Sequential Phased Displacement Cyclic Tests of Wood-frame Shear Walls with Various Openings and Base Restraint Configurations

Transcription

1 VPI&SU Report #TE Sequential Phased Displacement Cyclic Tests of Wood-frame Shear Walls with Various Openings and Base Restraint Configurations Virginia Polytechnic Institute and State University Department of Wood Science and Forests Products Brooks Forest Products Research Center Timber Engineering Center 1650 Ramble Road Blacksburg, Virginia Report No. TE By: J.D. Dolan Associate Professor of Wood Engineering C.P. Heine Research Assistant Prepared for: The NAHB Research Center, Inc. Upper Marlboro, MD With Sponsorship from: The National Association of Home Builders Washington, DC and The U.S. Department of Housing and Urban Development Office of Policy Development and Research Washington, DC. September 14, 1997

2 2 VPI&SU Report #TE INTRODUCTION Wood frame shear walls are a primary lateral force-resisting element in conventionally constructed wood frame structures. Traditional engineered shear wall design requires fully sheathed wall sections restrained against overturning. Their behavior is often considered analogous to a deep cantilever beam with the end framing members acting as "flanges" or "chords" to resist overturning moment forces and the panels acting as a "web" to resist shear. This simple analogy is generally considered appropriate and conservative for wind and seismic design. Overturning, shear restraint, and chord forces are easily calculated using principles of engineering mechanics. While shear resistance can be calculated, tabulated values for shear resistance with varying fastener schedules are usually used. Adoption of the empirical-based perforated shear wall method by the Standard Building Code (SBC) in 1994 provides an alternative for designing shear walls with openings. This new methodology requires mechanical tie-down devices at each end of the entire wall rather than at the end of each fully sheathed segment. Consequently this method reduces the number of tie-down devices required for shear walls with openings, compared to traditional engineered design, but requires two tie-downs where conventional construction used none. A significant number of monotonic tests of one-third scale models and shorter fullsized walls provide validation of the perforated shear wall method. Results of monotonic and cyclic tests of full-scale, forty-feet long plywood sheathed walls are presented in Dolan and Johnson (1996 a and b). A detailed description and discussion of the

3 VPI&SU Report #TE investigation used to validate the perforated shear wall method is presented in Johnson (1997). This study provides additional information about the performance of long, full-sized, shear walls with and without openings, and with and without overturning restraint tested under monotonic and reversed cyclic loads. To improve the accuracy of designs when no overturning restraints are provided, tests of shorter wall specimens with corner framing were also tested. Together with the results of Johnson (1997), the results of this investigation give information useful in determining at what wind and seismic force levels the different numbers of overturning restraints are required for a given wall configuration. This will provide information necessary to determine which regions of the United States need not require overturning restraint due to low lateral load levels and which regions should require overturning restraint due to high wind or seismic loads. Monotonic tests were performed to determine the performance under static or wind loading. Sequential Phased Displacement (SPD) tests were performed to establish conservative estimates of performance during a wind or seismic event. Results of SPD tests of walls sheathed with oriented strand board (OSB) are presented in this report and monotonic test results of OSB sheathed walls are reported in a companion report in Dolan and Heine (1997 a). A detailed description and discussion of the effect of overturning restraint on the performance of timber framed shear walls including the effects of corner framing is presented in Heine (1997).

4 4 VPI&SU Report #TE OBJECTIVE Results of an experimental study of the performance of shear walls with and without openings, and with and without overturning restraint are reported. The primary objective of this report was: to quantify the effects of overturning restraint on full-size wood frame shear walls with and without openings tested reversed cyclically. BACKGROUND Traditional design of exterior shear walls containing openings, for windows and doors, involves the use of multiple shear wall segments. Each is required to be fully sheathed and have overturning restraint supplied by structure weight and/or mechanical anchors. The design capacity of shear walls is assumed to equal the sum of the capacities for each shear wall segment. Sheathing above and below openings is typically not considered to contribute to the overall performance of walls. An alternate empirical-based approach to the design of shear walls with openings is the perforated shear wall method which appears in the Standard Building Code 1994 Revised Edition (SBC, 1994) and the Wood Frame Construction Manual for One- and Two- Family Dwellings High Wind Edition (WFCM, 1995). The perforated shear wall method consists of a combination of prescriptive provisions and empirical adjustments to design values in shear wall selection tables for the design of shear walls containing openings. The empirical equation developed by Sugiyama (1981) forms the basis of adjustment factors in Table in the SBC (1994) and Table 3B in the

