DYNAMIC PERFORMANCE OF PERFORATED LIGHT-FRAME SHEAR WALLS WITH VARIOUS END RESTRAINTS.
J. DANIEL DOLAN [*]
Established engineered design of exterior shear walls containing window and door openings involves the use of multiple, fully sheathed shear wall segments typically restrained against overturning forces at both ends. An alternate empirical-based approach is the Perforated Shear Wall Method, which requires mechanical overturning restraints at each end of the entire wall rather than at the end of each fully sheathed shear wall segment. Yet, under low to moderate wind and seismic conditions, mechanical overturning restraint may not be required at all due to the effects of gravity loads. The Perforated Shear Wall Method has not been adopted for "conventional" construction, which does not use overturning restraints, due to lack of information on unrestrained, fully sheathed walls. Results of three investigations of monotonic and cyclic response of light-frame shear walls are presented. Twenty-six full-scale wall specimens were constructed employing methods typical for platform construction and tested using monoto nic and sequential phased displacement (SPD) patterns. Five different wall configurations with various door and window openings, three anchorage conditions, and two loading conditions were used on 40-foot-long walls. Furthermore, to improve the accuracy of design when no overturning restraints (i.e., hold-down brackets) are provided, the interaction of mutually perpendicular walls was investigated on four 3.7-m-long walls. Based on results obtained, transverse walls alter the failure mode and impact shear wall performance. The number of overturning restraints is positively correlated with shear wall capacity and elastic stiffness. Anchorage requirements are accurately quantified for shear wall design.
Established engineered design of exterior shear walls containing window and door openings involves the use of multiple, fully sheathed shear wall segments typically restrained against overturning forces at both ends. Design capacity is assumed to be the sum of the shear capacities of each fully sheathed segment. Contribution to shear strength of sheathed segments below and above openings is generally neglected. However, Tissel and Rose  noted that the sheathed area below or above large openings contributed significantly to the overall wall strength and stiffness.
In Japan, Yasumura and Sugiyama  studied the influence of openings on stiffness and capacity in shear walls without intermediate overturning restraints employing static, monotonic tests. They developed a design method that is based on a "shear strength ratio." The shear strength ratio indicates the capacity reduction of a wall containing openings when compared to a fully sheathed wall with the same dimensions. In 1994, Sugiyama and Matsumoto introduced an empirically derived equation to calculate shear strength reduction at a shear deformation angle of 1/100 radian . This equation has been proposed to predict the shear strength ratio at capacity and formed the basis of the so-called Perforated Shear Wall Method, which has been incorporated in two standards [4,15]. As opposed to traditional engineered design, the Perforated Shear Wall Method does not neglect the contribution of structural sheathing above and below openings and stipulates mechanical overturning restraints at the ends of the entire wall only.
Walls in conventionally constructed homes in the United States normally do not contain mechanical tie-down anchors. Questions have been raised whether homes in low to moderate wind and seismic regions need to be mechanically restrained. The higher the aspect ratio (height/length) of a given shear wall, the more it will act like a cantilevered beam, and overturning restraints will obviously improve the strength when loaded in shear. However, for walls with relatively small aspect ratios, it is hard to believe that the wall would behave under a lateral load as a cantilevered beam. Many engineers question if wood shear walls with low aspect ratios will experience as high an overturning force during earthquakes as other walls. Anchorage effect on shear walls with small aspect ratios has yet to be quantified.
Monotonic tests of shear walls date back to the 1940s. Current shear wall design values are based on monotonic tests conducted by the American Plywood Association . Foschi  reported on experimental and numerical investigations that he conducted to provide sufficient evidence for waferboard to be considered an equivalent to plywood as a structural sheathing material for racking resistance. Based on Foschi's research, composite materials such as waferboard and oriented strandboard (OSB) are used interchangeably with plywood as structural sheathing material in wood buildings.
