Durham + CASHEW/ 'NSF: None/ Damping: 1%

._-' _..._- - - - W = 24730 lb. total'-I

0 0.5 1 1.5 2 2.5 3

Sa(g)

Figure 5.46 Fragility curve for wall with garage door opening, ömax (relative toground) at second story

209

1

0.9

0.8

0.7

0.6

d 0.5

0.4

0.3

0.2

0.1

n

-10 -

2% ///

'.) /0

/ /

/ /

/ // / Structure Type: Two Story (20 x 16)

/ / ' Nailing Schedule: 8d@6'/12 and 3/12"

/ " Sheathing: OSB (/a)/ / Shearwall HP: Durham + CASHEW

/ / NSF: None

/ ' Damping: 1%

" _.- W= 24730 lb. total

0 0.5 1 1.5 2 2.5 3 3.5

Sa(g)

Figure 5.47 Fragility curve for wall with pedestrian door opening, max (relative toground) at first story

1

0.9

0.8

0.7

0.6

0.5

0.4

0.3

0.2

0.1

n

Structure Type: Two Story (20 x 16)Nailing Schedule: 8d@6"/12" and 3/12"Sheathing: OSB (I)Shearwall HP: Durham + CASHEWNSF: NoneDamping: 1%W = 24730 lb total

////

0 0.5 1 1.5 2 2.5 3

Sa(g)

Figure 5.48 Fragility curve for wall with pedestrian door opening, 6max (relative toground) at second story

210

0.9

0.8

0.7

0.6

0.4

0.3

0.2

0.1

(1

// I

/

/

/

/

-" " Interstory (GD)

2' Story (GD)

2 Story (PD)

1' Story (GD)

1S Story (PD)

Structure Type: Two Story (20 x 16')Nailing Schedule: 8d@6"/12" and 3/12"Sheathing: OSB (/8")

Shearwall HP: Durham + CASHEWNSF: NoneDamping: 1%Hazard Level: 10 (50%/50yrs)W = 24730 lb. total

0 0.5 1 1.5 2 2.5

Sa(g)

Figure 5.49 Comparison of fragility curves for shearwall in two-story structure (JO,50/50, 1% drift limit)

1

0.9

0.8

0.7

0.6

0 0.5

0.4

0.3

0.2

0.1

(1

--Interstory(GD)

nd

2nd Story (PD) '_,/"1st Story (GD)1st Story (PD)

///7

/ / Structure Type: Two Story (20 16'),i Nailing Schedule: 8d@6"/12" and 3/12"


/ / Shearwall HP: Durham + CASHEWNSF: None/ / Damping: 1%/ Hazard Level: LS (10%/5oyrs)

- W = 24730 lb. total

0 0.5 1 1.5 2 2.5 3

Sa(g)

Figure 5.50 Comparison of fragility curves for shearwall in two-story structure (LS,10/50, 2% drift limit)

211

Figures 5.51 and 5.52 present the fragility curves for the East wall (2EW) and

West wall (2WW), built with nonstructural finish materials (stucco and gypsum

wallboard), with partition walls and considering the JO (50/50) performance level (1%

drift limit). As noted earlier, NSF materials significantly improve the displacement

performance of wood shearwalls. For this wall, only the immediate occupancy (with

JO, 50/50) with the 1% drift limit case could be considered since very low failure

probabilities were obtained for the LS and CP performance levels when NSF materials

were included. The fragility curves in Figures 5.53 and 5.54 show that wall

constructed with NSF materials (with partition walls) can sustain higher seismic

demand than walls sheathed with OSB only (without partition walls). For example, if

the seismic demand variable (Sa) is 1.5g, the probability of failure of wall 1WW

considering NSF materials and partition walls is about 10%, versus about 80% for the

equivalent bare wall. Similarly, the probability of failure for wall 2WW is about 3%

considering NSF materials and partition walls, versus about 85% for the bare wall.

212

0.9

0.8

0.7

0.6

o.5

0.4

0.3

0.2

0.1

Structure Type: Two Story (20 x 16)Nailing Schedule: 8d@6"/12" and 3/12'Sheathing: OSB (/")Shearwall HP: Durham + CASHEWNSF: GWB + StuccoDamping: 1%Hazard Level: ID (50%/50yrs)W = 24730 lb. total

First story(with Partition Walls)

/:/ Second story

(with Partition Walls)

0 0.5 1 1.5 2 2.5 3

Sa(g)

Figure 5.51 Fragility curves for shearwall with NSF materials (2EW) in two-storystructure

0.9

0.8

0.7

0.6

o 0.5

0.4

0.3

0.2

0.1

A

Structure Type: Two Story (20 x 16')Nailing Schedule: 8d@6"/12" and 3/12"Sheathing: OSB (3/.)

Shearwall HP: Durham + CASHEWNSF: GWB + StuccoDamping: 1%Hazard Level: 10 (50%/5oyrs)W = 24730 lb. total

First story(with Partition Walls)

/// /

/ /

/ Second story

/ (with Partition Walls)

/

0 0.5 1 1.5 2 2.5 3

Sa(g)

Figure 5.52 Fragility curves for shearwall with NSF materials (2WW) in two-storystructure

213

1

0.9

0.8

0.7

0.6

o 0.5

0.4

0.3

0.2

0.1

Structure Type: Two Story (20 x 16)Nailing Schedule: 8d@6'112" 3/12"andSheathing: OSB (/)

// /

Shearwall HP: Durham + CASHEWNSF: GWB + StuccoDamping: 1%

,

Hazard Level: 10 (50%I5Oyrs)W = 24730 lb. total ,' Bare wall (GD)

(without Partition Walls

Bare wall (PD)(without Partition Walls)

/ NSF wall (GD)/,,(with Partition Walls')

NSF wall (PD)

/(with Partition Walls')

/

0 0.5 1 1.5 2 2.5 3

Sa(g)

Figure 5.53 Comparison of fragility curves showing contribution of NSF materials,max (relative to ground) at first story

0.9

0.8

0.7

0.6

ci 0.5

0.4

0.3

0.2

0.1

n

Structure Type: Two Story (20' x 16') .'Nailing Schedule: 8d@6'/12" and 3/12" /Sheathing: 058 (/') /Shearwall HP: Durham + CASHEW / /NSF: GWB + Stucco //Damping: 1% /Hazard Level: 10 (50%/5Oyrs) // IW = 24730 lb. total // /

I // NSF Wall (GD)

// / NSF Wall (PD)

/ (with Partition Walls

/1 Bare wall (GD) /(without Parti7n Walls)

Bare wall (PD)1'(without Partjyon Walls)

0 0.5 1 1.5 2 2.5 3

Sa(g)

Figure 5.54 Comparison of fragility curves showing contribution of NSF materials,ömax (relative to ground) at second story

214

6. CONCLUSIONS AND RECOMMENDATIONS

Wood is the most common material used in low-rise construction in the United

States. Light-frame wood structures have a number of advantages including aesthetics,

beauty, construction cost and time, versatility, flexibility in floor plans, and so forth.

Most woodframe structures consist of floors, walls, and roof systems tied together by

fasteners. Shearwalls provide the primary resistance to lateral forces (along with

diaphragms), such as these arising from earthquake loading, in most woodframe

structures.

The objective of this research was to explore the potential for the application of

performance-based engineering concepts to the design and assessment of woodframe

structures subject to earthquakes. To accomplish this, shearwalls either were treated as

isolated subassemblies or were assumed to act as part of complete structure. Nonlinear

dynamic time-history analysis was used to predict the performance of shearwalls

considering a suite of suitably scaled characteristic ordinary ground motions to

represent the seismic hazard.

Sensitivity studies were performed to investigate the relative effects of

damping, sheathing properties, fastener type and spacing, panel layout. and other

properties on the performance of wood shearwalls. In addition, the effects of

uncertainty in ground motions and variability in sheathing-to-framing connection

hysteretic parameters were investigated. Issues such as the contribution of

nonstructural finish materials, different seismic hazard regions, and construction

quality also were investigated and modification factors to adjust peak displacement

215

distributions were developed. The peak displacement distributions were then used to

construct performance curves and design charts as a function of seismic weights for

two baseline walls (BW1 and Owl), considering different levels of construction

quality, and different seismic hazard regions. In the consideration of shearwalls acting

as part of a complete structure, interstory drift and the effects of considering partition

walls also were investigated.

Incremental dynamic analysis (IDA) using baseline isolated shearwall and the

worst performance wall in two-story structure was performed in efforts to quantify an

approach drift limit for collapse prevention. In most cases considered, this value was

close to the drift limit specified by FEMA 356. Examples of fragility curves

(considering both peak displacement and ultimate uplift force) were developed. The

shearwalls to construct fragility curves were designed by considering different nailing

schedules (2"/12", 3"/12", 4"/12", and 6"/12"), corresponding allowable seismic

weights (back-calculated from the UBC '97 allowable unit shear values), and various

overstrength (R) factors.

6.1 Conclusions

The following conclusions were drawn from the results of this research:

1. Performance-based design concepts can be applied to the design and

assessment of woodframe structure and can provide valuable information for

understanding the performance of shearwalls subject to earthquake loading.

