The racking performance of light-frame shear walls with

The racking performance of light-frame shear walls with various tie-down restraintsSalenikovich, Alexander J.1, Dolan, J. Daniel 2

ABSTRACT

Overturning restraint of engineered shear walls is typically provided by means of mechanical tie-down devices at the wallends. As opposed to engineered construction, conventional walls are built without such mechanical overturning restraintsand overturning forces are typically resisted by the sheathing fasteners. Nevertheless, conventional buildings throughoutlarge territory of the US may become subject to substantial wind and seismic forces. The response of conventional light-frame timber shear walls to lateral forces is not well studied. The objective of the presented experimental study is toexamine and to compare performance parameters of conventional and engineered shear walls with various height-to-length ratios. Presented are results of monotonic and reversed cyclic tests of full-size shear walls with height-to-lengthratios 4:1, 2:1, 1:1, and 2:3 with oriented strandboard sheathing. Three overturning restraint conditions were tested: 1)Tie-down anchors at the end studs and shear bolts along the bottom plate; 2) Shear bolts along the bottom plate, and 3)Nailing along the bottom plate. To obtain conservative estimates, no dead load was applied in the wall plane during thetests. Test results revealed that stiffness and strength per unit length of conventional walls were significantly lower thanthose of engineered walls, and decreased as the wall aspect ratio increased. The response of walls with the nailed bottomplates was similar to that of walls with shear bolts, provided a sufficient number of nails was used to prevent separation ofthe wall from the platform. It is shown that current nailing schedules for attachment of conventional walls to lowerstructures are not adequate to prevent wall overturning.

INTRODUCTION

Light-frame shear walls in a typical North American house are major components of a lateral force resisting system. Ifdesigned to resist high wind and/or seismic forces, walls at each story often require mechanical fasteners, such as tie-down anchors and shear bolts, to provide continuous and complete load paths from the top of the building to thefoundation. Shear forces are transferred by the shear bolts, which are typically installed equally spaced along the top andbottom plates of the walls. Tie-down anchors are designed to resist overturning moments caused by the shear forces, andthey are attached to the end studs of each fully-sheathed wall segment. As opposed to the engineered design,conventionally-built walls are secured to underlying structures by nails or shear bolts only. Previously, numerousexperimental studies were conducted on shear walls fully restrained against overturning according to ASTM standards E72 and E564 (ASTM 1995a and ASTM 1995b). Those tests served as a basis for establishing design values forengineered shear walls. Although many non-engineered buildings throughout large territory of the US may also becomesubject to substantial wind and seismic forces, the authors are unaware of any previous attempts to measure the responseof non-anchored shear walls.

At Virginia Tech, a comprehensive study has been launched that combines experimental and numerical analyses of shearwalls of various configurations including conventional and engineered walls. The experimental program includes tests offull-size shear wall specimens under various loading regimes. A part of the study presented in this paper is focused oncomparison of the racking performance of conventional (non-anchored) walls with engineered fully-anchored walls. Theobjective of the tests is to reveal the effects of various means of attaching the wall to the platform on the wallperformance under monotonic and cyclic loading. Quantification of shear resistance for conventional walls is of interestfor several reasons. Evaluating strength and deflection capacity of non-anchored walls will help estimate the potentialhazard of structural damage and failure from natural disasters. Evaluating stiffness parameters will help estimate thecontribution of non-anchored walls into the structural response of engineered buildings. Tests of walls with various 1 Graduate Research Assistant, Department of Wood Science and Forest Products, Virginia Polytechnic Institute and StateUniversity, Brooks Center (0503), 1650 Ramble Rd., Blacksburg, VA 24061, USA2 Associate Professor, Department of Wood Science and Forest Products, Virginia Polytechnic Institute and StateUniversity, Brooks Center (0503), 1650 Ramble Rd., Blacksburg, VA 24061, USA

aspect (height-to-length) ratios will demonstrate effects of the wall length on the shear stress distribution. Informationobtained in this study improves our understanding of shear wall performance and can be used in development of amechanics-based model for shear wall analysis, which would account for various details of shear wall assembly.

EXPERIMENTAL

Load regimesDisplacement history plots for monotonic and cyclic test procedures are shown in Figure 1. Each test was stopped whenthe specimen fully exhausted its ability to resist load. Monotonic load was applied at 15 mm/min (0.6 in./min.) in a singlestage. Cyclic load was applied at a frequency of 0.25 Hz with three cycles in each phase following three initial phases ofsingle cycles. The amplitudes of the phases were 2.5, 5.0, 7.5, 10, 20, 30 percent, etc. of the ultimate monotonicdisplacement. The average ultimate displacement of 76 mm (3.0 in.) was observed during preliminary monotonic tests.