In recent years, areas of increased interest have been dynamic and reversed cyclic loading. This is attributed to the fact that more catastrophic failures occur during dynamic loading associated with natural disasters than due to gravity loads. The Loma Prieta earthquake (San Francisco), in 1989, the Northridge earthquake, in 1994, hurricanes Andrew and Iniki, in 1992, and the most recent earthquakes in Turkey and Taiwan have increased the public's interest in improving the reliability of structures.
The purpose of this study is to quantify the effects of overturning restraints and to investigate transverse (or corner) walls on the performance of light-frame shear walls with various openings, as they are subjected to reversed cyclic and monotonic loading. Results provided are useful to refine current design methodologies for shear walls and examples are presented on how to incorporate the results in building codes. In order to expand the scope of this study to the fullest, the analysis includes data from a previous experimental investigation conducted by Johnson .
Properties essential for analysis can be determined from the load-drift curves. Drift describes the movement in loading direction of the wall top plate relative to its bottom plate. Unless stated otherwise, the property definitions used in this study were based on definitions put forth by the American Society for Testing and Materials (ASTM) .
For structures tested cyclically, the load-drift plot is a series of hysteresis loops. In order to compare walls tested monotonically and reversed cyclically, the load envelope or backbone curve was introduced. A load envelope curve is the locus of extremities of the hysteresis loops and resembles the shape of a load-drift curve obtained from monotonic tests. Due to the loading pattern applied, two types of load envelope curves were obtained. The initial cycle envelope curve contains the peak load from the first cycle of each phase of sequential phased displacement (SPD) loading, whereas the stabilized cycle envelope curve represents the peak loads from the last cycle of each phase of SPD loading.
Ultimate or initial capacity, [F.sub.peak(initial)], is the highest average of the absolute positive and negative peak load occurring in the first cycles of each phase of SPD loading. Stabilized capacity, [F.sub.peak(stabilized)], is defined accordingly for the stabilized cycles of each phase. Monotonic capacity is simply the highest load, [F.sub.peak], resisted by the wall during a monotonic test.
Sheathing area ratio classifies walls based on the area of openings and total length of structural full-height sheathing panels present. The ratio can be calculated by the following expression :
r = 1/1 +/[sigma][A.sub.i]/H[sigma][L.sub.i] 
where r designates the sheathing area ratio; [sigma][A.sub.i] area of all openings; [sigma][L.sub.i] = sum of the length of full-height sheathing; H = wall height.
Unit shear was determined using the following equation:
g = [F.sub.max]/[sigma][L.sub.i] 
where g = unit shear; [F.sub.max] = wall capacity. Consequently, the theoretical uplift is obtained by multiplying the unit shear (g) times the normal distance between applied load and foundation, which equals the wall height in this investigation.
MATERIALS AND METHODS
Twelve long full-scale walls without transverse walls were tested using a monotonic and an SPD pattern. Walls were 12.2 m long and 2.4 m high. Together with the results obtained by Johnson , a total of five different wall configurations and three anchorage conditions were included (Fig. 1). The long specimens were tested only once due to the high costs and time required, whereas the shorter walls with corner framing were tested twice for each configuration. Corner wall specimens were 3.7 m in length and 2.4 m in height. Attached to the ends of the walls were 1.2 m by 2.4 m and 0.6 m by 2.4 m wall segments, respectively, oriented perpendicular to the shear wall of interest.
Shaded areas in the figures included in Figure 1 represent the sheathing panels (OSB or plywood on one side and gypsum sheathing on the other side). The walls were constructed with the sheathing panels oriented vertically (i.e., the long panel dimension ran parallel to the studs). Wall configurations presented were selected to cover the range of opening sizes and locations typically encountered in light-frame construction (Table 1). Wall C was intended to represent walls where a minimum of area is covered with structural sheathing. The middle area of this wall type is commonly sheathed with non-structural sheathing such as insulation board.