216

2. The greatest contributors to variability in predicted shearwall response are the

ground motions. Therefore, caution must be used when specifying the seismic

hazard used to develop performance-based design requirements. Other

uncertainties contributing variability in shearwall response are model

uncertainty, sheathing-to-framing connection hysteretic parameters, and

construction quality.

3. The contributions of nonstructural finish materials to the performance of

woodframe shearwalls may be significant, especially at large demands, and

therefore should be considered when developing performance-based design

guidelines. In particular, the application of stucco serves to greatly reduce peak

shearwall displacements.

4. Construction quality issues such as missing or misplaced fasteners, overall

levels of construction quality, and quality of NSF material application can

significantly influence shearwall performance under earthquake loading.

5. Different earthquake scaling methods will result in different predicted

shearwall performance. However, the median displacement values are similar.

6. Simple deterministic modification factors can be developed to account for

variation in sheathing to-framing connection hysteretic parameters, effects of

different levels of construction quality, and contributions of nonstructural

finish materials. These factors can be used to adjust peak displacement

distributions obtained by nonlinear dynamic time-history analysis.

217

7. Performance curves and design charts can be developed using seismic weight

as the design variable. These permit selection of a particular sheathing type and

fastener spacing for a given seismic weight to meet specific performance

objectives at different hazard levels. The procedure to develop performance

curves and design charts is sufficiently modular to allow different information

on shearwall properties, seismic hazard, and so forth to be included.

8. Incremental dynamic analysis (IDA) can be used to quantify appropriate drift

limits for collapse prevention. In most cases considered in this study, this value

was close to the drift limit specified by FEMA 356.

9. The shearwalls designed using UBC '97 allowable unit shear values provide

relatively consistent levels of safety, as evidenced by the fact that the resulting

fragility curves were quite close for all nailing schedules. Thus, a single

fragility curve can be constructed for a given R factor.

10. Interstory drift may be used the most appropriate (conservative) displacement

criteria to consider in displacement-based design of woodframe structures.

11. The performance of a shearwall acting as part of a one- or two-story structure

is quite different from a shearwall acting as an isolated assembly. The

difference in peak displacement distributions is relatively small at the 10

hazard level, however increases as the hazard level increases. It therefore may

be appropriate to modify isolated shearwall performance to take into account

overall system performance of a woodframe structure, particularly for high

hazard levels. This modification, however, may be structure-dependent. In both

218

cases (isolated shearwall and wall acting as part of a complete structure),

however, peak displacements are significantly reduced with the addition of

NSF materials.

12. Partition walls have a significant effect on the performance of shearwalls at all

hazard levels. Also, the variability in peak wall displacement is reduced when

the effects of partition walls are considered.

6.2 Recommendations

The following might be suggested as topics for future study:

1. While construction quality issues (i.e., missing fasteners and level of

construction quality) were investigated in this study, there are a number of

other construction quality issues which could significantly influence overall

shearwall behavior. Among these are misplaced fasteners and anchors,

deterioration of structural and nonstructural finish materials, improper

selection of fasteners, under-driven or over-driven fasteners, missing blocking,

the use of smaller panel segments, cutouts in framing members (e.g., for

installation of conduit), and so forth.

2. The numerical model (CASHEW) used to develop global hysteretic parameters

for shearwall given nail hysteretic parameters assumes rigid hold-downs and

assumes that all fasteners have the same hysteretic parameters. Both of these

restrictions should be removed to permit more comprehensive investigations of

219

walls having different (and more realistic) anchorage and connection

properties.

3. Durability is an important consideration in woodframe structures. The primary

durability issues for wood structures are decay, corrosion of connectors, and

insect attack. Recently completed tests of connections with various levels of

decay are available. These results can be used, for example, to develop new

fastener hysteretic parameters for use with the CASHEW model, and thereby

investigate the effects of decay on shearwall performance.

4. Unidirectional earthquake records were used in this study. A suite of bi-

directional earthquake records, which can be obtained from the SAC project

(e.g.,), can be used to more accurately reflect ground motion characteristics

and their effects on the three-dimensional structure. The SAWS program can

be used with bi-directional earthquake records. Also, near-fault ground motion

records were not considered in most of this study. Therefore, a more

comprehensive study considering bi-directional and near-fault records might

be conducted.

5. Performance curves and design charts were developed considering only limited

sheathing types/thickness and nail types (sizes). In order to extend these design

tools to include a wider range of products, additional sheathing and fastener

types will need to be considered.

6. While it is recognized that the presence of NSF materials can significantly

improve shearwall performance, and evidence of this has been documented in

220

a number of recent studies, the degree to which this benefit (1) can be counted

upon for the design life of the structure, and (2) can be quantified for design

purposes, remains to be studied.

7. While the drift limits used in this study were adopted directly from FEMA 356,

their definition is qualitative. More quantitative definitions may be needed

which correlate more realistically with observed damage following actual

earthquake. Suitable drift limits and other performance measure are being

investigated by other researchers.

8. Woodframe construction elsewhere in the U.S. is conventional rather than

engineered and does not rely on large shearwalls, seismic hold-downs, or dense

nailing schedules. It may be useful, particularly for assessment and evaluation

purposes, to develop fragility curves using the approach developed in this

study for these types of structures.

221

REFERENCES

AF&PA (2001), National Design SpecfIcation for Wood Construction, AmericanForest & Paper Association, Washington, DC.

AISC (2001), "Description of Performance-Based Design Procedures and RelatedIssues," White Paper Distributed at AISC Workshop on Performance-BasedDesign of Steel Buildings for Seismic Loads, June 2001.

ASCE (2000), Minimum Design Loads for Building and Other Structures, ASCE 7-98,American Society of Civil Engineers, Reston, VA.

CALTRANS (1999), Seismic Design Criteria Version 1.1, California Department ofTransportation, CA.

Camelo, V.S., Beck, J.L. and Hall, J.F. (2002), Dynamic Characteristics ofWoodframe Structures, CUREE Publication No. W- 11, Consortium of Universities forResearch in Earthquake Engineering, Richmond, CA.

Ceccotti, A. and Foschi, R.O. (1998), "Reliability Assessment of Wood Shear Wallsunder Earthquake Excitation," Proceedings of International Conference onComputational Stochastic Mechanics, Santorini, Greece, June 1998.

Cheung, C,K., Itani, R.Y. and Polensek, A. (1988), "Characteristics of WoodDiaphragms: Experimental and Parametric Studies," Wood and Fiber Science,20(4):438-456.

Cobeen, K.E. (2001), Recommendations for Earthquake Resistance in the Design andConstruction of Woodframe Buildings, Part I and II- Code, Standards and GuidelinesDesign and Analysis Procedures and Construction Practice, Element 3 Draft Reportof CUREE Woodframe Project, San Diego, CA.

Cobeen, K.E. (2003), Personal Communication.

Deam, B.L. (1997), Seismic Ratings for Residential Timber Buildings, BRANZ StudyReport 73, Judgeford, New Zealand.

Deam, B.L. (2000), "Rating Seismic Bracing Elements for Timber Buildings," 12th

World Conference on Earthquake Engineering, Auckland, New Zealand.

Deierlein, G.G. and Kanvinde, A.M. (2003), Seismic Performance of Gypsum WallsAnalytical Investigation, CUREE Publication No. W-23, Consortium of Universitiesfor Research in Earthquake Engineering, Richmond, CA.

222

DerKiureghian, A. (1996), "Structural Reliability Methods for Seismic SafetyAssessment: a Review," Engineering Structures, 18(6) :412-424.

Dinehart, D.W. and Shenton, H.W. (1998), "Comparison of Static and DynamicResponse of Timber Shear Walls," Journal of Structural Engineering, ASCE,124(6):686-695.

Dinehart, D.W. and Shenton, H.W. (2000), "Model for Dynamic Analysis of WoodFrame Shear Walls," Journal of Engineering Mechanics, ASCE, 126(9): 899-908.

Dorey, D.B. and Schriever, W.R. (1957), Structural Test of a House under SimulatedWind and Snow Loads, ASTM Special Technical Report No. 210, Philadelphia.

Dolan, J.D. (1989), "The Dynamic Response of Timber Shear Walls," Ph.D.Dissertation, Department of Civil Engineering, University of British Columbia,Vancouver, Canada.

Dolan, J.D. and Madsen, B. (1992), "Monotonic and Cyclic Tests of Timber ShearWalls," Canadian Journal of Civil Engineering, 19(3): 115-122.

Durham, J.P. (1998), "Seismic Response of Wood Shearwalls with OversizedOriented Strand Board Panels," MASc Thesis, University of British Columbia,Vancouver, Canada.

Ellingwood, B., Galambos, T.V., MacGregor, T.G. and Cornell, C.A. (1980),Development of a Probability Based Load Criterion for American National StandardA58: Building Code Requirements for Minimum Design Loads in Buildings and OtherStructures, U.S. Department of Commerce, National Bureau of Standards, NBSSpecial Publication 577, Washington, D.C.