5 VPI&SU Report #TE WFCM (1995). Tabulated adjustment factors are used to reduce the strength of a traditional fully sheathed engineered shear wall segment for the presence of openings. When designing for a given load, shear walls resulting from this method will generally have a reduced number of overturning restraints than a similar shear wall constructed with multiple engineered shear wall segments. Under low to moderate wind and seismic conditions overturning restraint may not be required at all. To the authors knowledge, no investigations have completed to quantify the capacities of monotonically and reversed cyclically loaded full-scale shear walls without overturning restraint, with perforated shear wall method restraint, and fully restrained according to engineered design. In order to make these comparisons, this study utilizes additional data obtained by Johnson (1997) on full-scale plywood sheathed shear walls. So that comparisons can be made the same materials and three of the same wall configurations were used. One change was made, instead of plywood, oriented strand board (OSB) sheathing was used. OSB equivalency to plywood sheathed walls is given in all three building codes and has been demonstrated by Foschi (1980) and Dolan (1989). Both, plywood and OSB, are classified as structural wood sheathing. Therefore, direct comparison of OS and plywood sheathed walls was performed in this study. The Structural Engineers Association of Southern California (SEAOSC, 1997) proposed a standard reversed cyclic testing procedure. The quasi-static test method incorporates procedures included in the Sequential Phased Displacement Procedure proposed by M.L. Porter (1987). The proposed procedure was developed by the Technical Coordinating Committee on Masonry Research (TCCMAR), a joint project between USA and Japan, and revised by SEAOSC in As with the monotonic tests,

6 6 VPI&SU Report #TE load is applied to the top of the wall. The displacement pattern is sinusoidal, fully reversing with increasing amplitudes over time. TEST PROGRAM Six walls, forty feet (12m) in length and eight feet (2.4m) in height were tested using monotonic and SPD patterns. Together with the results obtained by Dolan and Johnson (1996 a and b) a total of three different wall configurations, three anchorage, and two loading conditions were included (Table 1). Results of monotonic tests are reported in Dolan and Heine (1997 a). Configurations A, D, and E were selected to cover the range of sheathing area ratios, r, typically encountered in light-frame wood construction. Each wall used the same type of framing, sheathing, nails, and nailing patterns. Table 1: Specimen configurations No Tie-down Anchors Wall Configuration Anchors at End of Wall Only Maximum Amount of Tie-down Anchors Wall Type A Testing Procedure Monotonic and SPD D Monotonic and SPD E Monotonic and SPD The shear capacity ratio, or the ratio of the strength (or stiffness) of a shear wall segment with openings to the strength (or stiffness) of a fully sheathed shear wall segment without openings, is determined by Equation (1):

7 VPI&SU Report #TE r F = 3 2 r where F is the shear capacity ratio and r is the sheathing area ratio. (1) Sheathing area ratio parameters are illustrated in Figure 1, and the ratio can be calculated by the following expressions: r = 1 1 = ( α + A 1 0 β ) ( 1 H ) + L i (2) α = A 0 H L (3) Li β = L (4) where: α = opening area ratio, β = wall length ratio, ΣA 0 = area of all openings, ΣL i = sum of the length of full height sheathing, L = shear wall length, and H = wall height.

8 8 VPI&SU Report #TE Figure 1: Sheathing area ratio variables The shear capacity ratio F is multiplied by the capacity of a fully-sheathed wall of the same length to obtain the capacity of the wall with openings. Table 2 lists the opening dimensions and illustrates the opening locations for each wall configuration included in this investigation. Table 2: Opening sizes for wall configurations Wall Sheathing Area Opening Size Configuration Ratio (r) Door Window (1) '-8" x 4'-0" 6'-8" x 12'-0" (Sheathed at ends) (2) 8'-0" x 28'-0" 4'-0" x 7'-10 1 / 2 " - (1): The top of the window is located 16 inches from the top of the wall. (2): Wall E has studs 16 in. o.c. for the full length of wall but is sheathed only at the ends of the wall. Materials and Fabrication Details Table 3 summarizes materials and construction details used for the wall specimens. Included are the sizes of headers and jack studs used around openings. Wall framing

9 VPI&SU Report #TE consisted of double top plates, single bottom plates, double end studs, and double or triple studs around doors and windows. Studs were spaced 16 in. on center. All framing consisted of spruce-pine-fir, Stud or better grade lumber, purchased from a local lumberyard. Members were arbitrarily chosen when placed in the wall specimens. Exterior sheathing was 7/16 in., OSB rated sheathing. All full height panels were 4 ft. by 8 ft. and oriented vertically. To accommodate openings, the OSB was cut to fit above and below the doors and windows without overlapping the header to king stud connection. Interior sheathing was 4 ft. by 8 ft. sheets of 1/2 in. gypsum wallboard, oriented vertically. As with OSB, the gypsum was cut to fit above and below the doors and windows. All joints in the interior sheathing were taped and covered with drywall compound. Compound drying times complied with manufacturer s recommendation and were adjusted to ambient temperature and humidity. Both exterior and interior sheathing were able to rotate past the test fixture at the top and bottom (i.e. the steel test fixture was narrower than the wood framing used for top and bottom plates.) Three specimens were tested reversed cyclically without overturning restraint. Simpson Tie-down HTT22 with 5/8 in. diameter anchor bolts were used on an additional three walls with one tie-down at the end of each fully sheathed segment. Tie-down anchors were attached to the bottom of the end studs of the respective segment by thirtytwo (32) 16d (0.148 in. diameter and 3.25 in. length) sinker nails. A 5/8 in. diameter bolt connected the tie-down, through the bottom plate, to the rigid structural steel tube test fixture.