Each wall was composed of the same type of framing and sheathing nails, and used the same nailing pattern (Fig. 2). Studs spaced 406 mm on center (o.c.), double top plates, single bottom plates, double end studs, and double or triple studs around doors and windows were the main components of the wall framing. All framing lumber was spruce-pine-fir and was graded Stud or Better. Lumber was stored a minimum of 2 weeks in a covered laboratory prior to assembly. Walls investigated by Johnson  were sheathed with plywood instead of OSB but were otherwise constructed equivalently to all of the other wall specimens. The size of all fullheight sheathing panels was 1.2 by 2.4 m. Exterior sheathing was either 11-mm OSB rated sheathing or 12-mm plywood sheathing. Interior sheathing consisted of 13-mm gypsum wallboard. To be compatible with common construction practices, a gap of 3 mm between exterior sheets was maintained to account for potential swelling of the sheets. Drywall joints were taped and covered with dry wall compound. Top plates of the corner framing were braced using 11mm-thick OSB sheets, fastened with 8d common nails (152mm o.c.) to simulate the assumed "lower bound" stiffening effect of a floor or ceiling diaphragm (Fig. 2). For the 2-foot-long corner walls, the OSB and gypsum wallboard sheathing panels were cut in half lengthwise. Additional details on specimen configuration and construction details are described in a previous study .
TEST SET-UP AND PROCEDURES
Walls were tested in a horizontal position and raised 400 mm above the ground to facilitate installation of the measuring equipment and load cell. Specimens were not loaded to account for gravity load effects, such as the weight of a roof or a second story. A hydraulic actuator, with a range of +/- 152 mm and capacity of 245 kN displaced the top corner of each shear wall. Figure 3 shows the load application and attachment to the test frame. A steel tube distributed the loading to the walls' double top plate. With the exception of tie-down anchor bolts, all bolts attaching top or bottom plates to the steel tubes were located a minimum of 1 foot from the edges of each sheathing panel and studs adjacent to openings. Walls restrained against overturning forces contained Simpson Tie-down HTT22 with [theta] 16-mm anchor bolts.
The cyclic loading protocol used was a modification of the "Sequential Phased Displacement Procedure" developed by the Technical Coordinating Committee on Masonry Research (TCCMAR). The procedure was first described by Porter  and later revised by Dolan , and Dolan and Johnson . The procedure was proposed for adoption by ASTM to form a standard testing procedure for shear walls under cyclic loading but has since been viewed as a very severe test protocol in terms of energy demand.
In general, the SPD protocol is displacement controlled and involves triangular loading cycles grouped in phases at incrementally increasing displacement levels. Each phase is associated with a respective displacement level and contains one initial, three decay, and three stabilization cycles (same displacement level as initial). The amplitude of each consecutive decay cycle decreases by 25 percent of the initial displacement. Stabilized response is an important characteristic to assess structural performance after high-wind and repetitive cyclic earthquake loading and is defined as a decrease in load between two successive cycles of not more than 5 percent. For nailed wood joints and nailed shear walls, it has been determined that three stabilization cycles are sufficient to obtain a stabilized response .
The incremental displacement increase between each consecutive phase is controlled by the experimentally determined displacement at the first major event (FME). The FME is defined as the displacement at which the structure starts to deform inelastically (anticipated yield displacement). For the walls tested in this study, an FME of 25 mm was determined. The displacement amplitude increase between each successive phase was twice the FME displacement. The cyclic frequency was held constant at 0.5 Hz and data were recorded between 15 and 35 times per second.
Monotonic (one-directional) loading displaced the walls at a constant rate of 15 mm per minute. Monotonic test specimens were not subjected to load increments as specified in the ASTM E72 procedure . Rather, the monotonic test was a continuous increase in displacement until failure occurred. During monotonic tests, the data-acquisition system recorded the data 10 times per second.
RESULTS AND DISCUSSION
EFFECT OF OVERTURNING RESTRAINTS
Only one specimen of each 12.2-inlong wall configuration was tested due to the high cost and effort associated with full-scale testing. As a result, no statistical information can be revealed. However, it is well established that response of light-frame shear walls is primarily influenced by the load-slip characteristics of the sheathing-to-framing nails (6,8,10,16). In light of the fact that a full-scale specimen as tested in this investigation contains many nails (600 to 1,500), the average response (or close to average response) of the wall can be expected and comparisons can be made.