Ellingwood, B.R. and Rosowsky, D.V. (1991), "Duration of Load Effects in LRFD forWood Construction," Journal of Structural Engineering, ASCE, 117(2):584-599.

Ellingwood, B.R. (1994), "Probability-Based Codified Design for Earthquakes,"Engineering Structures, 1 6(7):498-5 06.

Ellingwood, B.R. (2000), "Performance-Based Design: Structural ReliabilityConsiderations," Advanced Technology in Structural Engineering, Proceedings ofStructures Congress 2000, Philadelphia, PA, ASCE, Reston, VA.

Ellingwood, B.R., Rosowsky, D.V., Li, Y. and Kim, J.H. (2003), "FragilityAssessment of Light-Frame Wood Construction Subjected to Wind and EarthquakeHazards," Paper submitted to ASCE Journal of Structural Engineering, in Review.

223

FEMA (2000a), Prestandard amd Commentary for the Seismic Rehabilitation ofBuildings, Federal Emergency Management Agency, Washington, DC.

FEMA (2000b), Global Topics Report on the Prestandard and Commentary for theSeismic Rehabilitation of Buildings, Federal Emergency Management Agency,Washington, DC.

Filiatrault, A. (1990), "Static and Dynamic Analysis of Timber Shear Walls,"Canadian Journal of Civil Engineering, 17(4): 643-651.

Filiatrault, A., Foschi, R.O. and Folz, B. (1990), "Reliability of Timber Shear Wallsunder Wind and Seismic Loads," Proceedings of International Timber EngineeringConference, Tokyo, Japan, October 1990.

Filiatrault, A. and Folz, B. (2000), "Performance-Based Seismic Design of WoodFramed Buildings," Journal of Structural Engineering, ASCE 128(1):39-47.

Filiatrault, A., Uang, C., Folz, B., Chrstopoulos, C. and Gatto, K. (2001),Reconnaissance Report of the February 28, 2001 Nisqually (Seattle-Olympia)Earthquake, Report No. SSRP-2001/02, Dept. of Structural Engineering, University ofCalifornia San Diego, La Jolla, CA

Fischer, D., Filiatrault, A., Folz, B., Uang, C-M. and Seible, F. (2001), Shake TableTests of a Two-Story Woodframe House, CUREE Publication No. W-06, Consortiumof Universities for Research in Earthquake Engineering, Richmond, CA.

Foliente, G.C. (1995), "Hysteresis Modeling of Wood Joints and Structural Systems,"Journal of Structural Engineering, ASCE, 121(6): 1013-1022.

Foliente, G.C., Paevere, P., Saito, T. and Kawai, N. (2000), "Reliability Assessment ofTimber Shear Walls under Earthquake Loads," Proceedings of 12th World Conferenceon Earthquake Engineering (l2th WCEE), Auckland, New Zealand.

Folz, B. and Filiatrault, A. (2000), CASHEW- Version 1.0: A Computer Program forCyclic Analysis of Wood Shear Walls, CUREE Publication No. W-08, Consortium ofUniversities for Research in Earthquake Engineering, Richmond, CA.

Folz, B. and Filiatrault, A. (2001), "Cyclic Analysis of Wood Shear Walls," Journal ofStructural Engineering, ASCE, 127(4) :433-441.

Folz, B. and Filiatrault, A. (2002), SAWS Seismic Analysis of Woodframe StructuresVersion 1.0, CUREE Publication No. W-21, Consortium of Universities for

Research in Earthquake Engineering, Richmond, CA.

224

Fonseca, F.S., Rose, S.K. and Campbell, S.H. (2002), Nail, Screw and Staple FastenerConnections, CUREE Publication No. W- 16, Consortium of Universities for Researchin Earthquake Engineering, Richmond, CA.

Foschi, R.O. (1977), "Analysis of Wood Diaphragms and Trusses, Part I:

Diaphragms," Canadian Journal of Civil Engineering, 4(3): 345-352.

Gatto, K. and Uang, C.M. (2002), Cyclic Response of woodframe Shearwalls: LoadingProtocol and Rate of Loading Effects, CUREE Publication No. W-13, Consortium ofUniversities for Research in Earthquake Engineering, Richmond, CA.

Haldar, A. and Mahadevan, S. (2000), Probability, Reliability and Statistical Methodsin Enginerring Design, John Wiley & Sons, Inc., New York.

Home, J.C. (2002), Personal Communication.

Hurst, H.T. (1965), The Wood Frame House as a Structural Unit, Technical ReportNo. 5, National Forest Products Association, Washington D.C.

ICBO (1997), Uniform Building Code, International Conference of Building Officials,Whittier, CA.

Isoda, H., Folz, B. and Filiatrault, A. (2002), Seismic Modeling of Index WoodframeBuildings, CUREE Publication No. W-12, Consortium of Universities for Research inEarthquake Engineering, Richmond, CA.

Jones, S.N. and Fonseca, F.S. (2002), "Capacity of OSB Shear Walls with OverdrivenSheathing Nails," Journal of Structural Engineering, ASCE, 128(7): 898-907.

Krawinkler, H. and Seveviratna, G. T. (1998), "Pros and Cons of a Pushover Analysisin Seismic Performance Analysis," Engineering Structures, 20(4):452-464.

Krawinkler, H., Parisi, F., Ibarra, L., Ayoub, A. and Medina, R. (2000), Developmentof a Testing Protocol for Wood Frame Structures, CUREE Publication No. W-02,Consortium of Universities for Research in Earthquake Engineering, Richmond, CA.

Mahaney, J.A. and Kehoe, B.E. (2002), Anchorage of Woodframe BuildingsLaboratory Testing Report, CUREE Publication No. W-14, Consortium ofUniversities for Research in Earthquake Engineering, Richmond, CA.

McCutcheon, W.J. (1985), "Racking Deformations in Wood Shear Walls," Journal ofStructural Engineering, 111(2):257-269.

225

McMullin, K.M. and Merrick, D. (2002), Seismic Performance of Gypsum WallsExperimental Test Program, CUREE Publication No. W-15, Consortium ofUniversities for Research in Earthquake Engineering, Richmond, CA.

Moller, 0., Foschi, R.O. and Rubinstein, M. (2001), "Reinforced Concrete Structuresunder Seismic Excitation: Reliability and Performance Based Design," Proceedings ofthe 8th International Conference on Structural Safety and Reliability (ICOSSAR '01),Newport Beach, CA

Mosalam, K., Machado, C., Gliniorz, K., Naito, C., Kunkel, E. and Mahin, 5. (2003),Seismic Evaluation of Asymmetric Three-Story Woodframe Building, CUREEPublication No. W-19, Consortium of Universities for Research in EarthquakeEngineering, Richmond, CA.

NAHB (1994), Assessment of Damage to Residential Buildings Caused by theNorthridge Earthquake, NAHB Research Center, Upper Marlboro, MD.

NAHB (1997), The State-of-the-Art of Building Codes and Engineering Methods forSingle-Family Detached Homes: an Evaluation of Design Issues and ConstructionCosts, HADTE Final Report Prepared for NAHB and HUD, NAHB Research Center,Upper Marlboro, MD.

Ohashi, Y., Sakamoto, I. and Kimura, M. (1998), "Shaking Table Tests of a RealScale Wooden House Subjected to Kobe Earthquake," Proceedings of the InternationalConference on Timber Engineering, Montreux, Switzerland.

Paevere, P. and Foliente, G.C. (2000), "Hysteretic Pinching and Degradation Effectson Dynamic Response and Reliability," International Conference on Applications ofStatistics and Probability.

Pardoen, G., Waltman, A., Kazanjv, R., Freund, E. and Hamilton, C. (2003), Testingand Analysis of One-Story and Two-Story Shearwalls under Cyclic Loading, CUREEPublication No. W-25, Consortium of Universities for Research in EarthquakeEngineering, Richmond, CA.

PATH (2000), Residential Structural Design Guide: 2000 Edition, Partnership forAdvancing Technology in Housing, Washington, DC.

Patton-Mallory, M., Wolfe, R.W., Soltis, L.A. and Gutkowski, R.M. (1985), "Light-Frame Shear Wall Length and Opening Effects," Journal of Structural Engineering,ASCE, 11 1(10):2227-2239.

226

Paevere, P.J., Foliente, G.c. and Kasal, B. (2003), "Load-Sharing and Redistribution ina One-Story Woodframe Building," Journal of Structural Engineering, ASCE,129(9): 1275-1284.

Performance Criteria Resource Document for Innovative Construction (1977), ReportNBSIR 77-13 16, National Bureau of Standards, Washington, DC (available fromNTIS).

Plywood Design Specification (1998), APA The Engineered Wood Association,Tacoma, WA.

Phillips, T.L. (1990). "Load sharing of three-dimensional wood diaphragms." M.S.Thesis, Department of Civil and Environmental Engineering, Washington StateUniversity, Pullman, WA.