-127

-102

-76

-51

-25

0

25

51

76

102

127

0 24 48 72 96 120 144 168Time (sec.)

Dis

plac

emen

t (m

m)

-5

-4

-3

-2

-1

0

1

2

3

4

5

(in.)

Cyclic loadMonotonic load

Figure 1. - Monotonic and cyclic test protocols.

The definition of the cyclic protocol is consistent with principles of the proposed ISO quasi-static reverse-cyclic testmethod for joints with mechanical fasteners (ISO 1998). At the same time, the cyclic protocol can be described in termsof the sequential phased displacement (SPD) procedure adopted by the Structural Engineers Association of SouthernCalifornia (SEAOSC 1997) for testing framed walls. According to the SPD protocol, the similar pattern of cycling wouldcorrespond to the first major event (FME) of 7.6 mm (0.3 in.) with decay cycles eliminated. The decay cycles wereexcluded from the test protocol to reduce the energy demand imposed on the wall and, therefore, to make it more realistic.

SpecimensAll specimens were 2.4-m (8 ft.) tall. Walls with the following aspect ratios were tested (see Figure 2):• 4:1 - the size often used to accommodate garage doors and wide ‘view’ windows.• 2:1 - the minimum width of the conventional wall allowed in US model codes to resist high wind or high seismic

loads. It is the most typical width of fully-sheathed wall segments used to fill the space between windows and doorsin residential buildings.

• 1:1 - the specimen size historically used in ASTM standard tests. It is a reference point for comparison with anyother tests.

• 2:3 - the size used to investigate if longer walls have the same performance characteristics as square walls. It isuncommon that fully-sheathed wall segments in a modern house exceed 3.6 m in length.

Figure 2 shows schematic layouts of wall framing and location of tie-down anchors and shear bolts when those areapplied. The framing for each specimen was assembled of 38×89 mm (2×4 in.-nominal) spruce-pine-fir (SPF) stud grademembers spaced 406 mm (16 in.) on centers, except for the walls with the aspect ratio of 4:1 with studs spaced 533 mm(21 in.) on centers. End studs consisted of two members fastened by two 16d (∅4.1×89 mm) common nails every 0.6 m(2 ft.). The studs were attached to the single bottom plate and the double top plate with two 16d common nails at eachend. A single layer of oriented strandboard sheathing, 11 mm (7/16 in.) thick, was attached to one wall side bypower-driven 8d (∅3.3×63.5 mm) common SENCO® nails at 152 mm (6 in.) on centers along the edges and 305 mm

(12 in.) on centers along intermediate studs. The long dimension of the sheathing was oriented parallel to the studs.Control specimens had sheathing fastened to the framing with 19-mm (3/4-in.) edge distance along the top and bottomplates. Other specimens were intentionally nailed with 10-mm (3/8-in.) edge distance to represent the minimumallowable by current design provisions (BSSC 1997).

0.6 m

4:1

2 ft.

1:1

2.4 m8 ft.

2:1

1.2 m4 ft.

3.6 m12 ft.

2:3Double top plate

Single bottom plate

Single intermediate studs

Dou

ble

end

stud

2.4

m (8

ft.)

Tie-down anchors

Shear bolts

Figure 2. - Schematic of wall specimens.

Each wall configuration was tested with the following overturning restraint conditions (details are shown in Figure 3):• FA - ‘full attachment’. Represented engineered construction where end studs of fully sheathed wall segments areattached to foundation with tie-down anchors. As a means of overturning restraint, Simpson Strong-Tie® HTT-22connectors were fastened to the end studs by thirty-two 16d (∅3.8×82.6 mm) sinker nails and anchored to the base with15.9-mm (5/8-in.) diameter Grade A bolts. These bolts were installed in oversized holes and were instrumented withstrain gages to measure uplift forces resisted by the anchors. Bottom plate was attached to the base with 15.9-mmdiameter shear bolts every 0.6 m (2 ft.).• IA - ‘intermediate attachment’. Represented conventional construction where bottom plate is attached to foundationwith shear bolts only. The same size and location patterns of shear bolts were used as for walls with full attachment (FA).No tie-down anchors were applied.• NA - ‘nail attachment’. Represented conventional construction where no mechanical devices other than nails are usedto attach bottom plate of the wall to the platform.