10 10 VPI&SU Report #TE Table 3: Wall materials and construction data Component Fabrication and Materials Framing Members Sheathing: Exterior Interior Headers: 4'-0 opening Stud, Spruce-Pine-Fir, 2 x 4 inch 16 in. o.c. Structural Oriented Strand Board, 7/16 in., 4 ft. x 8 ft. sheets installed vertically. Gypsum wallboard, 1/2 in., installed vertically, joints taped (2) 2 x 4 inch (nom.) with intermediate layer of 7/16 in. OSB. One jack stud at each end. 12 opening (2) 2 x 12 inch (nom.)with intermediate layer of 7/16 in. OSB. Two jack studs at each end. Tie-down Anchor Bolts Simpson HTT 22, nailed to end studs with 16d sinker nails. 5/8 in. diameter A307 bolt to connect to foundation. 5/8 in. diameter A307 bolt with 3 in. square x 1/4 in. steel plate washers. Table 4 shows the four different types of nails and different nail schedules used in constructing the wall specimens. All framing connections used 16d (0.162 in. diameter and 3.5 in. length) brite common nails. Brite common 8d (0.131 in. diameter and 2.5 in. length) nails attached the OSB sheathing to the framing, 16d (0.148 in. diameter and 3.25 in. length) sinker nails attached tie-down anchors to the end studs, and 13 gage x 1-1/2 in. drywall nails attached gypsum wallboard to the framing. A nail spacing of 6 in. around the perimeter and 12 in. for intermediate framing was used for the OSB sheathing and 7 in. perimeter and 10 in. field for the gypsum wallboard. The bottom plate was attached to the steel structural tube test fixture using A307 or SEA grade 2. All bolts were 5/8 in. diameter National Coarse thread.

11 VPI&SU Report #TE Table 4: Fastener schedule Connection Description No. and Type of Connector Connector Spacing Framing Top Plate to Top Plate (Face-nailed) 16d common per foot Top / Bottom Plate to Stud (End-nailed) 2-16d common per stud Stud to Stud (Face-nailed) 2-16d common 24 in. o.c. Stud to Header (Toe-nailed) 2-16d common per stud Stud to Sill (End-nailed) 2-16d common per stud Header to Header (Face-nailed) 16d common 16 in. o.c. along edges Tie-down Anchor/ Anchor Bolts Tie-down Anchor to Stud (Face-nailed) Tie-down Anchor to Foundation Anchor bolts Sheathing: 32-16d sinker 1-A307 5/8 in. dia. Bolt or: strain-sert load bolt 1-A307 5/8 in. dia. bolt per tie-down per tie-down 24 in. o.c. and within 1 ft. of wall ends OSB 8d 6 in. edge / 12 in. field (2 rows for end stud) Gypsum wall board 13 ga x 1½ in. (3/8 in. head) 7 in. edge / 10 in. field Wall Attachment to Test Frame and Instrumentation Tests were performed with the shear walls in a horizontal position as shown in Figure 2. The walls were raised approximately 16 inches above the ground to allow instruments and load cell sufficient clearance to be attached to the wall. The bottom plate was secured to a fixed steel structural tube with 5/8 inch diameter bolts 24 inches on center, beginning one foot from the end of the wall. Steel plate washers, 3 inch x 3 inch square, 1/4 inch thick supported the bolts. Oversize of bolt holes was limited to 1/32 inch to minimize slip. With the exception of tie-down anchor bolts, all bolts were located a minimum of one foot from the edges of each sheathing panel and from the studs adjacent to openings.

12 12 VPI&SU Report #TE Figure 2: Wall orientation and test set-up A hydraulic actuator, with a range of ±6 inches and capacity of 55,000 lbs., displaced the top left corner of each shear wall (Figure 2). A steel tube was used to distribute the loading to the wall s double top plate. The steel tube and the double top

13 VPI&SU Report #TE plate were attached using 5/8 in. diameter bolts 24 in. on center, beginning one foot from the end of the wall. Eight casters allowed horizontal motion, as shown in Figure 2. The casters were fixed parallel to loading and did not rotate during testing. A test was conducted to determine the amount of friction created by the wheels. It was found that the magnitude of the friction was negligible (0.5-2%) when compared to the capacity of the walls. Nevertheless, the recorded load resisted by each wall was corrected to account for friction. Figure 2 depicts the location of three linear variable differential transducers (LVDT) that were attached to the foundation to measure wall displacements. LVDT #0 was built in the hydraulic actuator, measuring the relative displacement of the actuator. In order to obtain the actual displacement of the top of the wall the data of LVDT #0 were corrected using similar triangles to account for the depth of the steel load distribution beam. The actuator also contained a sensor recording the load resisted by the wall. LVDT #1 and LVDT #2 measured the compression and uplift displacement of the end studs relative to the foundation. These sensors determined the amount of crushing in the sill plate, or uplift of the end stud, depending on which corner of the wall was in compression or tension, respectively. For walls with tie-down anchors the data recorded were corrected to compensate for amplifications caused by the geometry of the LVDT fixtures. Anchored studs were assumed to rotate about the bolt attaching the anchor to the foundation. This ensured that actual compression and uplift displacements of the end studs were measured. However, this assumption could not be made for the walls without overturning restraint because the end stud did not have a defined point of rotation. In this case the data recorded by LVDT #1 and 2 were not corrected.