Philips, Y.L., Itani, R.Y. and McLean, D.L. (1993) "Lateral Load Sharing byDiaphragms in Wood-Frame Buildings," Journal of Structural Engineering, ASCE,1 19(5):1556-1571.

Rosowsky, D. and Ellingwood, B. (2002), "Performance-Based Engineering of WoodFrame Housing: A Fragility Analysis Methodology," Journal of StructuralEngineering, ASCE, 128(1): 32-38.

Rosowsky, D. (2002), "Reliability-Based Seismic Design of Wood Shear Walls,"Journal of Structural Engineering, ASCE, 128(11): 1439-1453.

Rosowsky, D.V. and Kim, J.H. (2002a), Reliability Studies, CUREE Publication No.W-10, Consortium of Universities for Research in Earthquake Engineering,Richmond, CA.

Rosowsky, D.V. and Kim, J.H. (2002b), "Performance-Based Seismic Design ofWood Shearwalls," Proceedings of World Conference on Timber Engineering (WCTE2002), Selangor, Malaysia.

Rosowsky, D.V. and Kim, J.H. (2003), "A Probabilistic Framework for Performance-Based Design of Wood Shearwalls", Proceedings: 9th International Conference onApplications of Statistics and Probability in Civil Engineering (ICASP9), SanFrancisco, CA.

SAC Joint Venture (1995), Interim Guidelines: Evaluation, Repair, Modfication, andDesign of Welded Steel Moment Frames, Report FEMA 267, Federal EmergencyManagement Agency, Washington, D.C.

227

Salenikovich, A.J. (2000), "The Racking Performance of Light-Frame Shear Walls,"Ph.D. Dissertation, Department of Wood Science and Forest Products, VirginiaPolytechnic Institute and State University, Blacksburg, VA.

Sasani, M. and Der Kiureghian, A. (2001), "Seismic Fragility of RC Structural Walls:Displacement Approach," Journal of Structural Engineering, ASCE, 127(2) :219-228.

SEAOC (1999), Recommended Lateral Force Requirements and Commentaiy, BlueBook, Structural Engineers Association of California, Sacramento, CA.

SEAOSC (2001), Report of a Testing Program of Light-Framed Walls with Wood-Sheathed Shear Panels, Final Report to the City of Los Angeles Department ofBuilding and Safety, Structural Engineers Association of Southern California andDepartment of Civil and Environmental Engineering, UC-Irvine, December.

Seible, F., Filiatrault, A. and Uang, C.M. (1999), Proceedings of the InvitationalWorkshop on Seismic Testing, Analysis and Design of Wooframe Construction,CUREE Publication No. W-01, Consortium of Universities for Research inEarthquake Engineering, Richmond, CA.

Seo, J., Choi, I. and Lee, J. (1999), "Experimental Study on the Aseismic Capacity ofa Wooden House using Shaking Table," Earthquake Engineering and StructuralDynamics, 28(10): 1143-1162.

Serrette, R.L., Encalada, J., Juadines, M. and Nguyen, H. (1997), "Static RackingBehavior of Plywood, OSB, Gypsum, and Fiberbond Walls with Metal Framing,"Journal of Structural Engineering, ASCE, 123(8): 1079-1086.

Shinozuka, M., Feng, M.Q., Kim, H.K. and Kim, S.H. (2000), "Nonlinear StaticProcedure for Fragility Curve Development," Journal of Engineering Mechanics,ASCE, 126(12): 1287-1295.

Shinozuka, M., Feng, M.Q., Lee, J. and Naganuma, T. (2000), "Statistical Analysis ofFragility Curves," Journal of Engineering Mechanics, ASCE, 126(12): 1224-123 1.

Shome, N. (1999), "Probabilistic Seismic Demand Analysis of Nonlinear Structures,"Ph.D. Dissertation, Department of Civil and Environmental Engineering, StanfordUniversity, Stanford, CA.

Simpson Strong-Tie (2002), "Catalogue of Wood Construction Connectors,"Catalogue C-2003, Dublin, CA.

228

Somerville, P., Smith., Punyamurthula, S. and Sun, J. (1997), Development of GroundMotion Time Histories for Phase 2 of the FEMA/SAC Steel Project, Report No.SAC/BD-97/04, SAC Joint Venture, Sacramento, CA.

Stewart, W.G. (1987), "The Seismic Design of Plywood Sheathed Shear Walls," Ph.D.Dissertation, Department of Civil Engineering, University of Canterbury,Christchurch, New Zealand.

Stewart, A., Goodman, J., Kliewer, A. and Salsbury, E. (1988); "Full-Scale Tests ofManufactured Houses under Simulated Wind Loads," Proceedings of the InternationalConference on Timber Engineering, Forest Products Research Society, Madison, WI.

Sugiyama, H., Naoto, A., Takayuki, U., Hirano, S. and Nakamura, N. (1988), "FullScale Test on a Japanese Type of Weo-Story Wooden Frame House Subjected toLateral Load," Proceedings of the International Conference on Timber Engineering,Forest Products Research Society, Madison, WI.

Tuomi, R.L. and McCutcheon, W.J. (1974), Testing of a Full-Scale House underSimulated Snow Loads and Wind Loads, Research Paper FPL 234, Forest ProductsLaboratory, Madison, WI.

Tuomi, R.L. and McCutcheon, W.J. (1978), "Racking Strength on Light-Frame NailedWalls," Journal of the Structural Division, ASCE, 104(11): 1131-1140.

Vamvatsikos, D. (2002), "Seismic Performance, Capacity and Reliability of Structuresas Seen through Incremental Dynamic Analysis," Ph.D. Dissertation, Department ofCivil and Environmental Engineering, Stanford University, Stanford, CA.

Van de Lindt, J.W. and Walz, M.A. (2003), "Development and Application of WoodShear Wall Reliability Model," Journal of Structural Engineering, ASCE, 129(3):405-413.

Wen, Y .K. and Foutch, D.A. (1997), Proposed Statistical and Reliability Frameworkfor Comparing and Evaluating Predictive Models for Evaluation and Design, andCritical Issues in Developing Such Framework, Report No. SAC/BD-97/03, SACJoint Venture, Sacramento, CA.

Wen, Y.K., Foutch, D.A., Eliopoulis, D. and Yu, C.Y. (1994), "Seismic Reliability ofCurrent Code Procedures for Steel Buildings," Proceedings of 5th National Conferenceon Earthquake Engineering, Chicago, IL, pp. 4 17-426.

Wen, Y.K., Hwang, H. and Shinozuka, M. (1994), Development of Reliability-BasedDesign Criteria for Buildings Under Seismic Load, Technical Report 94-0023,National Center for Earthquake Engineering, SIJNY Buffalo, NY.

229

Wood Handbook Wood as an Engineered Material (1999), Forest Products Society,Madison, WI.

Yancy, C.W., and Somes, N.F. (1973), Structural Test of a Wood Framed HousingModule, NB SIR Report 73-121, National Bureau of Standards, Washington DC.

Yokel, F.Y., Hsi, G, and Somes, N.F. (1973), Full Scale Test on a Two-Story HouseSubjected to Lateral Load, Building Science Series 44, National Bureau of Standards,Washington, DC.

Yun, S. Y., Foutch, D.A. and Lee, K. (2000), "Reliability and Performance BasedDesign for Seismic Loads," Proceedings of 8th ASCE Specialty Conference onProbabilistic Mechanics and Structural Reliability (PMC2000), Notre Dame, IN.

230

APPENDICES

231

APPENDIX A: Example showing convolution of hazard curve and fragility curve

232

Fragility curves such as those developed in Section 4.4.2.1 can be convolved

with a hazard curve to evaluate failure probability. The probability of failure can be

obtained using the following equation;

Pf =fPx(x)FR(x)dx (A-i)

where, Pf = failure probability, Px(x) = probability density function of hazard (in this

case 50-year seismic hazard), and FR(x) = fragility. One example using a hazard curve

for southern California (specifically, Landers region) and a fragility curve for an

isolated shearwall considering a 3"/12" nailing schedule, various R factors, and the

LS (10/50) hazard level is shown in Figure A.1.

0.9

0.8

- 0.7

80.a)L.