TEplhymca

a ) c

Bottom plate

3×5-in. steel beam

Dou

ble

end

stud

Simpsontie-down

Ø5/8-in. bolts

Figure 3. - Overturning restraint de

est setupach specimen was tested in a horizontal posane of the wall, which conservatively repredraulic actuator was distributed through a m (5/8-in.) bolts spaced 610 mm (24 in.) opacity of 245 KN (55 Kips), was secured

b

Bottom plate

Dou

ble

end

stud

Ø5/8-in. bolt

3×5-in. steel beamtails: a) FA walls, b) IA walls, and c) NA w

ition, as is shown in Figure 4. In this setupsented a wall parallel to floor joists. The r76×127-mm (3×5-in.) steel beam attached n centers. The actuator, with a displacemebetween the support and the load distributi

)

Dou

ble

end

stud

Wood sill plate

16d nails

Bottom plate

alls. (1 in. = 25.4 mm)

, no dead load was applied in theacking load from a programmableto the wall top plate with ∅15.9-

nt range of ±152 mm (6 in.) and aon beam by means of the hinged

connections shown in Figure 4. Two casters were fixed to the beam, to allow free movement of the top of the specimen ina parallel direction to the applied load. The casters rolled along the greased surface of plastic pads laid on the floor toreduce any friction induced by the wall weight. Horizontal translation of the wall top was measured by the LVDT built inthe load cell. In addition, several transducers were applied to measure vertical displacement of the studs, horizontal slipof the bottom plate, and sheathing movement.

Load distributionbeam on casters

PLAN VIEW

A - A

Steel beamsanchored tofoundation

A A

Plastic pads onconcrete floor

Hinges

Hydraulicactuator

Reinforced-concretefoundation

Figure 4. - Test setup.

Definition of performance parametersTo compare the performance of walls with different aspect ratios, all response parameters discussed below are evaluatedon a unit length basis. The hysteresis curve in Figure 5 represents a typical load-deflection response observed in thecyclic tests of engineered walls. Two envelope curves, as shown in Figure 5, are obtained from the hysteresis curve: onefor the first and the other for the last cycle in each phase of the loading. These curves are referred to as initial andstabilized response curves and are analyzed like the monotonic response curves. Since the envelope curves includereversed sides, the absolute values of parameters of the negative and the positive curves are averaged during the analysis.

Deflection (∆)

Uni

t she

ar lo

ad (v

)

Hysteresis curve

Initial envelope

Stabilized envelope

Figure 5. - Hysteresis and envelope curves.

Deflection (∆)

Uni

t she

ar lo

ad (v

)

vpeakvyield

vfailure= 0.8 vpeak

0.4vpeak

∆peak∆yield ∆failure

Equivalent energy elastic-plastic curveObserved monotonic or envelope curve

Figure 6. - Performance parameters.

Wall strength (vpeak), deflection capacity (∆peak), deflection at 0.4vpeak, and failure point (vfailure, ∆failure) are determined foreach monotonic, initial, and stabilized response curve using a plot similar to that shown in Figure 6. The failure point is

considered at 0.8 vpeak (i.e., when a 20% decrease in resistance occurs) (ISO 1998). The area under the curve limited bythe failure point approximates the work that can be done by the wall during a monotonic test. Using these data, theequivalent energy elastic-plastic (EEEP) curve is derived for each specimen, as is shown in Figure 6. The initial slope ofthe EEEP curve, drawn through 0.4 vpeak on the observed curve, determines the shear modulus. The yield point (vyield,∆yield) is found by equating the areas under the observed and EEEP curves. The bilinear EEEP curves obtained this wayallow comparison of the nonlinear performance of different walls on an equivalent energy basis.

RESULTS AND DISCUSSION

Test results and design parameters of engineered fully-anchored (FA) walls have been discussed in the previouspublication (Salenikovich and Dolan 1999). In summary, it was found that the response of the walls did not depend ontheir size with the exception of 0.6-m (2-ft.) walls, which were significantly weaker and more flexible than other walls.In monotonic tests, the FA walls resisted up to 10 KN/m (0.69 Kip/ft.) at deflections exceeding 64 mm (2.5 in.). Cyclicloading and reduction of the edge distance caused significant decrease of strength and deflection capacity of the FA walls.On the average, 1.2-m and longer walls developed stabilized cyclic strength of 7.3 KN/m (0.50 Kip/ft.) at 44-mm (1.7-in.)deflection. In this publication, we would like to discuss the performance of conventional walls in more detail.