Peak load of initial and stabilized cycles was positively correlated with the number of overturning restraints (Table 2). Walls with tie-down devices at the extreme ends had capacities in between conventional construction (no overturning restraints) and engineered construction (maximum restraint) due to lifting of the bottom plate at the ends of unrestrained fully sheathed segments. This lifting resulted in localized damage of the sheathing nails along the bottom of the wall, which failed before other sheathing nails were loaded significantly. The size of openings influenced the magnitude of capacity increase or decrease between anchorage conditions. Walls with a small percentage of openings had the smallest influence from anchorage. Based on engineered construction, initial and stabilized peak loads of Wall A (fully sheathed) decreased by less than 5 percent when overturning restraints were omitted. Yet, ultimate capacity of Wall C with the largest opening and no overturning restraints was 43 percent lower when compared to the same configuration with maximum restraint.
It is apparent from Table 2 that the monotonic testing procedure yields capacities in between the stabilized and initial response from the SPD test method for walls without overturning restraints. Initial capacities were between 3 and 8 percent higher than monotonic capacities. This is in part attributed to the wall specimens separating from the bottom plate at both ends during the SPD procedure. The walls crushed the bottom plate at one end and "unzipped" at the other end, and vice versa. Until nails were completely withdrawn from the studs or fatigued, and the wall was separated from the bottom plate, there was more cold working of the nails or energy dissipation involved during cyclic tests than during monotonic tests. These results refute widespread belief that cyclic testing yields lower capacities than a monotonic testing procedure. However, the difference is most likely due to the unrestrained wall specimens racking at the base during the early stages of the test, resulting in a reduction in cold work ing and associated early fatigue of the sheathing nails.
For all wall configurations, the difference between monotonic and SPD capacity (initial and stabilized) increased with increasing overturning restraint. The highest disparity was experienced by Wall C with maximum restraint. This is due to increased restraint producing a more uniform distribution of the load to the sheathing nails, which increased the overall damage to the sheathing connections prior to reaching peak load during the SPD tests. This illustrates an important point. The influence of overturning restraints diminishes under cyclic loading. While overturning restraints significantly increases the capacity of monotonically loaded walls, requiring them in wood construction based solely on monotonic test results may be an unnecessary expense for many wood buildings in the United States since most lateral loading is of a cyclic nature. (This inference is based on the results of tests using the SPD test protocol. Other protocols may yield different results.)
The ratio of stabilized to initial peak load is almost constant for all wall configurations. According to the ratio, reduction of ultimate resistance between initial and stabilized SPD cycles is, on average, 14 percent (standard deviation [SD] = 1%). This demonstrates that strength degradation of shear walls is a result of crushing of the timber around the nail shanks, cold working of the nail, and partial nail withdrawal. The reduction magnitude is independent of opening size and anchorage condition.
Due to being fully sheathed, wall configuration A was tested with only two anchorage conditions (see footnote of Table 2). Table 2 indicates that there is a difference in performance between the two restrained walls. Differences cannot be explained by the number of anchors, but may be a result of statistical variation brought about by different manufacturing and testing personnel (see footnote of Table 2), or a slight difference between plywood and OSB nailbearing strength.
EFFECTS OF TRANSVERSE WALLS
Specimens with 0.6-m corners attached show a fairly high variation relative to the mean shear capacity (Table 3). The sample size of two walls is not sufficient to make inferences whether the variation is due to random effects, but the variation may stem from the different failure mode that was observed. It is remarkable that except for the nails attaching the sheathing to the bottom plate, there were no typical damage signs of racking of the sheathing panels with corner framing. The taped joints between the drywall panels experienced no damage. The corner segments hindered the free rotation of the sheathing with respect to the framing. Basically, the walls rotated as rigid bodies and eventually separated from the bottom plate. This behavior may be a result of the relatively short wall length of 3.7 m. Nails at the bottom plate tore and pulled through the sheathing. Nails simply withdrew from the framing along the bottom of the wing walls. There was no nail fatigue observed in the corner specimens due to the minimal relative movement of the sheathing and framing, which is a significant change from the behavior observed on long and straight walls. The wing walls rotated about the bottom nail furthest away from the corner of the specimens.