QU) 0.2

0.1

Hazarçi Curve (Landers)

/,/ /

I\ / Fragility Curve1 \ I R = 5.5 (3/12") -3 Pf= 0.16880

/ / ,'\ / R = 4.5 (3/12") P = 0.12600,' \/ R = 3.5 (3/12") Pf =0.08417

// R = 2.5 (3/12") -* Pf = 0.03697

-I-,

0 0.5 1 1.5 2 2.5Sa(g)

Figure A.! Convolution of hazard curve and fragility curve

3

233

APPENDIX B: Deterministic modification factors for construction quality

234

Index F0 K0 F0 F1

BuildingType

Quality Story Sheathing.

kipkips r1

kip kipr2 r3 a 3

OSB only 1.00 1.00 1.00 1.00 1.00

SmallTypical OSB +

GWB1.00 1.00 1.00 1.00 1.00

House OSB only 1.00 1.00 1.00 1.00 1.00Poor

1.00 1.00 1.00 1.00 1.00

OSB only 0.87 0.87 0.97 0.87 0.86 1.02 1.01 1.01 1.01

OSB+

Stucco0.89 0.88 0.98 0.89 0.88 1.01 1.00 1.00 1.00

I 0.86 0.86 0.98 0.86 0.85 1.01 1.00 1.00 1.00

OSB +Stucco + 0.87 0.87 0.99 0.87 0.87 1.01 1.00 1.00 1.00

TypicalGWB

OSB only tL84 0.85 1197 0.84 0.84 1.04 1.00 1.01 1.00

OSB+

Stucco0.87 0.88 0.99 0.87 0.87 1.02 1.00 1.00 1.00

2 0.84 0.85 0.99 0.85 0.85 1.02 1.00 1.00 1.00

OSB +

Stucco + 0.86 0.87 0.99 0.86 0.87 1.01 1.00 1.00 1.00

Large GWBHouse OSB only 0.62 0.60 1.00 0.61 0.61 0.96 1.02 1.01 1.01

OSB+

Stucco0.66 0.65 1.00 0.65 0.66 0.98 1.01 1.00 1.00

0.68 0.67 1.00 0.68 0.68 0.98 1.01 1.00 1.00

OSB +

Stucco + 0.69 0.68 1.00 0.69 0.69 0.99 1.01 1.00 1.00

oorGWB

OSB only 0.66 0.59 1.01 0.65 0.58 0.97 1.01 1.00 1.00

OSB+

Stucco0.68 0.64 1.00 0.68 0.64 0.99 1.00 1.00 1.00

2 0.71 0.67 1.00 0.70 0.67 0.99 1.00 1.00 1.00

OSB+Stucco+ 0.70 0.68 1.00 0.70 0.68 0.99 1.00 1.00 1.00

GWBTown OSBonly 0.85 0.87 0.99 0.84 0.85 0.93 1.00 1.00 1.00

House 05B+Stucco

0.87 0.88 1.00 0.87 0.87 0.96 1.00 1.00 1.00

1 0.85 0.86 1.00 0.85 0.85 0.96 1.00 1.00 1.00

OSB +Stucco + 0.87 0.87 1.00 0.86 0.86 0.98 1.00 1.00 1.00

TypicalGWB

OSB only 0.85 0.87 0.99 0.85 0.85 0.97 1.00 1.00 1.00

OSB+

Stucco0.88 0.88 0.99 0.88 0.87 0.98 1.00 1.00 1.00

2 0.85 0.86 0.99 0.85 0.85 0.98 1.00 1.00 1.00

OSB +

Stucco + 0.87 0.87 1.00 0.87 0.87 0.99 1.00 1.00 1.00

GWBPoor I OSB only 0.63 0.63 0.99 0.62 0.62 0.93 1.00 1.00 1.00

OSB+Stucco 0.66 0.67 0.99 0.66 0.66 0.97 1.00 1.00 1.00

0.69 0.69 0.99 0.69 0.68 0.97 1.00 1.00 1.00

235

OSB +

Stucco+ 0.69 0.69 1.00 0.69 0.69 0.98 1.00 1.00 1.00

GWBOSB onjy 1162 1161 1198 0.62 (161 1197 1.00 L00 1.00

OSB+ 0.66 0.65 0.99 0.66 0.66 0.98 1.00 1.00 1.00

Stucco

2 0.69 0.68 0.99 0.68 0.68 0.98 1.00 1.00 1.00

OSB +

Stucco+ 0.69 0.69 0.99 0.69 0.69 0.99 1.00 1.00 1.00

GWBOSB only 0.86 0.82 1.00 0.86 0.86 0.97 1.00 1.00 1.00

OSB+Stucco

0.88 0.86 1.00 0.88 0.88 0.99 1.00 1.00 1.00

1 0.85 0.83 1.00 0.85 0.88 0.99 1.00 1.00 1.00

OSB +

Stucco + 0.87 0.86 1.00 0.87 0.88 0.99 1.00 1.00 1.00

TypicalGWB

OSB only (184 (186 (199 (184 (184 (197 L00 1.00 L00OSB

+ 0.87 0.88 0.99 0.87 0.87 0.99 1.00 1.00 1.00Stucco

2,3 0.85 0.86 0.99 0.85 0.85 0.99 1.00 1.00 1.00

OSB +

Stucco + 0.86 0.87 1.00 0.86 0.86 0.99 1.00 1.00 1.00

Apartment GWBBuilding OSB only 0.65 0.68 0.97 0.65 0.65

OSB+

Stucco0.67 0.69 0.99 0.67 0.68

1 0.70 0.72 0.99 0.70 0.72

OSB +

Stucco + 0.70 0.71 0.99 0.70 0.71

oorGWB

OSB only 0.61 0.58 1.00 0.61 0.60OSB

+Stucco

0.65 0.64 1.00 0.65 0.65

2,3 0.68 0.67 1.00 0.68 0.68

OSB +Stucco + 0.69 0.68 1.00 0.69 0.69

GWB

OSB only 0.85 0.86 0.99 0.85 0.85 0.99 1.00 1.00 1.00

TypicalNSF 0.86 0.87 0.99 0.86 0.87 0.99 1.00 1.00 1.00

Modification OSB onjy 1163 1161 (199 1163 (161 (197 LOl 11)0 L00OSB

+ 0.67 0.66 1.00 0.66 0.66 0.98 1.00 1.00 1.00FactorPoor Stucco

OSB +

NSF 0.69 0.69 1.00 0.69 0.69 0.98 1.00 1.00 1,00

(GWB)

Table B. 1 Deterministic modification factors for construction quality

236

APPENDIX C: Scaling earthquake records to response spectra considering differentscaling methods

2.5

2.0

1.5

(I)

1.0

0.5

237

0.0

0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0

Peirod (sec)

Figure C. 1 20 0GM records (CUREE) scaled over the plateau re ion of the response

spectrum (LS, 10/50)

3.0

2.5

2.0

0)1.5

U)

1.0

0.5

0.0

0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0

Period (sec)

Figure C.2 20 0GM records (CUREE) scaled at a period of 0.2 sec to the response

spectrum (LS, 10/50)

238

3.0

2.5

2.0

C)1.5

C,)

1.0

0.5

0.0

0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0

Period (sec)

Figure C.3 20 0GM records (CUREE) scaled at a period of 0.5 sec to the responsespectrum (LS, 10/50)

239

APPENDIX D: Earthquake records used in this study

240

EQ EventFile Station MW1

D (2)

FM (3) fDT (4) D5 ST

I PGA 1 PGV r PGD

km sec sec1

j[

cm& Year j

SUPI Brawley 6.7 18.2strike- 0.010

21.96D 0.116 17.2 8.6

slip 0Superstition

Hills SUP2El Centro

6.7 13.9strike- 0005 D 0.258 40.9 20.2

imperial slip 5(1987)SUP3

Plaster6.7 21.0

strike-0.010

22.22D 0.186 20.6 5.4

City slip 0

NOR2Beverly

6.7 19.6reverse- 0010 29.98

C 0.416 59.0 13.1Hills slip 0

NOR3Canoga

6.7 15.8reverse- 0.010

24.98D 0.356 32.1 9.1

Park slip 0

NOR4Glendale-

6.7 25.4reverse-

0.01029.98

D 0.357 12.3 1.9Las Palmas slip 0

NorthridgeNOR5

LA-6.7 25.5

reverse-0.020

39.98D 0.231 18.3 4.8

(1994) Hollywood slip 0

NOR6LA N.

6.7 23.9reverse-

0.01029.98

D 0.273 15.8 3.3Faring slip 0

NOR9North

6.7 14.6reverse-

0.01021.91

C 0.271 22.2 11.7Hollywood slip 0

NORIOSunland

6.7 17.7reverse- 0.010

29.98C 0.157 14.5 4.3

Mt slip 0

LP1 Capitola 6.9 14.5reverse-

0.00539.95

D 0.529 36.5 9.1oblique 0

LP2Gilroy

6.9 14.4reverse-

0.00539.94

D 0.555 35.7 8.2Array #3 oblique 0

Loma LP3Gilroy

6.9 16.1reverse-

0.00539.94

D 0.417 38.8 7.1Array #4 oblique 5

Prieta(1989) LP4

Gilroy6.9 24.2

reverse-0.005

39.94D 0.226 16.4 2.5

Array #7 oblique 5

LP5Hollister

6.9 25.8reverse-

0.00539.63

D 0.279 35.6 13.1Diff. Array oblique 5

LP6Saratoga-

6.9 13.7reverse-

0.00539.94

C 0.332 61.5 36.4West Val. oblique 5

Cape CM1Fortuna

7.1 23.6reverse- 0.020

43.98C 0.116 30.0 27.6

Boulevard slip 0Mendocino

(1992) CM2Rio Dell

7.1 18.5reverse- 0.020

35.98C 0.385 43.9 22.0

Overpass slip 0

LAN1Desert Hot

7.3 23.3strike-

0.02049.98

C 0.154 20.9 7.8Landers Springs slip 0(1992)

LAN2Yermo Fire

7.3 24.9strike- 0.020

43.98D 0.152 29.7 24,7

Station slip 0

Table D.1 Set of LA ordinary ground motion records (CUREE project)

(1) Moment magnitude(2) Closest source-to-site distance(3) Faulting mechanism(4) Recording time interval(5) Duration

241

EQ EventI

I D (2) fNumber-

FDT (4)

______

D5 J

PGAI lO%/ I I

File Station MW1 I

I

I FM (31 of I

I

I

I

IST

I

ISOyrs

0GM& Year

[

I

I

PointsI

I

i

SF

L Jso [

I sec I

.___________

Imperial LAO!