According to prescriptive code (ICC 1998), no more than three 16d nails at 406 mm (16 in.) on centers are required tofasten the bottom wall plate to the lower structure. To demonstrate that this requirement is not adequate to prevent walloverturning, a series of monotonic tests was conducted on 1.2-m (4-ft.) NA walls attached to the platform with 16d nailsin various patterns. Figure 7 shows load-deflection curves recorded during these tests. In the absence of gravity load inthe wall plane, walls attached with 12, 18, and 24 nails simply ‘walked’ off the foundation without any sign of walldamage. Only 36 nails in three rows at 3 in. on centers were sufficient to hold down the bottom plate of the 1.2-m wall.This nailing pattern was used in tests of longer NA walls, and there was no practical method to attach the 0.6-m (2-ft.) NAwalls.

0

1

2

3

4

5

6

0 13 25 38 51Displacement (mm)

Uni

t sh

ear

load

(K

N/m

)

0.00

0.07

0.14

0.21

0.27

0.34

0.0 0.5 1.0 1.5 in.

Kip

/ft.

12 2418 36Number of nails in bottom plate:

Figure 7. - Effect of nailing to the platform.

0

1

2

3

4

5

6

0 13 25 38 51Displacement (mm)

Uni

t sh

ear

load

(K

N/m

)

0.00

0.07

0.14

0.21

0.27

0.34

0.0 0.5 1.0 1.5 in.

Kip

/ft.

19-mm edge distance10-mm edge distance

Figure 8. - Effect of reduced edge distance.

Two monotonic tests of 2.4-m (8-ft.) square walls with sheathing attached at 19-mm (3/4-in.) edge distance representedthe control sample for non-engineered walls. The IA walls were attached to the platform with shear bolts. To determineeffects of reduced edge distance on the wall performance, two additional monotonic tests were performed on similar wallswith the sheathing attached at the 10-mm (3/8-in.) edge distance. Figure 8 shows the average load-deflection curves foreach of these samples. The walls with reduced edge distance exhibited approximately 20% lower strength (vpeak) and30% lower deflection at failure (∆failure). Consequently, the work to failure of these walls was decreased almost by half.In other words, the walls with larger edge distance along the top and bottom of the sheathing panel were tougher and hadsignificantly higher deflection capacity. Similar conclusions were made from the tests of the FA walls.

To obtain conservative estimates, all tests discussed below were performed on walls with sheathing attached at theminimum allowable edge distance (BSSC 1997). Figure 9 shows load-deflection curves representing the averagemonotonic response of non-anchored walls. The cyclic response of these walls is represented by stabilized envelope

curves shown in Figure 10. Both monotonic and cyclic tests revealed common trends in response of the non-anchoredwalls. These walls were considerably weaker than the anchored walls and their response was dependent upon the aspectratio in the entire range of tested configurations. After reaching peak loads at deflections between 25 and 33 mm (1.0 and1.3 in.), longer walls degraded rapidly; in the contrast, narrow walls exhibited considerable ductility.

Measurements of the stud uplifting and sheathing movement suggest that rigid body rotation controlled the response ofnon-anchored walls. During monotonic tests, these walls, with minimum racking, rocked around the wall toe causingnails to tear through the sheathing edge (“unzip”) at the bottom plate, as shown in Figure 11a. Most of the load wasresisted by sheathing nails on one end of the bottom plate. Long walls, having longer arms of rotation than narrow walls,imposed high deflection demands on these nails. The peak load was registered as soon as the first sheathing nail in thetension corner tore through the panel edge. Bottom plates in the IA walls experienced considerable bending between theend of the wall and the first shear bolt located 305 mm (12 in.) away from the corner (see Figure 3b). The deflection ofthe bottom plate delayed failure of sheathing connections allowing load shearing between the nails in that corner.Conversely in the NA walls, bottom plates were fastened to the sill starting from the corner (see Figure 3c). Thisattachment prevented bending of the bottom plate and lead to earlier unzipping of sheathing connections and toconsiderably weaker performance of the NA walls in monotonic tests. During the cyclic tests, the center of wall rotationmoved away from the corners. This reduced the translation demand on sheathing nails along the bottom plate andallowed distribution of load between the nails on both wall ends. Therefore, initial cyclic response parameters weresometimes higher than in corresponding monotonic tests.