The results of tests with transverse walls must be considered with the important limitation that the specimens were loaded parallel to the 3.7-m wall segment only. Wall segments perpendicular to the load were not directly loaded and essentially resisted uplift of the corner only. In a building loaded either by a hurricane or earthquake, both walls would be loaded and the demand on uplift restraint at a corner would depend on whether the two walls were loaded in phase or out of phase. If the walls were loaded in phase, the uplift demand on the corner would be the sum of the two walls' overturning. Therefore, counting on the uplift restraint provided by corner restraint without knowing whether the loading components are in or out of phase is not recommended. However, to provide information for use in future studies, the results are reported here.
On average, walls with 1.2-m corners reached higher ultimate capacities than walls with shorter perpendicular segments. The average reduction between initial and stabilized capacity was 17 percent (SD = 1.1%) and practically constant, which is in line with observations made on long walls with different anchorage conditions but loaded equivalently.
Considering average values, walls with 1.2-m corners show higher ultimate unit shear values, whereas the ultimate unit shear values of walls with 0.6-m corners are lower than walls without corners and overturning restraint (Fig. 4). However, this assumes that the shear load is distributed uniformly to the top of the wall, when in fact the shear load is distributed according to stiffness. Therefore, for this study to be complete, the results should be compared to straight shear walls that are 3.7 m long. In addition, the high variation of the values obtained from the walls with 0.6-m corner framing and the small sample size may also be contributing to the observed difference. It is interesting to note that the coefficient of variation of the tested unit shear for the two end-wall configurations is only 8.2 percent (Table 3).
Elastic stiffness was not significantly influenced by the length of the corner segments (Fig. 4), but is substantially higher when compared to straight walls on a unit-length basis. It should be noted that unit shear values were determined by Equation  using actual test results. This equation does not account for the effects of wall length on unit shear. In other words, the assumption was made that unit shear remains constant and uniform with changing wall length, and the data reflect the observed unit shears. This is in fact not true since the shear is distributed to the wall in proportion to the wall segment stiffness, with the end segment, next to the corner, carrying a higher portion of the lateral load.
Measurements of end stud movement suggest that perpendicular wall segments reduce the uplift by providing some hold-down effect. On average, total stud movement at capacity was reduced by 36 and 41 percent for walls with 0.6-m and 1.2-m corner framing, respectively, compared to a straight wall with no overturning restraint. However, the fact that the wall segment oriented perpendicular to the load was not directly loaded should be considered when reviewing the results.
Sugiyama and Matsumoto  developed empirical equations for estimating the racking strength of light-frame wood shear walls. The perforated shear wall method of design  adopted one equation as a method of designing shear walls to resist lateral loads with fewer mechanical connectors resisting the overturning moment than required for fully engineered shear walls. The method relates the capacity of a wall containing openings to the capacity of a fully sheathed wall using the following equation:
F = r/3 - 2r 
where F denotes the shear strength ratio as a function of the sheathing area ratio (r). The shear strength ratio accounts for the fact that the shear capacity for a wall with a given length decreases as the size of the openings increases. However, this equation was derived for walls with uplift restraint at the ends of the wall, and does not necessarily apply to shear walls without hold-down devices, as the failure mechanism of unrestrained walls differs greatly from mechanically anchored walls.
The predominant mode of failure for walls with maximum overturning restraint and restraints at wall ends tested cyclically was nail fatigue between framing and OSB sheathing at larger displacements (greater than that associated with wall capacity), and nail tear through at the top and bottom of sheathing panels after peak load was reached. Nails attaching OSB sheathing to the framing partially withdrew on the perimeter of the panels near corners, but failed predominantly due to fatigue. 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. Less nail fatigue occurred in walls with no overturning restraints than in walls with a maximum number of overturning restraints because the racking of the sheathing relative to the framing was less distinct.
