Combining wood and viscoelastic material

David E. Lewicki1 and David W. Dinehart2 ABSTRACT Passive energy dissipation (PED) devices, when incorporated into a structure, work by adding supplemental damping to the system which reduces the structural response during an earthquake or high winds. Viscoelastic dampers are one type of PED device that has emerged as an innovative solution to improving the seismic performance and wind response of structures. Viscoelastic dampers have been especially successful at gaining widespread acceptance by the engineering and construction industries as evidenced by its increased application over the last 10 years. Traditionally, viscoelastic dampers have been used exclusively in steel and reinforced concrete structures; however, there have been some recent studies demonstrating the feasibility of using viscoelastic dampers in woodframe structures. (Dinehart et. al., 1999) These initial studies used dampers consisting of viscoelastic material bonded to steel plates, typical of devices found in steel and concrete structures. To date no attempts have been made to utilize viscoelastic material bonded to wood. The objective of this study is to show that wood and viscoelastic material can be combined effectively and to characterize and quantify the differences between viscoelastic dampers constructed from steel and wood. Tests were conducted on double-lap shear viscoelastic dampers of various thickness constructed with wood and steel to examine the performance under fully reversed cycles of loading. Tests were conducted on matching steel and wood dampers. The dampers were constructed using, 0.2 (5 mm), 0.4 (10 mm), and 0.6 (15 mm) thick VE, self adhesive, material with a shear area of 8 in2 (5161 mm2) per pad or 16 in2 (10323 mm2) per damper. Cyclic tests were conducted at frequencies of 0.1 and 0.5 Hz and strains of 10 and 50 percent. Each damper test was conducted twice for a total 48 tests. A comparison of the dampers constructed with the steel and wood demonstrates several important characteristics. Overall, when subjected to low strains and frequencies, viscoelastic material performed well when integrated with wood. The damper stiffness of the steel and wood dampers was within 4 percent while the energy dissipation capacity was within 8 percent. There were no signs of any degradation of stiffness or energy dissipation capacity during the constant amplitude cycling, indicating that the viscoelastic material maintained a complete bond to wood when subjected to the test parameters stated herein. Further testing will be conducted on these dampers overtime to investigate the bond integrity over time. These tests show that viscoelastic material and wood can be combined effectively. This finding directly leads to an increase in the utilization of viscoelastic material in wood structures. More importantly, these results allow engineers to investigate new applications of combining wood and viscoelastic material. Applications may include improving the dynamic performance of woodframe structures subjected to wind or earthquakes or reducing vibrations of floor systems. INTRODUCTION The response of civil structures subjected to dynamic loads has demonstrated the importance of structural ductility. The presence of ductility enables a structure to absorb wind or seismic energy. Classical design procedures require isolated critical regions to undergo plastic deformations to dissipate dynamic energy and reduce the corresponding inter-story deflections. This method, in practice, has enabled tall structures to resist the dynamic response associated with wind loads and seismic events. However, this method requires predetermined secondary members to undergo stressing beyond their elastic potential. The primary concern of this design method is the cost affiliated with maintenance and replacement
1

Graduate Assistant, Department of Civil and Environmental Engineering, Villanova University, 800 Lancaster Avenue, Villanova, PA 19085 2 Assistant Professor, Department of Civil and Environmental Engineering, Villanova University, 800 Lancaster Avenue, Villanova, PA 19085