0

1

2

3

4

5

6

0 13 25 38 51Displacement (mm)

Uni

t sh

ear

load

(K

N/m

)

0.00

0.07

0.14

0.21

0.27

0.34

0.0 0.5 1.0 1.5 in.

Kip

/ft.

2:1 4:12:3 1:1Aspect ratio:

0

1

2

3

4

5

6

0 13 25 38 51Displacement (mm)

Uni

t sh

ear

load

(K

N/m

)

0.00

0.07

0.14

0.21

0.27

0.34

0.0 0.5 1.0 1.5 in.

Kip

/ft.

2:12:3 1:1Aspect ratio:

Figure 9. - Monotonic response curves: a) IA walls and b) NA walls. Wall height = 2.4 m (8 ft.)

0

1

2

3

4

5

6

0 13 25 38 51Displacement (mm)

Uni

t sh

ear

load

(K

N/m

)

0.00

0.07

0.14

0.21

0.27

0.34

0.0 0.5 1.0 1.5 in.

Kip

/ft.

2:1 4:12:3 1:1Aspect ratio:

0

1

2

3

4

5

6

0 13 25 38 51Displacement (mm)

Uni

t sh

ear

load

(K

N/m

)

0.00

0.07

0.14

0.21

0.27

0.34

0.0 0.5 1.0 1.5 in.

Kip

/ft.

2:12:3 1:1Aspect ratio:

Figure 10. - Stabilized cyclic response curves: a) IA walls and b) NA walls. Wall height = 2.4 m (8 ft.)

a) b)

a) b)

Apart from geometry, performance of walls without tie-downs was controlled by the number of sheathing nails in thebottom plate and their distance from the panel edge. Observations of failure modes explain the early wall degradationduring the tests. The walls failed when sheathing nails were torn through the panel edge along the bottom plate followingthe separation of studs from the plate. The shorter the edge distance, the less deflection it took to ‘unzip’ the connection.Sheathing nails along studs had their heads pulled through the sheathing, and their edge distance did not control thefailure.

For comparison, Figure 11b illustrates the failure of a FA wall due to significant racking deformations of individual wallsegments. Measurements of axial forces in tie-down bolts indicated increase of tension in the anchors on both ends of thewall. It means that the anchors not only hold down the tension stud but also restrain pivoting of the compression studproviding additional resources to the wall strength, stiffness, and ductility. The nail fatigue occurred only during cyclictests of the FA walls when the energy demand exceeded 5 KN⋅m/m (1.12 Kip⋅ft./ft.). The NA walls dissipated less than2 KN⋅m/m (0.45 Kip⋅ft./ft.) in cyclic tests, which was insufficient for developing nail fatigue.

NEHshearand dappro

where

and w

a )

Figure 11. - Failure modes: a) non-anchored w

COMPARISON OF DESIGN PA

RP Provisions (BSSC 1997) establish guidelines for seismic desig walls such as shear resistance (v), structural overstrength coeffeflection amplification factor (Cd). Corresponding performaximated and compared with the published design values using the

Ω0* = vpeakλφ / v ≤ Ω0 = 2.5

R* = Rd Ω0* ≥ R = 7

Cd* = ∆peak /∆elastic ≥ Cd = 4.5

, v = 4.8×0.82, factored shear resistance in KN/m (= 0.33×0.8framing members of spruce-pine-fur (specific gravλ = 1.0, time effect factor, from Table 12.4.3-2 (BSSC 19

Rd = ∆design /∆yield, ductility reduction factor;∆design = 0.025H or ∆peak, whichever is smaller, H = 2.4 m (8 ft∆elastic = 0.025HI/Cd, elastic deflection demand, I = 1.0, occup

here, vpeak, ∆peak, and ∆yield determined according to Figure 6.

b

alls and b) anchored walls.

RAMETERS

n of buildings, including parameters of light-frameicient (Ω0), response modification coefficient (R),nce parameters of the tested shear walls were following formulas:

(1)

(2)

(3)

2 in Kip/ft) for seismic forces on shear walls withity G ≥ 0.42), φ = 0.65, resistance factor, and97);

.), wall height;ancy importance factor;

Table 1. Performance parameters.