Therefore, a new relationship for the shear load ratio as a function of the sheathing area ratio may need to be developed for conventional construction. Yet, the relatively small sample size of three different configurations tested does not permit the development of a prediction model with a high level of confidence. Future research will focus on analytical modeling of walls with no anchors to extend the sample size most economically and to advance a prediction model similar to Sugiyama's equations.
Experimentally determined and predicted shear strength ratios at capacity as a function of sheathing area ratio under cyclic loading are shown in Figure 5.
To see how the expression F = r was derived, consider a wall with openings whose height equals total wall height. Then, derived from Eurocode 5, the shear strength ratio may be determined by the following equation (2):
F = [F.sub.f,d] * [b.sub.1] * 1/s/[F.sub.f,d] * [b.sub.2] * 1/s = [b.sub.1]/[b.sub.2] 
But r may be expressed as:
r = 1/1+[sigma][A.sub.i]/H[sigma][L.sub.i] = 1/1+[sigma][A.sub.i]/H[b.sub.i] 
which leads to:
[b.sub.1] = [b.sub.2] H - [b.sub.1]H/H/r - H = [b.sub.2]r 
F = [b.sub.1]/[b.sub.2] = r 
where [b.sub.1] and [b.sub.2] denote the total length of full-height sheathing for the wall with openings and the total length of the wall without openings, respectively; s describes fastener spacing. Note that experimentally determined shear strength ratios for fully restrained walls almost coincide with the prediction equation. This highlights an important point. It means that for fully anchored walls, the relation F = r still holds when contributions of sections above and below openings are considered. If this proves to be true for a larger sample size, this very simple prediction model could form a more accurate design procedure in building codes. Figure 5 also indicates that the perforated shear wall method prediction model is fairly conservative when compared to actual shear strength ratios obtained during cyclic tests. Yet, for the purpose of design, there is no obvious reason why this relation should be conservative since it is directly proportional to a fully sheathed wall with a safety factor alread y applied. Equation  has been proposed to predict the shear strength ratio at capacity, although this equation was derived from data obtained at a shear deformation angle of 1/100 radian or 24-mm story drift for a 2.4-m-tall wall. However, most displacements at capacity for the walls tested here were significantly higher. Given the fact that Equation  predicts fairly conservative values, the following relation, which was developed by Sugiyama and Matsumoto  to predict the shear strength ratio at 41-mm story drift, may be more appropriate to predict the shear strength ratio at capacity (Fig. 5):
F = r/2 - r 
Based on the limited number of tests, the number of overturning restraints is positively correlated with ultimate capacity and elastic stiffness. The magnitude of influence depends on the number of openings in the wall. The greater the openings, the higher the increase in capacity and stiffness when overturning restraint is increased from zero to maximum. Without overturning restraints, shear walls exhibit a pronounced rigid body rotation arising from uplift and separation along the bottom plate.
Cyclically tested wall specimens ultimately resisted higher loads (initial capacity) compared to specimens tested monotonically when overturning restraints were omitted.
In these tests, corner framing provided some hold-down effect when compared to straight walls with no overturning restraint and no perpendicular walls attached. However, tests that load both walls simultaneously should be conducted to quantify the hold-down effect of corner walls under more realistic conditions. While buildings are typically designed assuming one-directional loading, they really experience loading in three orthogonal directions during an earthquake. Depending on whether the loadings in each direction are in phase or not determines the magnitude of hold-down effect for corner framing. If the loading were as tested, the hold-down effect provided by corner framing would be sufficient to provide for development of unit shear slightly less than, but comparable to, straight walls with hold-down devices. Straight shear walls that are 3.7 m in length should be tested to better quantify the effect of the corner walls.
Walls with corners attached exhibited different failure modes compared to straight walls. Based on this observation, shear wall tests carried out in the future should always have corner framing to insure a more realistic performance.