___________

El Centro

_______

6.9 10.0strike- 2674 0.020

,

53.460 D 452.03 2.01 0.229Valley slip

(1940) LAO2 El Centro 6.9 10.0strike 2674 0.020 53.460 D 662.88 2.01 0.336

slip

LAO3 Array#5 6.5 4.1strike-

3939 0.010 39.380 0 386.04 1.0! 0.390slip

Imperial LAO4 Array #5 6.5 4.1strike-

3939 0.010 39.380 D 478.65 1.01 0.483Valley(1979) LAOS Array #6 6.5 1.2

strike-3909 0.010 39.080 D 295.69 0.84 0.359

slip

LAO6 Array #6 6.5 1.2strike-

3909 0.010 39.080 D 230.08 0.84 0.279slip

LAO7Landers,

7.3 36.0strike- 4000 0.020 79.980 0 412.98 3.20 0.132

Barstow slip

LAO8Landers,

7.3 36.0strike- 4000 0.020 79.980 D 417.49 3.20 0.133

Landers Barstow slip(1992)

LAO9Landers,

7.3 25.0strike- 4000 0.020 79.980 0 509.70 2.17 0.240

Yermo slip

LAIOLanders,

7.3 25.0strike- 4000 0.020 79.980 D 353.35 2.17 0.166

Yermo slipLAII Gilroy 7.0 2.0 oblique 2000 0.020 39.980 D 652.49 1.79 0.372

Loma LAI2 Gilroy 7.0 12.0 oblique 2000 0.020 39.980 D 950.93 1.79 0.542Prieta

LA13 Newhall 6.7 6.7 3000 0.020 59.980 0 664.93 1.03 0.658(1989)

LA14 Newhall 6.7 6.7 3000 0.020 59.980 0 644.49 1.03 0.638

LAI5 Rinaldi RS 6.7 7.5 2990 0.005 14.945 0 523.30 0.79 0.675

LAI6 Rinaldi RS 6.7 7.5 2990 0.005 14,945 0 568.58 0.79 0.734Northridge

LA 17Northridge

6.7 6.4 3000 0.020 59.980 D 558.43 0.99 0.575(1994)

, Sylmar

LAI8Norlhridge

6.7 6.4 3000 0.020 59.980 D 801.44 0.99 0.825,SylmarNorth

LAI9 Palm 6.0 6.7 oblique 3000 0.020 59.980 D 999.43 2.97 0.343N. Palm

SpringsSprings(1986) North

LA2O Palm 6.0 6.7 oblique 3000 0.020 59.980 0 967.61 2.97 0.332Springs

Table D.2 Set of LA earthquake ground motions with 10% probability of exceedencein 50 years (SAC project)

(1)Moment magnitude

(2)Closest source-to-site distance

(3)Faulting mechanism

(4)Recording time interval

(5)Duration

242

1

PGAEQ D12 Number

DT14 D5 l0%/Event & File Station MW' I FM 3) of I

ST I

f0GM

Yearj

I Points I

5oyrs SF

s_ I sec ISCI

Imperial Long6.5 1.2 Strike-slip 3909 0.010 39.080 D 170.55 0.49 0.355

Beach,Valley(1979) SF02

Long6.5 1.2 strike-slip 3909 0.010 39.080 D 132.70 0.49 0.276

Beach,Morgan

SE03 Hill, 6.2 15.0 strike-slip 3000 0.020 59.980 D 378.82 2.84 0.136Morgan

GilroyHill

(1984)Morgan

SEO4 Hill, 6.2 15.0 strike-slip 3000 0.020 59.980 D 649.80 2.84 0.233GilroyWest

SEO5 WA, 6.5 56.0subduction 4000 0.020 79.980 0 376.18 1.86 0.206intraplale

WestSF06 WA, 6.5 56.0

subduclion 4000 0.020 79.980 0 345.11 1.86 0.189iniraplate

Olympia Olympia(1949) West

SF07 WA, 6.5 80.0subduction

3335 0.020 66.680 D 289.19 5.34 0.055Seattle___________intraplate

WestSE08 WA, 6.5 80.0

subduction3335 0.020 66.680 D 381.26 5.34 0.073

intraplaleSeattleNorth

SEO9 Palm 6.0 6.7 oblique 3000 0.020 59.980 0 576.45 1.71 0.344N. Palm SpringsSprings(1986)

NorthSEIO Palm 6.0 6.7 oblique 3000 0.020 59.980 D 558.10 1.71 0.333

Springs

SEt IWA,

7.1 80.0subduction 4092 0.020 81.820 0 737.82 4.30 0.175

Olympia, intraptate

SF12WA,

7.1 80.0subduction 4092 0.020 81.820 D 584.52 4.30 0.139

Olympia, intraplate

WA,SF13 Federal 7.1 61.0

subduction 3705 0.020 74.080 0 362.31 5.28 0.070intraplate

Seattle OFC B(1949) WA,

SF14 Federal 7.1 61.0subduction 3705 0.020 74.080 D 297.30 5.28 0.057intraplate

SEISTacoma

7.1 60.0subduction

3000 0.020 59.980 13 284.72 8.68 0.033County inlraptate

SEI6 Tacoma7.1 60.0

subduction3000 0.020 59.980 D 563.47 8.68 0.066

County intraplale

SEI7 Llolleo,8.0 42.0

subduclion 4000 0.025 99.975 0 684.27 1.24 0.563Chile inlerplate

SF18Llolleo,

8.0 42.0subduction 4000 0.025 99.975 13 657.89 1.24 0.541

Chile interplate

Valparaiso Vinadel(1985) SF19 Mar, 8.0 42.0

subduclion4000 0.025 99.975 13 531.05 1.69 0.320

interplate

VinadelSF20 Mar, 8.0 42.0

subduction 4000 0.025 99.975 D 376.88 1.69 0.227interplate

Chile

Table D.3 Set of Seattle earthquake ground motions with 10% probability ofexceedence in 50 years (SAC project)

(1) Moment magnitude(2) Closest source-to-site distance(3) Faulting mechanism(4) Recording time interval(5) Duration

243

1i:

PGAEQ

D 2 Number DT (4) D3Event & File Station I FM 3) of I

ST °Year I Points

km j sec[

see j cn5sc

Simulation,BOOl Hanging 6.5 30.0 reverse 3000 0.010 29.990 121.97 0.39 0.319

Reverse WallSimulation,

B002 Hanging 6.5 30.0 reverse 3000 0.010 29.990 D1t61 72.93 0.39 0.191

Wall

B003 Simulation,6.5 30.0 reverse 3000 0.010 29.990 D161 141.37 0.54 0.267

Reverse Foot Wall2

B004 Simulation,6.5 30.0 reverse 3000 0.010 29.990 D16t 109.65 0.54 0.207

Foot WallNew

BOOSNew

4.3 8.4 reverse 3847 0.005 19.230 B 564.78103 0.054

Hampshi hampshirere

B006 New4.3 8.4 reverse 3847 0.005 19.230 0 309.51

10.70.029

(1982) hampshire 5

B007 Nahanni 6.9 9.6 4068 0.005 20.335 i5 86.29 0.09 0.978

BOO8 Nahanni 6.9 9.6 4068 0.005 20.335 81.18 0.09 0.920

Nahanni B009 Nahanni 6.9 6.1 3752 0.005 18.755 59.48 0.20 0.303(1985) BOlO Nahanni 6.9 6.1 3752 0.005 18.755 72.23 0.20 0.368

B011 Nahanni 6.9 18.0 3804 0.005 19.015 i5Y 130.69 0.92 0.145

B0l2 Nahanni 6.9 18.0 3804 0.005 19.015 133.21 0.92 0.148

BOI3 Saguenay 5.9 96.0 reverse 3548 0.005 17.735 D 196.50 1.57 0.128

BOI4 Saguenay 5.9 96.0 reverse 3548 0.005 17.735 D 268.44 1.57 0.174

BOI5 Ssguenay 5.9 980 reverse 2958 0.010 29.570 D1 513.58 3.21 0.163

Ssguena B0l6 Saguenay 5.9 98.0 reverse 2958 0.010 29.570 243.68 3.21 0.077

y (1988) B0l7 Ssguenay 5.9 118.0 reverse 3906 0.010 39.050 179.47 3.25 0.056

BO18 Ssguenay 5.9 118.0 reverse 3906 0.010 39.050 222.98 3.25 0.070B0l9 Ssguenay 5.9 132.0 reverse 3325 0.010 33.240 172.96 3.34 0.053

B020 Saguenay 5.9 132.0 reverse 3325 0.010 33.240 267.23 3.34 0.082

Table D.4 Set of Boston earthquake ground motions with 10% probability ofexceedence in 50 years (SAC project)

(I) Moment magnitude(2) Closest source-to-site distance(3) Faulting mechanism(4) Recording time interval(5) Duration(6) Rock converts to soil

244

APPENDIX E: Peak displacement distributions considering different R factors

245

0.9

0.8

0.7

0.6

0.5U-

0.4

0.3

0.2

0.1

n

/ R2.5, W1940 lbs/ft

/ / / R=3.5, W=2720 lbs/ft

/ /R4.5, W3500 lbs/ft

I R=5.5 W=4270 lbs/ft

I//8ft.