of structural members. This method has been integrated into some of the current design codes such as the National Building Codes of Canada (NBCC) and the Structural Engineering Association of California (SEAOC). (Cherry and Filiatrault, 1993) In recent years, however, the integration of passive energy dissipation systems (PEDs) has been investigated to enhance ductility. Viscoelastic (VE) dampers have been on type of PED that has been widely accepted in steel construction to mitigate deflections associated with wind loads on relatively tall building over the past 20 year. The United States has incorporated 20,000 VE dampers in the 1969 construction of the Twin Towers in New York City, New York. (Mahmoodi et. al., 1987) The Columbia SeaFirst and Two Union Square building, in 1982 and 1988 respectively have also incorporated VE dampers to limit wind vibrations. (Nielson et. al., 1994) Recently, analytical studies have been performed to predict the seismic response of viscoelastic dampers. Various damper configurations and steel frames have been investigated in accordance to recorded seismic events. (Chang et. al., 1998, Kasai et. al., 1993, Zhang and Soong 1992, and Aiken et. al., 1990). Viscoelastic dampers have been thoroughly evaluated and shown to increase structural damping, reduce deflections and accelerations, and limit yielding of structural steel members. VE dampers have also been evaluated as supplemental seismic protection in the construction and retrofit of concrete structures. Many existing concrete structures do not maintain the necessary stiffness or ductility required in resisting a moderate earthquake. Supplemental dampers have been one method of retrofit examined in recent years to reduce the inter-story drifts and base shear. (Dumoulin et. al., 1998; Soong et. al., 1998; and Chang et. al., 1994) While VE dampers have gained popularity in the design and retrofitting of steel and concrete structures, it has gone largely overlooked in woodframe construction. Wood structures subjected to dynamic lateral loads have performed in accordance to current building codes. The good performance of woodframe structures subjected to cyclic lateral loads can be attributed to their low weight, redundant lateral load paths, good ductility, and their ability to dissipate large amounts of energy without collapsing. In this respect, wood structures have generally satisfied the spirit of the modern seismic building codes, the goal of which is to maintain life safety. However, those same attributes that allow the structure to maintain its overall integrity also lead to costly non-structural damage. Severe as well as non-structural damage can result in economic failure of a woodframe structure following a hurricane or seismic event. Hurricanes in the eastern United States have demonstrated the susceptibility of woodframe construction to severe winds. Hurricane Andrew, which struck Florida on August 24, 1992, damaged or destroyed more than 60,000 WF homes, with an associated cost exceeding 20 billion dollars (Keith and Rose, 1994). Statistics from the 1994 Northridge earthquake demonstrated that woodframe structures are also vulnerable to earthquakes. In Los Angeles County, roughly 60,000 residential units were significantly damaged and were uninhabitable (Holmes and Somers, 1995). The cost of the damage to WF structures from the Northridge earthquake was estimated at over 20 billion dollars (CUREe No. W-01, 1999). Research has been conducted on implementing VE dampers into standard wood shear walls (Dinehart et. al., 1999). This research involved using traditional VE dampers, typical of those used in steel and concrete structures, that utilized shear dampers with layers of steel bonded to the VE polymer. The emphasis of this investigation is to evaluate the performance of self-adhesive viscoelastic material bonded exclusively to untreated wood. This is the first such study into the behavior of adhering VE material directly to wood surfaces. TEST SPECIMEN Viscoelastic material, ISD Polymer 111, manufactured by 3M was used to construct the dampers tested. The dampers were fabricated into a double lap configuration to load in pure shear. A total of six dampers were formed; three wood dampers of VE layer thickness 0.2 (5 mm), 0.4 (10 mm) and 0.6 (15 mm). Three steel dampers of identical specifications were constructed to serve as the baseline for the various cyclic tests. Each damper was subjected to four test regimens utilizing different combinations of strain and frequency. The three steel dampers were fabricated from A36 3/8 thick (9.5 mm), 2 (50.8 mm) width bar stock. Viscoelastic pads used in the damper construction measured 2 by 4 (50.8 mm by 101.6 mm) with varying thicknesses of 0.2 (5 mm), 0.4 (10 mm) and 0.6 (15 mm) respectively. The double lap configuration, shown in Figure 1, was designated such that loading would be pure shear in the VE layers. The wood damper design is shown in Figure 1 (b). Four-ply APA (American Plywood Association) rated, 15/32 (11.9 mm), sheathing and SPF (Spruce-Pine-Fur) 2 by 4, 1.5 by 3.5 (38