Using sheathing-area ratio as a variable to classify walls including openings, a very simple and less conservative design model for fully restrained walls was found. Results for three constrained wall specimens (conventional construction) indicate that the perforated shear wall method provides a conservative prediction of capacity. Additional test results would, however, increase confidence in applying the perforated shear wall method to conventional construction.
The design method employed by the Perforated Shear Wall Method using one of Sugiyama's equations proved to be conservative based on the limited data available. A better predictor was another equation developed by Sugiyama for larger story drifts.
The authors are, respectively, Research Associate and Professor, Brooks Forest Products Research Lab., Dept. of Wood Sci. and Forest Products., Virginia Polytechnic Institute and State Univ., 1650 Ramble Road, Blacksburg, VA 24061-0503. This paper was received for publication in January 2000. Reprint No. 9083.
*Forest Products Society Member.
[C]Forest Products Society 2001.
Forest Prod. J. 5l(718):65-72.
(1.) Adams, N.R. 1987. Plywood shear walls. Res. Rept. 105. APA-The Engineered Wood Assoc., Tacoma, WA.
(2.) Alsmarker, T. 1995. Diaphragms and shear walls. Step/Eurofortech Vol. 1 Timber Engineering. Almere, The Netherlands. Distributed by the National Representative Org. (NRO).
(3.) American Forest & Paper Association. 1995. Wood frame construction manual for one- and two-family dwellings - SBC high wind edition. AF&PA, Washington, DC.
(4.) American Society for Testing and Materials. 1997. Annual Book of ASTM Standards. ASTM, West Conshohocken, PA.
(5.) _____. 1998. Standard test method for cyclic properties of connections assembled with mechanical fasteners. Proposed Standard by Task Group 11 of ASTM E6.13. ASTM, West Conshohocken, PA.
(6.) Dolan, J.D. 1989. The dynamic response of timber shear walls. Ph.D. thesis. Univ. of British Columbia, Vancouver, BC, Canada.
(7.) _____. 1993. Proposed test method for dynamic properties of connections assembled with mechanical fasteners. Pap. presented at 26th Meeting of CIB W18 in Athens, GA.
(8.) _____ and B. Madsen. 1992. Monotonic and cyclic nail connection tests. Canadian J. of Civil Engineering 19:97-104.
(9.) _____ and A.C. Johnson. 1996. Cyclic tests of long shear walls with openings. Timber Engineering Rept. TE-1996-002. Virginia Polytechnic Inst. and State Univ., Blacksburg, VA.
(10.) Dowrick, D.J. 1986. Hysteresis loops for timber structures. Bull. of the New Zealand National Soc. of Earthquake Engineering 19(20): 143-152.
(11.) Foschi, R.O. 1980. Summary on racking tests conducted on waferboard sheathing. TECO Project 80-37 Rept., Washington, DC.
(12.) Heine, C.P. 1997. Effect of overturning restraints on the performance of fully sheathed and perforated timber framed shear walls. M.S. thesis. Virginia Polytechnic Inst. and State Univ., Blacksburg, VA.
(13.) Johnson, A.C. 1997. Monotonic and cyclic performance of long shear walls with openings. M.S. thesis. Virginia Polytechnic Inst. and State Univ., Blacksburg, VA.
(14.) Porter, M.L. 1987. Sequential phased displacement procedure (SPD) for TCCMAR testing. In: Proc. of the Third Meeting of the Joint Technical Coordinating Committee on Masonry Research. U.S.-Japan Coordinated Earthquake Res. Program, Tomamu, Japan.
(15.) Standard Building Code. 1994. Southern Building Code Congress Inter., Burmingham, AL.
(16.) Stewart, W.G. 1987. The seismic design of plywood sheathed shearwalls. Ph.D. thesis. Univ. of Canterbury, NZ.
(17.) Sugiyama, H. and T. Matsumoto. 1994. Empirical equations for the estimation of racking strength of a plywood sheathed shear wall with openings. Mokuzai Gakkaishi 39(8):924-929.