FA IA NAAspect ratio Response

Ω0* R* Cd

* Ω0* R* Cd

* Ω0* R* Cd

*

4:1monotonic

cyclic initialcyclic stabilized

1.211.080.99

3.294.384.29

9.726.356.46

0.340.240.21

1.871.160.71

2.023.392.97

NOT TESTED

2:1monotonic

cyclic initialcyclic stabilized

1.461.361.20

6.746.064.52

4.813.963.53

0.400.440.38

1.481.290.97

2.482.662.41

0.410.400.35

1.402.161.15

3.032.822.41

1:1monotonic

cyclic initialcyclic stabilized

1.641.311.14

7.546.264.00

5.423.382.97

0.660.720.60

2.011.861.21

1.901.841.56

0.560.730.63

1.062.891.28

1.481.981.42

2:3monotonic

cyclic initialcyclic stabilized

1.631.441.25

7.276.574.77

5.903.233.25

0.990.880.74

3.192.761.57

2.501.841.56

0.740.860.76

2.541.781.57

1.821.701.70

Coefficients given in Equations (1) to (3) provide sufficient information on the load and deformation capacity of the wallssubjected to lateral loads and allow direct non-dimensional comparisons with the design values. Average parameters forvarious wall configurations, including monotonic and cyclic responses, are summarized in Table 1.

Results in Table 1 indicate that only the monotonic response of the FA walls (with 19-mm edge distance) was adequate tocurrent design values if the aspect ratio requirements were satisfied (≤ 2:1). The cyclic responses of the FA walls withthe reduced edge distance were significantly below the design requirements; in other words, the walls lacked deformationcapacity and strength. None of the tested IA and NA walls satisfied design requirements under monotonic or cyclicloading. Although the values shown in Table 1 are based on a few replications, they represent the range of actions (forcesand deflections) that they are capable of sustaining during a hazard event. Performance parameters of non-anchored wallsdepend greatly on the wall size; therefore, the estimates of shear wall resistance should consider the wall aspect ratio.

CONCLUSION

Monotonic and cyclic tests of anchored and non-anchored walls of various sizes proved that high aspect ratio andinsufficient sheathing edge distance for fasteners were the weakest links in shear wall resistance. To provide adequateracking performance of engineered shear walls, the minimum requirements for the edge distances along the top andbottom plates should be increased. This conclusion is based on the tests of walls with the long dimension of sheathingpanels oriented parallel to the studs. Based on results of the cyclic tests, design parameters for shear walls need revision.

Test results revealed that stiffness and strength per unit length of conventional walls were significantly lower than thoseof engineered walls, and decreased as the wall aspect ratio increased. The response of walls without tie-down anchorsunder monotonic and cyclic loading was strongly dependent on quality of sheathing attachment at the bottom plate. Rigidbody rotation contributed significantly to the wall performance. The response of walls fastened to foundation by nailswas similar to that of walls with shear bolts, provided a sufficient number of nails was used to prevent separation of thewall from the platform. It is shown that current nailing schedules for attachment of conventional walls to lower structuresare not adequate to prevent wall overturning and must be reconsidered. Direct comparison of engineered andconventional walls allowed conservative estimate of the lateral forces and deflections they are capable of withstanding.

REFERENCES

American Society for Testing and Materials (ASTM). 1995a. Standard test methods of conducting strength tests of panelsfor building construction. ASTM E 72-95, ASTM, West Conshohocken, PA.

_____. 1995b. Standard practice for static load test for shear resistance of framed walls for buildings. ASTM E 564-95.ASTM, West Conshohocken, PA.

Building Seismic Safety Council (BSSC). 1997. NEHRP recommended provisions for seismic regulations for newbuildings and other structures. BSSC, Washington, D.C.

International Code Council (ICC). 1998. International residential code for one-and two-family dwellings. First Draft.ICC, Whittier, CA.

International Organization for Standardization (ISO). 1998. Timber structures – Joints made with mechanical fasteners –Quasi-static reversed-cyclic test method. WG7 Draft. ISO TC 165 Secretariat, Standards Council of Canada.

Salenikovich A. J. and Dolan J. D. 1999. Effects of aspect ratio and overturning restraint on performance of light-frameshear walls under monotonic and cyclic loading. In Proceedings of the Pacific Timber Engineering Conference Volume 3,14-18 March 1999, Rotorua, New Zealand, p.p. 205-212.

Structural Engineers Association of Southern California (SEAOSC). 1997. Standard method of cyclic (reversed) loadtests for shear resistance of framed walls for buildings. SEAOSC, Whittier, CA.