mm by 89mm) net, studs were selected to simulate use of the material between a typical sheathing-to-stud wood connection as shown in Figure 2. Self adhesive viscoelastic material was bonded to the A36 bar stock after proper cleansing of the surface metal with acetone to remove oil and moisture. Viscoelastic material bonded the stud to sheathing after smoothing the surface with fine-grain sandpaper. Following the bonding of the VE material to the steel and wood respectively, pressure was applied to the dampers such that the VE material was compressed to 20% of its original thickness. The constant pressure was applied for 24 hours to ensure a complete bond.
VE Material A36 Plate (TYP) Hex Bolts (TYP)

3/8 (TYP)

t1 = 49/64 (0.2 VE Pads) t2 = 1 11/64 (0.4 VE Pads) t3 = 1 37/64 (0.6 VE Pads)

16

(a): Steel

VE Material Plywood - 4 ply SPF Stud Hex Bolts (TYP)

15/32 1/2 (TYP) t1 = 29/32 (0.2 VE Pads) t2 = 1 19/64 (0.4 VE Pads) t3 = 1 11/16 (0.6 VE Pads)

1 1/2

11

21 1/2

(b): Wood Figure 1: Damper Dimensions

 (t) =

sin t o

2x4 - Stud Sheathing

Fastener

Viscoelastic Material

Figure 2: Possible use of VE material in a wood connection. TEST SETUP AND PROCEDURE All dampers were cycled using a procedure conducted in displacement control at a constant frequency for ten fullyreversed cycles of constant amplitude. The test protocol for the 0.1 in (5 mm) thick damper cycled at 0.5 Hz is shown in Figure 3. The cyclic tests were performed in a 110 kip (489 KN) capacity MTS (Material Testing System) TestStar Workstation. All tests were conducted at an ambient temperature of 65F (+/-2F).

10 cycles
0.1

Displacement (in

0.05

-0.05

-0.1 0 2 4 6 8 10 Time (sec) 12 14 16 18 20

Figure 3: Fully Reversed Dynamic Load

Tests were conducted on three different matched wood and steel dampers for four different combinations of strain (10% and 50%) and frequency (0.1 Hz and 0.5 Hz). All six dampers were tested twice in accordance to the four individual procedures combining for a total of 48 tests. The test setups of the steel and wood dampers are shown in Figure 4(a) and 4(b), respectively. Three independent slave channels within a digital controller recorded time, load, and displacement for the various tests. Windows-NT Testware-SX, templates were programmed in the PC interface to perform the tests and serve as the data acquisition system. Data was recorded through the TestStar system at a sampling rate of 100 data points per second.

(a): Steel Figure 4: Damper Test Setups TEST RESULTS

(b): Wood

For each test time histories and hysteresis plots were created, and effective stiffness and energy dissipation was calculated for each test cycle. An idealized hysteresis loop for a viscoelastic material is shown in Figure 5.
P(t) Ki = Effective Stiffness

(t)

EDi = Area of Ellipse

Figure 5: Theoretical Hystersis Loop. The stiffness of the damper in cycle i was calculated as
+ +

K Ei =

Pi Pi

i i

(1)

in which Pi+ is the force corresponding to the maximum positive displacement, i+, and Pi- is the force corresponding to the maximum negative displacement, i-. The energy dissipation for each cycle is the area bounded by the ellipse. A

representative hysteresis plot for the dampers tested is presented in Figure 6. This hysteresis plot corresponds to a steel damper with 0.2 (5 mm) thick viscoelastic pads tested at 0.5Hz and 50% strain.