14 14 VPI&SU Report #TE LVDT #3 measured horizontal displacement of the bottom plate relative to the foundation. This measurement allows rigid body translation of the wall to be subtracted from the global displacement to obtain interstory drift. Interstory drift was calculated by subtracting the readings obtained from LVDT #3 from the obtained LVDT #0 data. An instrumented tension bolt replacing a normal bolt that attached the bottom plate to the foundation, was inserted close to the end stud were the load was applied (Figure 2). When tie-down anchors were applied, the tension bolt replaced the bolt that fastened the anchor and bottom plate to the foundation. Dolan and Johnson (1996 a) concluded that anchor slip between tie-down and stud was negligible. Consequently, it was not measured in this study. The data acquisition system recorded the data 15 times per second. Loading Sequential phased displacement (SPD) loading consisted of two displacement patterns and is illustrated in Figure 3 and 4. The first pattern gradually displaced the wall to its anticipated yield displacement. Elastic behavior was observed in this section of the test. The second displacement pattern began once the wall had passed its anticipated yield displacement (i.e. inelastic behavior) or first major event (FME). A FME of 0.1 inches was used in these tests and was determined from monotonic test results. The first displacement pattern consisted of reversed-cyclic displacements for three cycles at each incremental level at low, elastic behavior displacements. The first set of three cycles displaced the wall at approximately 25 percent of the FME. The second set of three cycles displaced the wall 50 percent of the FME, and the final set of three cycles displaced the wall at 75 percent of the FME. The next cycle displaced the wall to

15 VPI&SU Report #TE approximately the FME. At this point, the wall started to behave in an inelastic manner and the second displacement pattern began X% of yield displacement FME Time Figure 3: Displacement pattern used in sequential phased displacement

16 16 VPI&SU Report #TE phase X% of yield displacement A B C D Time (sec) Figure 4: Single phase of sequential phased displacement pattern Figure 4 graphically illustrates one phase of the second displacement pattern in SPD loading. Once yielding occurs, the displacement of each set of cycles is based on the previous set of cycles. Peak displacement of a set of cycles is increased 200 percent of the FME displacement over the previous set of cycles. The first peak cycle of a set is followed by three decay cycles, with each magnitude being 25 percent less than the previous cycle (i.e. the first decay cycle is 75 percent of the peak displacement, second is 50 percent, and third is 25 percent). Following the decay cycles are three cycles at the peak displacement. Three cycles were previously determined to be sufficient in order to obtain a "stabilized" response for nailed connections and shear walls. "Stabilized" response is reached when the load resisted by the wall does not decrease more than 5 percent when displaced the same magnitude in two consecutive cycles.

17 VPI&SU Report #TE PROPERTY DEFINITIONS Initial and stabilized load envelope (or backbone) curves, similar to the ones shown in Figure 5, were determined for each wall. A typical initial load envelope curve consists of positive and negative peak loads and corresponding displacements of the first cycle (corresponding to points A and B in Figure 4) for each phase of SPD loading. Positive and negative peak loads and corresponding displacements of the last cycle for each phase (corresponding to points C and D in Figure 5) form a stabilized envelope curve. The highest average of the absolute values of peak positive and negative load occurring in the first cycle of each phase determined the initial capacity, F max,init. Likewise, the average value of the corresponding interstory drifts determined the displacement at initial capacity, denoted as max,init. Stabilized capacity, F max,stab, is the highest average load resistance occurring in the last cycle. The corresponding average interstory drift was denoted as max,stab. Fmax, init. Load (lbs) max, init. Initial envelope curve Stabilized envelope curve Interstory drift (in) Figure 5: Typical initial and stabilized load envelope curves

18 18 VPI&SU Report #TE Failure of the walls was defined as a significant drop in load resistance or 0.8 F max depending on what occurred first. Elastic stiffness, k e, was defined as the slope of the line passing through the origin and the point on each load envelope curve where the load equals 40 percent of F max. This stiffness represents a good estimate of the stiffness that shear walls will exhibit after being loaded a number of times at low to moderate magnitudes. An equivalent energy elastic-plastic curve, used for comparison purposes, was determined for each wall. This artificial curve, shown in Figure 6, depicts how an ideal perfectly elastic-plastic wall would perform and dissipate an equivalent amount of energy until failure. Therefore, the area under the equivalent elastic-plastic curve (EEPC) equals the area under the load-displacement envelope curve from zero drift to drift at failure, failure. The elastic portion of the EEPC contains the origin and has a slope equal to the elastic stiffness, k e. The plastic portion of the EEPC is a horizontal line equal to F yield. Integrating the obtained load displacement curve for each wall, F yield was determined according to: F yield = ± failure 1 ke 2 2 failure A ke (5) where A is the area under the respective true load-displacement curve. Displacement at yield, yield, was defined as the displacement at the intersection of the elastic and plastic lines of the EEPC. The ratio of failure and yield provides the ductility D for each wall:

19 VPI&SU Report #TE failure D = yield (6) This definition of EEPC was also used in the monotonic tests, and is similar to that proposed in the sequential phased displacement test developed for ASTM, ISO. The Structural Engineers Association of Southern California (SEAOSC) uses an alternate variable to quantify this behavior. Unit shear was determined using the equation: g F L = max (7) where g is the unit shear, F max stands for wall capacity, and L is the sum of the widths of full-height sheathing panels. Consequently, the theoretical uplift is obtained by multiplying the unit shear g with the normal distance between applied load and foundation, which equals the wall height in this investigation. Fmax Load-displacement envelope curve Equivalent elastic-plastic curve Fyield Load (lb) F 0.8 F yield m a x 0.4 Fmax yield max Interstory drift (in) failure Figure 6: Typical equivalent energy, elastic-plastic curve

20 20 VPI&SU Report #TE RESULTS AND DISCUSSION Strength Load resistance at capacity, and at interstory drifts of 0.32 in., 0.96 in., and 1.6 in was determined from the initial and stabilized load envelope curves and is presented in Table 5. Peak load of the initial and stabilized cycles is correlated with the amount of tiedown anchors. For all wall configurations, capacity increased with increasing overturning restraint. Similar trends were observed for the load resistance at all drift limits. The monotonic capacity of Wall E more than doubled when the number of tie-down anchors was raised from zero to maximum. Table 6 illustrates this trend by showing the relative capacities based on the values of the configuration with maximum overturning restraint as is used in fully-engineered construction. For walls tested monotonically the difference is more pronounced due to the different failure mode between monotonically and SPD tested walls. Furthermore, the increase in capacities with the number of tie-down anchors is also affected by the size of openings. Wall E with the largest opening shows the greatest increase while the capacity of Wall A increased the least amount (Table 5 and Table 6). The same trend is apparent at all drift limits. It should be noted that these test observations are relative to the nature of the test method and conditions. When corner framing is considered as in Dolan and Heine (1997 b), the strength and stiffness of conventionally-built (unrestrained) walls would be enhanced. Consideration of gravity load effects would also provide expected improvements in tested or predicted performance. Wall A was tested with only two anchorage configurations, due to being fully sheathed. In other words, Wall A was tested twice with tie-down anchors at the end. Therefore, differences in capacities between the two tests can not be explained by the

21 VPI&SU Report #TE amount of tie-down anchors. However, the differences could be a result of statistical variation between the two specimens. Capacity Table 5: Initial cyclic and stabilized cyclic load resistance data no tie-down anchors (1) Wall Specimens anchors at end of wall only (2) maximum amount of tie-down anchors (1) A D E A (3) D E A (3) D E Monotonic (4) Initial SPD (kips) Stabilized SPD (kips) Initial/Monotonic Stabilized/Monotonic Stabilized/Initial Load 1.6 in. interstory drift Monotonic (4) Initial SPD (kips) Stabilized SPD (kips) Load 0.96 in. interstory drift Monotonic (4) Initial SPD (kips) Stabilized SPD (kips) Load 0.32 in. interstory drift Monotonic (4) Initial SPD (kips) Stabilized SPD (kips) (1) These specimens had OSB sheathing (2) These specimens had plywood sheathing (Dolan and Johnson 1996 a and b) (3) Wall A has the same anchorage requirements for the anchors at the end of wall only and maximum amount of tie-down anchors due to being fully sheathed. (4) From Dolan and Heine (1997 a) The ratio of stabilized to initial peak load is almost constant for all wall configurations. According to that ratio, the reduction of ultimate resistance between initial and stabilized SPD cycles is, on average, 14 percent (Std Dev = 1%). However, the

22 22 VPI&SU Report #TE ratio of reduction between monotonic and initial SPD cycles ranges from 23 percent (Wall E, maximum amount of anchors) to 6 percent (wall D perforated restraint). For walls without overturning restraint the ratio shows an increase from monotonic to initial SPD capacity between 3 and 8 percent. For all wall configurations the ratio between monotonic and stabilized SPD peak loads shows a reduction ranging from 32 percent (Wall A, maximum amount of anchors) to 7 percent (Wall E, no anchors). Table 6: Relative capacities based on conventional construction (no tie-down anchors) Relative Capacity (%) no tie-down anchors (1) Wall Specimens anchors at end of wall only (2) maximum amount of tie-down anchors (1) A D E A (3) D E A (3) D E Monotonic (4) Initial SPD (kips) Stabilized SPD (kips) (1) These specimens had OSB sheathing (2) These specimens had plywood sheathing (Dolan and Johnson 1996 a and b) (3) Wall A has the same anchorage requirements for the anchors at the end of wall only and maximum amount of tie-down anchors due to being fully sheathed. (4) From Dolan and Heine (1997 a)

23 VPI&SU Report #TE Figure 7: Shear strength ratios at capacity (Eq. 1) as a function of sheathing area ratio Shear strength ratio, F shear strength ratio no anchors anchors at end of wall max amount of anchors Sheathing area ratio, r Figure 7 illustrates the conservatism of the predicted shear strength ratio by plotting the actual shear strength ratios and the theoretical shear strength ratio according to Equation (1). It should be noted that the walls with different overturning restraint conditions use different normalizing values for the shear strength ratio, F. While it is possible to compute F for this study due to the fully-sheathed condition being tested for all restraint conditions, the fully-sheathed capacity for unrestrained walls is not typically available to the design professional. Therefore, the application of the perferated shear wall method is not typically available for this condition. However, as shown in Error! Reference source not found., the actual shear capacity ratios are higher than the predicted ratios in all cases. This is the same behavior found by Dolan and Johnson (1996 a).