(18.) Tissel, J.R. and J.D. Rose. 1988. Plywood and shear walls in mobile homes. Res. Rept. 151. American Plywood Assoc., Tacoma, WA.
(19.) Yasumura, M. and H. Sugiyama. 1984. Shear properties of plywood-sheathed wall panels with opening. Transactions of the Architectural Inst. of Japan 338(4):88-98.
TABLE 1. -- Opening sizes for wall configurations. Sheathing Opening size Wall type area ratio Door Window a A 1.00 -- -- B 0.48 2.03 x 1.22 [m.sup.2] 1.22 x 2.40 [m.sup.2] 2.03 x 3.66 [m.sup.2] C 0.30 (sheathed at ends) -- 2.44x 8.53 [m.sup.2] D 0.76 2.03 x 1.22 [m.sup.2] 1.73 x 2.40 [m.sup.2] E 0.55 2.03 x 1.22 [m.sup.2] 1.22 x 2.40 [m.sup.2] 1.22 x 3.66 [m.sup.2] (a)The top of the window is located 16 inches from the top of the wal TABLE 2 Monotonic, initial cyclic, and stabilized cyclic capacity. a Wall specimens No tie-down Anchors at end anchors b of wall only c [F.sub.peak] A B C A d Monotonic (kN) 111.65 43.59 19.57 171.70 Initial SPD (kN) 118.77 44.93 21.35 142.34 Stabilized SPD (kN) 100.08 38.70 18.24 121.88 Stabilized/initial 0.84 0.86 0.85 0.86 Maximum amount of tie-down anchors b [F.sub.peak] B C A d B C Monotonic (kN) 53.38 35.59 153.91 68.95 48.04 Initial SPD (kN) 49.82 32.47 123.22 59.61 37.37 Stabilized SPD (kN) 42.70 28.02 105.42 52.04 31.58 Stabilized/initial 0.86 0.86 0.86 0.87 0.85 (a)The values presented are rounded. All calculations were done with original values. (b)These specimens had OSB sheathing. (c)These specimens had plywood sheathing (13). (d)Wall A has the same anchorage requirements for the anchors at the end of the wall only and a maximum number of tie-down anchors due to being fully sheathed. TABLE 3. -- Cyclic load resistance of 3.7-m-long walls with corner framing. Wall specimens 0.6-m corners 1.2-m corners [F.sub.peak] Wall 1 Wall 2 Average Wall 1 Initial SPD (kN) 36.92 30.69 33.81 37.81 Stabilized SPD (kN) 30.25 24.91 27.58 32.03 Stabilized/initial 0.82 0.82 0.82 0.84 Unit shear (kN/m) 10.10 8.39 9.25 10.33 [F.sub.peak] Wall 2 Average Initial SPD (kN) 37.81 37.81 Stabilized SPD (kN) 31.58 32.03 Stabilized/initial 0.84 0.84 Unit shear (kN/m) 10.33 10.33 Figure 4. -- Capacities and stiffness of walls with corner framing compared to long straight walls (black bars represent stiffness; white bars represent unit shear). Unit shear Elastic stiffness 0.6 m Corner (3.7 m) 9.25 1.78 1.2 m Corner (3.7 m) 10.33 1.72 No anchors (12.2 m) 9.75 0.86 Anchors (12.2 m) 10.89 0.98 Note: Table made from bar graph
|Printer friendly Cite/link Email Feedback|
|Author:||HEINE, CHRISTIAN P.; DOLAN, J. DANIEL|
|Publication:||Forest Products Journal|
|Article Type:||Statistical Data Included|
|Date:||Jul 1, 2001|
|Previous Article:||PARTICLEBOARD MADE FROM REMEDIATED CCA-TREATED WOOD: EVALUATION OF PANEL PROPERTIES.|
|Next Article:||PERMANENT BRACING DESIGN FOR MPC WOOD ROOF TRUSS WEBS AND CHORDS.|