/ / //

8ft.

// BW (8 x 8), 8d@2"/1 2, OSB (/8)

J I ED=3/8',G= 185ksi,=2%, 10(50/50)

0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2

6max (in.)

Figure E.1 Peak displacement distributions considering different R factors (2"/12",JO)

0.9

0.8

0.7

0.6

0.5U-

0.4

0.3

0.2

0.1

U

-

/ ;5 W=3500 lbs/ft

// R=5.5, W4270 lbs/ft

I /

I ///,, __I / ', 8ft.II, 0

'i" BW (8' x 8'), 8d@2"/12", 0S (/"),

ED=3/8",G=185ksi,2°4,LS(10/50)

0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 5.5 6

max (in.)

Figure E.2 Peak displacement distributions considering different R factors (2"/12",LS)

246

0.9

0.8

0.7

0.6

0.5

0.4

0.3

0.2

0.1

0

R=2.5, W=1940 lbs/ft

R3.5, W=2720 lbs/ft

R=4.5, W=3500 lbs/ft

R=5.5, W4270 lbs/ft

8ft.

8ft.

/ 0 BW(8>< 8'), 8d@2/12', OSB (/')ED=3/8',G= 185ksi,=2%,CP(2/5O)

0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 5.5 6

max (in.)

Figure E.3 Peak displacement distributions considering different R factors (2"/12",CP)

0.9

0.8

0.7

0.6

0.5

0.4

0.3

0.2

0.1

n

R2.5, W1147 lbs/ft

:=:: c::::: :R=5.5, W2521 lbs/ft

II

1

/

8LH

/ x 8'), 8d@4"112", OSB (/8"),

- ED=3/6",G=l8Oksi,ç=2%, 10(50/50)

0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2

max (in.)


247

0.9

0.8

0.7

0.6

0.5

0.4

0.3

0.2

0.1

n

- ----

/

/ /

/ / / R=4.5, W=2065 lbs/ft

/ / /R=5.5, W2521 lbs/ft

/LI

8ft.

/

1/" 8

BW(8 x8), 8d©4 /12, OSB (/ ),

- ED = G = 180 ksi, = 2%, LS (10/50)

0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 5.5 6

max(fl.)


0.9

0.8

0.7

0.6

LL

0.4

0.3

0.2

0.1

n

R=2.5, W=1 147 lbs/ft

R=3.5, W=1605 lbs/ft

R=4.5, W=2065 lbs/ft

R=5.5, W=2521 lbs/ft

/ /1

HH// 8ft.

° BW(8 x 8'), 8d@4"/12", OSB (/"),

ED3/8",G180ksi,2%,CP(2/50)

0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 5.5 6

6max (in.)


0.9

0.8

0.7

0.6

0.5LL

0.4

0.3

0.2

0.1

n

R=2.5, W=73OthsIft

7 :"L :::: :R=5.5, W1610 lbs/ft

Il

/ 8ft.

/ 8ft.

/ BW (8 >< 8), 8d@6/12", OSB (/)

ED=3/G 185ksi,=2%, 10(50/50)

0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2

6max (in.)


0.9

0.8

0.7

0.6

0.5U-

0.4

0.3

0.2

0.1

n

/

/ 4R2.5, W730 lbs/ft

/ /R3.5, W=1026 lbs/ft

/R4.5, W=1321 lbs/ft

I / / R5.5

//

'r/

Bft.

/ /,'/!'/ BW (8 x 8), 8d@6"/12, OSB (/8),

- ED =/", G= 185 ksi,ç=2%, LS(10/50)

0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 5.5 6

max (in.)


249

0.9

0.8

0.7

0.6

0.5LI

0.4

0.3

0.2

0.1

n

R=2.5, W=730 lbs/ft

R=3.5, W=1026 lbs/ft

R=4.5, W=1321 lbs/ft ,'R=5.5, W=1610 lbs/ft '1

8ft.

- / ° BW(8 x 8), 8d©6/12, OSB (/'),v-" ED =I', G = 185 ksi, = 2%, CP(2150)

0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 5.5 6

max (iii.)


250

APPENDIX F: Fragility curves for baseline wall (BW1) considering different hazardlevels

0.9

0.8

0.7

0.6

o 0.5

0.4

0.3

0.2

0.1

(1

0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2

Sa(g)

Figure F.l Fragility curves (R = 2.5, 2"/12")

0.9

0.8

0.7

0.6

o 0.5

0.4

0.3

0.2

0.1

n

8ft. / /BW (8 < 8'), 8d@3"/12", OSB (/8"), /ED=3/8",G=200ksi,/=2%, R=2.5/W = 1400 lbs/ft (52.1 kN total)

/

/ /

/

0 (50/50)

/ LS(10/50)

0 0.5 1 1.5

Sa(g)

Figure F.2 Fragility curves (R = 2.5, 3"/12")

2 2.5

251

0.9

0.8

0.7

0.6

o 0.5

0.4

0.3

0.2

0.1

0

//70 (50/50) / /LS

/// /

/ // / BW(8 x 8'), 8d@4'/12", OSB (/8"),/ / ED=3/5",G=180ksi,=2%,R=2.5/ W = 1147 lbs/ft (40.8 kN total)

0 0.5 1 1.5 2

Sa(g)


1

0.9

0.8

0.7

0.6

0.5

0.4

0.3

0.2

0.1

0

8ft. / /8W (8' x 8), 8d@6"/12", OSB (/'), /ED 3/, G 185 ksi, = 2%, R = 2.5 /W = 731 lbs/ft (26.0 kN total) /

//

// IO (50/50)

/ /LS1O/5O

0 0.5 1 1.5 2

Sa(g)


2.5

2.5

252

1

0.9

0.8

0.7

0.6

0.5

0.4

0.3

0.2

0.1

n

88.

SW (8' x 8'), 8d@2"/12", OSB

ED 3/, G = 185 ksi, ç = 2%, R -3.5

W = 2720 lbs/ft (96.8 kN total)

//

10 (50/50)

/

//

/ /,,,'

It,,,

0 0.5 1 1.5

Sa(g)

Figure F.5 Fragility curves (R 3.5, 2"/12")

1

0.9

0.8

0.7

0.6

0 o.s

0.4

0.3

0.2

0.1

n

2 2.5

88.

BW (8' x 8'), 8d@3"/12", OSB (3/../

ED'3/g",G'200ksi,/'2%, Rfr3.5W = 2051 lbs/ft (73.0 kN total)

/ ' 0 (50/50)

/ / '

/ /' CP (2/50)

//,,,,//,"

0 0.5 1 1.5

Sa(g)

Figure F.6 Fragility curves (R 3.5, 3"/12")

2 2.5

253

0.9

0.8

0.7

0.6

0.4

0.3

0.2

0.1

0

0 (50/50) /LS (10/50)

CP (2/50)

/ /,, J6

86.

/ / BW (8' x 8'), Sd©4'112", OSB (I"),

/ / ED = J8", G = 180 ksi, ç = 2%, R = 3.5

,/ - W = 1605 lbs/ft (57.1 kN total)

0 0.5 1 1.5

Sa(g)


I

0.9

0.8

0.7

0.6

o 0.5

0.4

0.3

0.2

0.1

(1

86.

8ft. //BW (8' x 8'), 8d©6"/12", OSB (/"), /ED=3/8',G= 185ksi,=2%, R=3.5/W = 1026 lbs/ft (36.5 kN total)

/ /CF (2/50)

0 0.5 1 1.5

Sa(g)


2 2.5

2 2.5

254

0.9

0.8

0.7

0.6

0.5

0.4

0.3

0.2

0.1

C)

/ /lO (50/50)

//

,.LS(10/50)

/CP (2/50)

/ /,,,

/ 8ft.

88.

/ BW (8 x 8), 8d@2"/12", OSB (/8"),

/ ED=3/8",G=l85ksi,ç=2%,R=4.5

-W = 3501 lbs/fl (124.6 kN total)

0 0.5 1 1.5

Sa(g)


I

0.9

0.8

0.7

0.6

o 0.5

0.4

0.3

0.2

0.1

n

2

/ /0 (50/50)

/ /,,"

::

/ /,,,'88.

/ //88.