24 24 VPI&SU Report #TE Table 7 lists the three parameters, F yield, elastic stiffness, and ductility determined from the EEPC. The average reduction of F yield between initial and stabilized cycles is 14 percent (Std Dev = 3.1%) for all wall configurations. Together with the capacity reduction this implies that the damage experienced by the walls is fairly uniform regardless of the amount of overturning anchorage present. In other words, the damage performance is more governed by the sheathing nails than it is by the tie-down anchors. Table 7: Equivalent elastic-plastic curve parameters no tie-down anchors (1) Wall Specimens anchors at end of wall only (2) maximum amount of tie-down anchors (1) A D E A (3) D E A (3) D E F yield Monotonic (4) Initial SPD (kips/in) Stabilized SPD (kips/in) Stabilized/ Initial Elastic Stiffness Monotonic (4) Initial SPD (kips/in) Stabilized SPD (kips/in) Stabilized/ Initial Ductility: Monotonic (4) Initial SPD ductility Stabilized SPD ductility Stabilized/ Initial ductility (1) These specimens had OSB sheathing (2) These specimens had plywood sheathing (Dolan and Johnson 1996 a and b) (3) Wall A has the same anchorage requirements for the anchors at the end of wall only and maximum amount of tie-down anchors due to being fully sheathed. (4) From Dolan and Heine (1997 a) As expected, F yield increased as the number of overturning restraints was increased from zero to maximum. The values from monotonic tests show the greatest increase. Data from the initial curve can be used to set design values for shear walls subjected to a one time peak load such as wind loading. Data from the stabilized curve

25 VPI&SU Report #TE can be used to set conservative design values for shear walls subjected to repetitive cycling such as a seismic event. The objectives of design govern which curve, initial or stabilized, should be used. However, an adjustment in the system response factor, R, that is used in calculating base shear, may be needed to compensate or calibrate the design process to expected response. Stiffness The average value for elastic stiffness calculated from the secant slope passing through the origin and a point on the positive and negative load-drift envelope curve where the resisted load equals 40 percent of capacity, is presented in Table 7. Stabilized and initial elastic stiffness values for a given configuration differ a maximum of 4 percent, which is not significant for most practical applications. However, stiffness values vary significantly between the monotonic and cyclic testing procedures. With no tie-downs applied, the elastic stiffness was on average 17 percent lower (Std Dev = 4.6%) for walls tested monotonically as opposed to the stiffness obtained from SPD testing. As soon as overturning restraint was applied, the stiffness values for monotonically tested walls were not consistently lower anymore (Table 7). This change in performance is in part attributed to the fact that energy was absorbed through increased racking and cold-working of the sheathing nails during the cyclic test when anchors were applied, whereas all the energy was absorbed through racking of the sheathing nails for all monotonic tests. When anchors were omitted, however, the internal damage experienced by the walls was mainly withdrawal from the bottom plate. In other words, without anchors the damage performance was more alike during the two different testing procedures.

26 26 VPI&SU Report #TE Table 7 further shows that stiffness values also increase with an increasing amount of overturning restraint for a given wall, but the increase is not as pronounced. This condition is unique to the small amplitude of displacements used in the elastic stiffness region, resulting in minor damage to the sheathing nails. In all three wall groups, the highest elastic stiffness values were determined for Walls A, fully sheathed. As with load resistance and capacity, stiffness values increased in magnitude as area of openings decreased. Stiffness values for Wall E reached only an average of 20 percent (Std Dev = 4.3%) of the stiffness of Wall A, regardless of the number of tie-downs. Ductility Equation (6) was used to determine the ductility ratio from the initial and stabilized equivalent elastic-plastic curves. The ductility values reported in Table 7, as well as the ratio of stabilized to initial ductility, show no apparent trend. Initial ductility ranged from 2.7 to 5.7, while stabilized ductility ranged from 3.7 to 7.3. Monotonic ductility generally appears to be higher than initial ductility. Vertical End Stud Displacement Table 8 shows the peak displacements of the end studs measured during the tests. The vertical movement (in y direction, Figure 2) of the end studs relative to the bottom plate was measured. The distance traveled by the end studs between peak positive and negative interstory drifts (i.e. between point A to B in Figure 4), recorded during the initial cycle at max load and failure, are given in Table 8. At this point, it should be mentioned that no pattern of movement for the end studs of walls without overturning restraint could be assumed. When comparing the displacements depicted in Table 8, recall that the values for walls without restraint were not corrected regarding geometric measurement