600 400 Load (lb) 200 0 -200 -400 -600 -0.1 -0.05 0 Displacement (in) 0.05 0.1

Figure 6: Representative Hysteresis Temperature increases have been shown to have an adverse effect on performance of VE dampers (Chang et. al., 1998; Tsai, 1994). Dinehart, 1998, showed that VE damper stiffness and energy dissipation capacity degrade during constant amplitude cycling at strains greater than 50%. Upon inspection of the damper responses tested at 10 and 50 percent strain at 0.1 and 0.5 HZ, no degradation of stiffness or energy dissipation was observed for the wood or steel dampers during the constant amplitude cycling. Consequently, the overall effective stiffness for a given strain, frequency, and temperature was simply be taken as the average stiffness of the individual cycles. Similarly, the total energy dissipation for each specific test was taken as the average of the individual cycles. Stiffness and energy dissipation capacities were averaged for the two separate replicate tests. A summary of the stiffness and energy dissipation for the wood and steel dampers tested at 10 percent strain is presented in Table 1. A summary of the stiffness and energy dissipation for the dampers tested at 50 percent strain is presented in Table 2. Table 1: Average steel and wood damper properties at 10% strain.

Stiffness, KE (lb/in) 0.1 Hz 0.5 Hz VE Thickness Steel Wood Steel Wood 0.2 1583 1818 3329 3372 0.4 782 773 1488 1499 0.6 530 535 887 925 Energy Dissipation, ED (lb-in) 0.1 Hz 0.5 Hz VE Thickness Steel Wood Steel Wood 0.2 2.1 2.1 2.1 2.2 0.4 2.4 2.5 3.6 3.9 0.6 2.5 2.9 4.7 5.1

Table 2: Average steel and wood damper properties at 50% strain.

Stiffness, KE (lb/in) VE Thickness 0.1 Hz 0.5 Hz (in) Steel Wood Steel Wood 0.2 1469 1503 2872 2579 0.4 713 722 1320 1304 0.6 427 416 759 800 Energy Dissipation, ED (lb-in) VE Thickness 0.1 Hz 0.5 Hz (in) Steel Wood Steel Wood 0.2 22.0 23.1 47.7 38.9 0.4 40.1 40.7 83.4 76.1 0.6 50.0 53.7 103.2 115.2
Table 1 clearly shows that the stiffness and energy dissipation of the wood and steel dampers were very similar, with the wood dampers exhibiting slightly higher stiffnesses and energy dissipations. For the dampers tested at 10 percent strain, the average stiffnesses of the wood and steel dampers were within 4 percent, while the energy dissipation varied on average by 6 percent. Table 2 shows that the performance of the wood and steel dampers at 50 percent strain were also similar. For the dampers tested at the higher strain the stiffnesses remained within 4 percent for the steel and wood dampers and the energy dissipations varied by approximately 9 percent. Note that the variation of the energy dissipation was greatest at a strain of 0.5 Hz and 50 percent strain. Further testing will be conducted on these dampers at higher frequencies and higher strains. CONCLUSIONS This study has been conducted to assess the use of viscoelastic material bonded directly to untreated wood. Twenty-four tests were performed on duplicate steel and wood dampers, to compare stiffness and energy dissipation capacities for various frequencies and strains. Effective stiffness of the wood and steel dampers was within 4%, while the energy dissipation showed an 8% difference. In most cases the wood dampers were stiffer and dissipated more energy than the steel dampers. These results suggest that the self-adhesive VE material can be bonded directly to wood without any reduction in VE material performance. These preliminary results show that combining VE material and wood to reduce the structural response of woodframe structures is feasible. An experimental test program is underway to investigate the application of VE sheet material between the sheathing and frame of shear walls to increase the wall performance under dynamic loading. This is only one possible application. The use of this material in wood structures is only limited by the innovation of the design engineers. Further work needs to be done to test the wood-VE performance at higher frequencies and higher strains, and testing should be conducted on wood dampers to investigate the integrity of the bond over time. ACKNOWLEDGMENTS The authors would like to extend their gratitude to Mr. M. Tanaka at 3M for arranging for the donation the VE material required for this study. We would also like to thank George Papas, a civil engineering technician at Villanova University, for aiding in the construction of the dampers. REFERENCES Aiken, I. D., Kelly, J. M., and Mahmoodi, P. 1990. The Application of Viscoelastic Dampers to Seismically Resistant Structures. Fourth U.S. National Conference on Earthquake Engineering. Earthquake Engineering Research Institute. El Cerrito, California. Vol. 3.