/ / BW (8 x 8'), 8d@3"112", OSB (I8"),/ /, ED=3/8",G200ksi,l2%, R=4.5W = 2637 lbs/ft (93.9 kN total)

0 0.5 1 1.5

Sa(g)

Figure F.1O Fragility curves (R = 4.5, 3"/12")

2.5

2 2.5

255

0.9

0.8

0.7

0.6

o 0.5

0.4

0.3

0.2

0.1

11

/ 10 (50/50)

/ / LS(10/50)

// I

CP (2/50)

/ /,,,,

/ /,,,'

/ 88.

8ff.

IBW (8 x 8), 8d@4"112", OSB (/8"),ED=3f8G- 180ksi,=2%, R=4.5W = 2065 lbsfft (73.5 kN total)

0 0.5 1 1.5

Sa(g)

Figure F. 11 Fragility curves (R = 4.5, 4"/12")

0.9

0.8

0.7

0.6

a 0.5

0.4

0.3

0.2

0.1

0

2 2.5

/ 10 (50/50)

/ /

/

/ /,,,

/ /:'88.

/ 8ft.

/ /' BW (8' x 8), 8d©6112", OSB (I8'),

/ ED=3/8",G=185ks1,ç=2%,R=4.5,/ / = 1321 lbs/ft (47.0 kN total)

0 0.5 1 1.5

Sa(g)

Figure F. 12 Fragility curves (R = 4.5, 6"/12")

2 2.5

256

257

APPENDIX G: Fragility curves for baseline wall (BW1) considering different Rfactors and nailing schedules

258

1

0.9

0.8

0.7

0.6

ci:- 0.5

0.4

0.3

0.2

0.1

('I

8d©3"/12", W1464 lbs/ft

8d@4"/12", W=1147 lbsIft/8d©2"/12", W=1940 lbs/ft /8d©6"/12", W730 lbs/ft /

7',,"

/,8ff.

/, SW (8 x 8'), OSB(31,"), ED =

- - ?=2%, R=2.5, 10 (50/50)

0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2

Sa(g)

Figure G. 1 Fragility curves considering R = 2.5 (JO, 50/50 hazard level)

1

0.9

0.8

0.7

0.6

o 0.5

0.4

0.3

0.2

0.1

n

8d@3"/12", W2051 lbs/ft 7/ ,'8d@4"/12", W1605 lbs/ft / I

8d@2"/12", W2720 lbs/ft / /8d@6"/12", W=1026 Ibs/ft77'/ /

/,

/'//'

/ /// __8ff.

/ // BW (8' 8'),OSB (3/,), ED =

-' /' =2%, R=3.5, 10(50/50)

0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2

Sa(g)

Figure G.2 Fragility curves considering R = 3.5 (JO, 50/50 hazard level)

259

0.9

0.8

0.7

0.6

0.5

0.4

0.3

0.2

0.1

n

/

/ 8d3"/12", W2639 lbs/ft,/ / 8d(4"/12", W2065 lbs/ft

/ // 8d@2"/12", W3500 lbs/ft

,,' / / 8d@6"/12", W1321 lbs/ft//// ///

/ /// __8ft.

/ //'/ BW (8' x 8'), OSB (/8"), ED = /8,

/ ç = 2%, R = 4.5, 10 (50/50)

0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2

Sa(g)

Figure G.3 Fragility curves considering R = 4.5 (10, 50/50 hazard level)

1

0.9

0.8

0.7

0.6

0.5

0.4

0.3

0.2

0.1

n

/ 8d3"/12", W3220 lbs/ft

//I8d@4"/12", W2521 lbs/ft

/// 8d2"/12", W4270 lbs/ft

/// 8d@6"/12", W1610 lbs/ft

////

1/

/i'8 ft.

/ BW (8' x 8'), OSB (/8"), ED =

2%, R = 5.5, 10 (50/50)

0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2

Sa(g)

Figure G.4 Fragility curves considering R = 5.5 (10, 50/50 hazard level)

I

260

1

0.9

0.8

0.7

0.6

Q0.5

0.4

0.3

0.2

0.1

n

;

8d©4"/12", W=1605 lbs/ft

8d@3"/12", W=2051 Ibs/ft

/ H/ BW (8 8'), OSB e/8"), ED =,

ç = 2%, R = 3.5, CP (2/50)

0 0.5 1 1.5 2 2.5

Sa(g)

Figure G.5 Fragility curves considering R = 3.5 (CP, 2/50 hazard level)

1

0.9

0.8

0.7

0.6

d 0.5

0.4

0.3

0.2

0.1

n

88. 188. ///

BW (8' x 8'), OSB (3/e") ED = /8",

ç=2%, R=4.5,CP(2/50) /

/ 8d©3"Il 2", W=2639 lbs/ft

0 0.5 1 1.5 2 2.5

Sa(g)


261

1

0.9

0.8

0.7

0.6

0.5

0.4

0.3

0.2

0.1

0

0

_8ff.BW (8' x 8), OSB (/'), ED =

= 2%, R = 5.5, OP (2/50)

8d@6"/12", W=1610 lbs/ft

8d@3"/12", W=3220 lbs/ft

8d@4"/12", W=2521 lbs/ft

8d©2"/12", W=4270 lbs/ft

0.5 1 1.5 2 2.5

Sa(g)


262

APPENDIX H: CDF for baseline wall (BW1) considering ultimate force with variousR factors

263

0.9

0.8

0.7

0.6

0.5U-

0.4

0.3

0.2

0.1

n

R=2.5

'1

I /

/I

/ / H /

I/j /

/ /

I /

/ / i, aft.

/ / 2 BW (8' x 8'), 8d@2"/12", OSB (/8").

j, ED I8", G = 185 ksi, = 2%, 0 (50/50)

2000 4000 6000 8000 10000 12000 14000 16000

Fmax (lbs.)

Figure H.1 CDF for ultimate force with various R factors (2"/12", JO)

0.9

0.8

0.7

0.6

0.5

0.4

0.3

0.2

0.1

H

H

H/

H / //

,''

/ /

//

,,/

// '1

8ft.

/ / ,'/ 88.

/ / " BW (8' >< 8'), 8d@2"/12", OSB (/8"),

_--2' ED=3/8",G l85ksi,ç=2%, LS(10/50)

6000 8000 10000 12000 14000 16000 18000 20000

Fmax (lbs.)

Figure H.2 CDF for ultimate force with various R factors (2"/12", LS)

264

0.9

0.8

0.7

0.6

0.5

0.4

0.3

0.2

0.1

C)

/1R=2.5 / /R=3.5

i/hV/ /R=4.5

II

R5.5

'I 88.

/'/7,, 88.

/ :' BW (8 x 8'), 8d@2"112", OSB (/8"),

ED = G = 185 ksi, = 2%, CP(2/50)

10000 12000 14000 16000 18000 20000

Fmax (lbs.)

Figure H.3 CDF for ultimate force with various R factors (2"/12", CP)

0.9

0.8

0.7

0.6

0.5U-

0.4

0.3

0.2

0.1

n

8L.

88.

:2.5

BW (8x8'), 8dq4"/12" / / ,'/ R3.5

OSB(/8),ED= /8, I / 'W.JG 180 ksi, = 2%, I ,' /- R=4.510(50/50) 1 I

R=5.5

/

I II '/ II

I/ ,' /

"

/ /,,'// 1(1)1::

___/ __/_____ -J

0 2000 4000 6000 8000 10000

Fmax (lbs.)


265

.1

0.9

0.8

0.7

0.6

0.5U-

0.4

0.3

0.2

0.1

n

4000 5000 6000 7000 8000 9000 10000 11000 12000

Fmax (lbS.)


0.9

0.8

0.7

0.6

0.5U-

0.4

0.3

0.2

0.1

n

4000 5000 6000 7000 8000 9000 10000 11000 12000

Fmax (lbs.)


266

0.9

0.8

0.7

0.6

40.5

0.4

0.3

0.2

0.1

n

88.

8ft. /BW (8 x 8'), 8dc6'I12", / / /OSB (I8"), ED = a" / / R=2.5G=l85ksi,ç=2%, /0(50/50)

/ /" R=3.5

I / R=4.5

/ ,' R=5.5

1000 2000 3000 4000 5000 6000Fmax (lbs.)


0.9

0.8

0.7

0.6

40.5

0.4

0.3

0.2

0.1

n

/ IT",;,

/ ,,/ R=5.5

/ ,,,/I

/ "788.

88.

/ ," / BW (8' x 8'), 8d6"/12",/ m OSB(3/8"),ED /8",/ ' / G=185ksi,=2%,

__/ LS(10150)

2000 3000 4000 5000 6000 7000 8000Fmax (lbs.)


267

1

0.9

0.8

0.7

O.5

0.4

0.3

0.2

0.1

0

R=4.5/

R=5.5

/8ft.

(0(0

8ft.I ii

/ / 8W (8' x 8'), 8d©6"/12", OSB (/8"),

/ / ED G = 185 ksi, = 2%,

/ CF (2/50)-J

3000 4000 5000 6000 7000 8000

Fmax (lbs.)


the objective of this research was to explore the

Documents