27 VPI&SU Report #TE amplification effects. However, the vertical stud movement for walls without tie-down anchors is an order of magnitude higher than for walls with anchors. This illustrates why walls tested without anchors tend to fail along the line of sheathing nails at the bottom of the wall in an unzipping manner, while walls with anchors show a distribution of the load to all sheathing nails indicating a higher degree of load sharing. However, if corner framing effects are considered stud uplift is significantly reduced for walls without tiedown anchors. Dolan and Heine (1997 b) report the hold-down effects in a companion report. Table 8: End stud displacement between positive and negative peak drifts during initial cycle of max load and failure no tie-down anchors (1) Wall Configuration anchors at end of wall only (2) maximum amount of tie-down anchors (1) A D E A (3) D E A (3) D E Left end stud (LVDT max load Right end stud (LVDT max load Left end stud failure _4) Right end stud (LVDT failure _5) (1) These specimens had OSB sheathing (2) These specimens had plywood sheathing (Dolan and Johnson 1996 a and b) (3) Wall A has the same anchorage requirements for the anchors at the end of wall only and maximum amount of tie-down anchors due to being fully sheathed. (4) Sensors did not record data (5) Data not available due to failure mode For all walls, vertical stud displacements increase beyond capacity. This is partly attributed to the fact that the stud rotation about the anchor bolt increases with increasing drift of the wall. In case of no overturning restraints, the applied force causes the wall to separate along the bottom plate and stud uplift increases until catastrophic failure. It

28 28 VPI&SU Report #TE should be noted that this failure mode could improve for unrestrained walls if corner framing is applied (Dolan and Heine, 1997 b). Since stud uplift of walls with tie-down anchors is relatively small it is doubtful that axial loading (i.e. dead load) would increase shear load capacity for walls with tiedown anchors of the kind used in this investigation. Conversely significant increases could be expected for the unrestrained walls. Tie-Down Tension Bolts Tension load bolts measured the uplift resisted by the tie-down anchors during SPD loading. Correlation of racking load resisted by the shear wall to load resisted by the bolt for each cycle in the test was determined and the results are shown in Table 9. For comparison, equivalent data of monotonic test results are listed in Table 10. The theoretical uplift force was determined by assuming that the shear load was uniformly distributed along the wall, and therefore a unit shear could be calculated. However, it should be noted that this is an apparent value since the shear load is distributed along the wall according to the relative stiffness of the wall segments, with the end wall segment resisting the highest percentage of the lateral force.

29 VPI&SU Report #TE Table 9: Maximum force resisted by tension bolt SPD loading no tie-down anchors (1) Wall Configuration anchors at end of wall only (2) maximum amount of tie-down anchors (1) A D E A (3) D E A (3) D E F max, tensionbolt (lbs.) N/A (4) N/A N/A F F max, tensionbolt (kips) N/A N/A N/A Initial capacity (kips) Interstory drift F max, tensionbolt (lbs.) N/A N/A N/A Unit shear (lbs./ft) F max, wall Theoretical uplift (lbs.) actual / theoretical (1) These specimens had OSB sheathing (2) These specimens had plywood sheathing (Dolan and Johnson 1996 a and b) (3) Wall A has the same anchorage requirements for the anchors at the end of wall only and maximum amount of tie-down anchors due to being fully sheathed. (4) Johnson (1997) did not measure uplift forces for these configurations (5) Unit shear calculated here is an apparent value since the lateral force is distributed to the wall segments in proportion to their stiffness. The ratio of actual to theoretical uplift exceeds unity for Wall A and E tested cyclically and for Wall D and E tested monotonically. The significant smaller peak loads recorded by the bolts in walls without overturning restraint elucidate how little uplift force is transferred to the foundation in a concentrated manner if tie-down anchors are omitted. The lower force experienced by the instrumented bolts were due to the inability of the sheathing nails to transfer the overturning loads. The associated damage to the sheathing nails resulted in significantly reduced capacities. This indicates possible advantage for using lighter but more disperse hold-down devices if lower capacities are adequate. The tension bolts experienced their maximum load at wall displacements slightly lower than where capacity was reached in walls with no overturning restraint. With tie-down anchors

30 30 VPI&SU Report #TE applied, maximum load in the tension bolts was reached at displacements slightly higher than where the wall capacity was reached. Table 10: Maximum force resisted by tension bolt, monotonic loading (1) Wall Configuration no tie-down anchors (2) Anchors at end of wall only (3) maximum amount of tie-down anchors (2) A D E A (4) D E A (4) D E F max, tens.bolt (lbs.) N/A 5) N/A N/A F F max, tensionbolt (kips) N/A N/A N/A Capacity (kips) Interstory drift F max, tens.bolt (lbs.) N/A N/A N/A Unit shear Theoretical uplift (lbs.) actual / theoretical (1) From Dolan and Heine (1997 a) (2) These specimens had OSB sheathing (3) These specimens had plywood sheathing (Dolan and Johnson 1996 a and b) (4) Wall A has the same anchorage requirements for the anchors at the end of wall only and maximum amount of tie-down anchors due to being fully sheathed. (5) Johnson (1997) did not measure uplift forces (6) Data not available due to damaged leads to tension bolt General Observations These tests were performed without an applied dead load in order to test the most conservative condition. If dead load had been present, the studs next to the openings that had no overturning restraint (i.e., no tie-down connectors) may not have lifted from the test frame as much. This may have reduced the damage to the nails attaching the sheathing to the bottom plate in these regions. The result may have been an improved overall performance. This is especially clear when one considers that studs next to openings have the highest compressive load due to applied dead load. In addition, corner framing effects, when considered, also improve the performance of unrestrained walls (Dolan and Heine, 1997 b).