Chang, K. C., Tsai, M. H., Chang, Y. H., and Lai, M. L. 1998. Temperature Rise Effects of Viscoelastic Damped Structures Under Strong Earthquake Ground Motions. Proceeding Sixth U.S. National Conference on Earthquake Engineering. Earthquake Engineering Research Institute. Oakland, California. Chang, K. C., Shen, K. L., Soong, T. T., and Lai, M. L. 1994. Seismic Retrofit of a Reinforced Concrete Frame with Added Viscoelastic Dampers. Fifth U.S. National Conference on Earthquake Engineering. Earthquake Engineering Research Institute. Oakland, California, Vol. III. Cherry, M. and Filliatrault, M. 1993. Seismic Response Control of Buildings Using Friction Dampers. Earthquake Spectra. Vol. 9. No. 3. August. CUREe Publication No. W-01. 1999. Preface. Proceedings of the Invitational Workshop on Seismic Testing, Analysis and Design of Woodframe Construction. Ed. F. Seible, A. Filiatrault, and C. Uang, Division of Structural Engineering. University of California, San Diego. March 5-6. Dinehart, D. W. 1998. The Dynamic Behavior of Wood Framed Shear Walls with Passive Energy Dissipation Devices. Ph.D. Dissertation. Department of Civil and Environmental Engineering. University of Delaware, Newark, Delaware. Dinehart, D. W., Shenton, H. W. III, and Elliott, T. E. 1999. The Dynamic Response of Wood Frame Shear Walls with Viscoelastic Dampers. Earthquake Spectra. 15(1) February. Dumoulin, C., Magonette, G., Taucer, F., Fuller, K. N. G., Goodchild, I. R., and Ahmadi, H. R. 1998. Viscoelastic Energy Dissipators for Earthquake Protection of Reinforced Concrete Buildings. Proceedings of the Eleventh European Conference of Earthquake Engineering. A. A. Balkema, Rotterdam. Holmes W. T. and Somers, P. 1995. Northridge Earthquake of January 17, 1994 Reconnaissance Report, Volume 2. Earthquake Spectra. Supplement C to Volume 11. Publication 95-03/2. Kasai, Kazuhiko, Munshi, Javeed, A., and Maison, B. F. 1993. Viscoelastic Dampers for Seismic Pounding Mitigation. Proceedings of Papers Presented at the Structures Congress 93. ASCE. New York. Keith, E. L. and Rose, J. D. 1994. Hurricane Andrew Structural Performance of Buildings in South Florida. Journal of Performance of Constructed Facilities. 8(3). Mahmoodi, P., Robertson, L. E., Yontar, M., Moy, C., and Feld, L. 1987. Performance of Viscoelastic Dampers in World Trade Center Towers. Proceedings of Structural Congress, 87. ASCE. New York. Nielson, E. J., Lai, M. L., Soong, T. T., and Kelly, J. M. 1994. Viscoelastic Damper Overview for Wind and Seismic Applications. Proceedings of the Third Conference on Tall Buildings in Seismic Regions. Los Angeles Tall Buildings Structural Design Council. Los Angeles. Tsai, C. S. 1994. Temperature Effects of Viscoelastic Dampers During Earthquakes. Journal of Structural Engineering. 120(2) February. Zhang, R. H. and Soong, T. T. 1992. Seismic Design of Viscoelastic Dampers for Structural Applications. Journal of Structural Engineering. 118(5) May.