Seismic performance investigation of Timber Shear Walls with Sprayapplied polyurethane foam infill

1
EXPERIMENTAL AND ANALYTICAL SEISMIC PERFORMANCE
INVESTIGATION OF SPRAY-APPLIED POLYURETHANE FOAM
INFILL TIMBER SHEAR WALLS
April 25, 2008 Darius Dodge, URS Corporation,
San Francisco Charles Chadwell, Ph.D., P.E.,
California Polytechnic State University, San Luis
Obispo
2
Overview

• Preliminary Interest
• Experimental Investigation
• Full-Scale Wall Testing
• Analytical Investigation
• Computer Based Earthquake Simulations
• Conclusions and Recommendations

3
Preliminary Interest

• NCFI Polyurethanes suggested duel use for
  spray-applied polyurethane foam (SPF) infill
• Currently used as thermal insulation in North
  America
• Additional strength for lateral force resisting
  systems (LFRS)
• NAHB testing showed a 22 increase in elastic
  stiffness
• Monotonic loading protocol used
• Non-standard construction methods used

4
Full-Scale Wall Testing

• CUREE-Caltech pseudo-static protocol used
• Provides good representation of energy applied
  during an earthquake (Krawinkler et al, 2001)
• Representative of dynamic testing
• Multiple dimension, standard construction timber
  shear walls
• 4 feet by 8 feet (Satisfy 21 aspect ratio for
  timber shear walls to achieve full
  capacity, IBC2006)
• 8 feet by 8 feet
• One control wall for each dimension, several
  walls with SPF

5
Full-Scale Wall Testing

• Representative boundary conditions applied
• Lateral and torsional buckling restrained
• 130 kip large displacement actuator
• Moveable concrete foundation

6
Full-Scale Wall Testing

• Representative boundary conditions applied
• Lateral and torsional buckling restrained
• 130 kip large displacement actuator
• Moveable concrete foundation

7
Results of Full-Scale Wall Testing

• 4 feet x 8 feet walls showed that SPF did not
  have a significant effect upon wall behavior

• 9.45 decrease in Yield Strength
• 1.75 increase in Peak Strength
• 14.9 decrease in Elastic Stiffness
• 40.5 increase in Post-Yield Stiffness

8
Results of Full-Scale Wall Testing

• Bond loss, found along perimeter members
• Area of debonding approx. 1 inch to1.5 inch from
  interior face of perimeter members
• Little debonding between SPF and OSB

9
Results of Full-Scale Wall Testing

• 8 feet x 8 feet walls showed that SPF did have a
  significant effect upon wall behavior

• 1.7 increase in Yield Strength
• 12.2 increase in Peak Strength
• 0.4 increase in Elastic Stiffness
• 129.1 increase in Post-Yield Stiffness

10
Computer Based Earthquake Simulations

• Nonlinear time-history analyses conducted with
  range of one- and two-story structures
• SDOF pinched hysteretic force-displacement
  articulation model (Ibarra et al, 2005)
• Three probability of exceedence ground motion
  suites3
• 10 in 50 years (475 year return period)
• 2 in 50 years (2475 year return period)
• 50 in 50 years (72 year return period)

3Ground motion suites from SAC, 1997
11
Computer Based Earthquake Simulations

• Model required hysteretic backbone properties
• Original backbones obtained from testing not used
  for calibration of model
• A bounding Hysteresis adopted as appropriate
  alternative

12
Computer Based Earthquake Simulations

• Model required hysteretic backbone properties
• Original backbones obtained from testing not used
  for calibration of model
• A bounding Hysteresis adopted as appropriate
  alternative

FEMA 356 Backbone
13
Computer Based Earthquake Simulations

• Proper behavior for each modeled structure was
  obtained by developing scaling relationships for
  wall properties

Py Yield Force
Pu Peak Force
Pc Peak Force
?1 Force Scale Factor
14
Computer Based Earthquake Simulations

• Proper behavior for each modeled structure was
  obtained by developing scaling relationships for
  wall properties

ke Initial Elastic Stiffness
?2 Stiffness Scale Factor
as Post-Yield Stiffness Coefficient
ac Post-Peak Stiffness Coefficient
15
Computer Based Earthquake Simulations
(1)

• Force scaling factor

(2)
Stiffness scaling factor
Hysteretic energy scaling factor
(3)
Units in lbs, inches
16
Computer Based Earthquake Simulations

• Scaled model calibrated by comparing model to lab
  results, using CUREE-Caltech displacement
  protocol
• Degradation properties controlled by comparison
  of absorbed hysteretic energy per cycle, to total
  capacity

Degradation properties controlled by ß-factor,
above
17
Computer Based Earthquake Simulations
Design Response Spectra
Ground Motions Obtained from SAC, 1997
18
Results From Computer Based Earthquake Simulations
One Story, Probability of Exceedence 10 in 50
Years
Two Story, Probability of Exceedence 10 in 50
Years
19
Results From Computer Based Earthquake Simulations
One Story, Probability of Exceedence 2 in 50
Years
Two Story, Probability of Exceedence 2 in 50
Years
20
Results From Computer Based Earthquake Simulations
One Story, Probability of Exceedence 50 in 50
Years
Two Story, Probability of Exceedence 50 in 50
Years
21
Results From Computer Based Earthquake Simulations
Ductility Demand ResultsPercentage Reduction
Using SPF

• Correlates to variation in response with
  building weight
• Maximum and average values vary considerably in
  certain cases
• Differences in hysteretic energy capacity
  conjectured to influence results

Drift Demand ResultsPercentage Reduction Using
SPF
22
Conclusions

• Walls of standard height and at the 21 aspect
  ratio were not shown to benefit with SPF infill
• Walls of a 11 aspect ratio were shown to have a
  12 increase in peak strength, much higher
  post-yield stiffness
• SPF bond fails well below expected drift demands
  (1.05)may significantly influence performance
• Brittle bond fracture seen to influence behavior
  past 1.05 drift

23
Conclusions

• Reduction in demands using SPF can be as much as
  40, but is not consistent for every practical
  case considered
• Demand reduction varies between ground motion
  suites or building periods of interest
• Certain cases do show a gain in demands

24
Final Conclusions and Recommendations

• Increased SPF bond strength should be further
  pursued
• Larger increase in strength seen prior to bond
  loss than at peak
• SPF/Timber system has advantages for use
• Further development encouraged

25
Acknowledgements
North Carolina Foam Industries, Inc.
California Polytechnic State University, San Luis
Obispo
Daniel Jansen, Ph.D., Cal Poly Civil and
Environmental Engineering Faculty
URS Corporation
26

• Questions ?