31 VPI&SU Report #TE Gypsum panels were observed to perform poorly during the cyclic tests. Even at low displacement magnitudes, drywall nails tore a path in the gypsum panels. Taped joints failed at lower interstory drifts than during monotonic tests. This matches experience in large magnitude seismic events. Walls with no overturning restraint separated almost completely from the bottom plate at large displacements, which was the typical failure mode. Panels above and below openings more or less rotated as rigid bodies. Nail fatigue occurred less in walls with no tie-down anchors than in walls with maximum amount of tie-down anchors because the racking of the sheathing relative to the framing was less distinct. However, when corner framing effects are included in the testing the failure mode for unrestrained walls becomes more comparable to the restrained wall tests (Dolan and Heine 1997 b). The predominant mode of failure for walls with maximum overturning restraint was nail fatigue between framing and OSB sheathing at larger displacements and nail tear through at the top and bottom of sheathing panels after peak load was reached. Close to capacity, OSB panel edges between two adjacent panels started to interfere and fail in bearing. At larger displacements, much greater than at capacity, studs and sheathing started to separate from the top plate, especially at the end of the wall away from the applied load. Nails attaching OSB sheathing to the framing partially withdrew on the perimeter of the panels near corners, but failed predominantly due to fatigue. Bolts attaching the bottom plate were located a minimum of 12 inches away from the studs adjacent to openings. This resulted in the bottom plate lifting when the stud next to an opening was in tension and did not have a tie-down anchor to resist the force. In turn, the nails attaching the sheathing to the bottom plate had to transfer this tension and

32 32 VPI&SU Report #TE were damaged significantly more than nails near tie-down anchors. When corner framing restraint is considered, the stud uplift in unrestrained walls is reduced (Dolan and Heine, 1997 b) Monotonic vs. SPD Performance Figures 8 through 10 reveal monotonic load vs. interstory drift curves, and initial and stabilized envelope curves for walls without overturning restraint. Values for stabilized and initial response in all graphs depicted are average values of the positive and negative curves for each wall. It is apparent from the three graphs that the monotonic testing procedure yields capacities in between the stabilized and initial response from the SPD test method for Walls A, D, and E without tie-down anchors As revealed in Figures 8 through 10 and in Table 7, elastic stiffness is lower when a wall with no tie-down anchors is tested monotonically as opposed to SPD testing. Wall A (no tie-down anchors) Load (kips) initial monotonic stabilized Interstory drift (in) Figure 8: Monotonic load vs. drift curve and initial and stabilized envelopes curves for Wall A (no tie down anchors, OSB sheathing)

33 VPI&SU Report #TE Wall D (no tie-down anchors) Load (kips) Interstory drift (in) Figure 9: Monotonic load vs. drift curve and initial and stabilized envelopes curves for Wall D (no tie down anchors, OSB sheathing) Wall E (no tie-down anchors) Load (kips) Interstory drift (in) Figure 10: Monotonic load vs. drift curve and initial and stabilized envelopes curves for Wall E (no tie down anchors, OSB sheathing)

34 34 VPI&SU Report #TE Load versus interstory drift curves and envelope curves for walls with overturning restraint at the ends only and maximum overturning restraint are presented in Figures 11 through 16. Monotonic capacities exceeded cyclic capacities by 19 percent, 5 percent, and 10 percent for Wall A, D, and E respectively, with anchors at the ends only and by 24 percent, 13 percent, and 17 percent for Wall A, D, and E respectively, with maximum amount of tie-down anchors. For all wall configurations, the discrepancy between monotonic and SPD capacity increased with increasing overturning restraint. This is due to the increased restraint producing a more uniform distribution of the load to the sheathing nails which increases the overall damage to the sheathing connections prior to reaching peak load during the SPD tests. Because the initial and stabilized curves actually represent envelopes of hysteretic loops (cycles), the walls have actually expended more energy at a given displacement relative to the monotonic tests.

35 VPI&SU Report #TE Wall A (anchors at end of wall) Load (kips) Interstory drift (in) Figure 11: Monotonic load vs. drift curve and initial and stabilized envelopes curves for Wall A (anchors at the end of wall only, plywood sheathing) Wall D (anchors at end of wall) Load (kips) Interstory drift (in) Figure 12: Monotonic load vs. drift curve and initial and stabilized envelopes curves for Wall D (anchors at the end of wall only, plywood sheathing)

36 36 VPI&SU Report #TE Wall E (anchors at end of wall) Load (kips) Interstory drift (in) Figure 13: Monotonic load vs. drift curve and initial and stabilized envelopes curves for Wall E (anchors at the end of wall only, plywood sheathing) Wall A (max. # of anchors) Load (kips) Interstory drift (in) Figure 14: Monotonic load vs. drift curve and initial and stabilized envelopes curves for Wall A (max. amount of tie down anchors, OSB sheathing)

37 VPI&SU Report #TE Wall D (max. # of anchors) Load (kips) Interstory drift (in) Figure 15: Monotonic load vs. drift curve and initial and stabilized envelopes curves for Wall D (max amount of tie down anchors, OSB sheathing) Wall E (max. # of anchors) Load (kips) Interstory drift (in) Figure 16: Monotonic load vs. drift curve and initial and stabilized envelopes curves for Wall E (max amount of tie down anchors, OSB sheathing)