Cold-Formed Steel Framed Gypsum

Shear Walls: In-Plane Response

Q. Peck1; N. Rogers2; and R. Serrette, A.M.ASCE3

Abstract: Cold-formed steel (CFS) framed gypsum shear walls are common elements on light-frame structures, and they may be designed as
the primary lateral force-resisting (LFR) system or function as secondary LFR elements. This paper presents the results of monotonic and
Downloaded from ascelibrary.org by New York University on 05/16/15. Copyright ASCE. For personal use only; all rights reserved.

reversed cyclic gypsum shear wall tests and discusses wall performance in detail. Increased strengths and wall toughness may be achieved
with closer intermediate panel fastener spacing, and the difference between monotonic and reversed cyclic shear wall responses, although
different, may not be significant. Additionally, aspect ratio and abutting supported vertical panel joints have little effect on shear wall per-
formance. The scope of testing also introduces three new panel attachment configurations for design. Overall, this paper’s discussion provides
a helpful supplemental commentary on the performance of CFS framed gypsum shear walls. DOI: 10.1061/(ASCE)ST.1943-541X.0000521.
© 2012 American Society of Civil Engineers.
CE Database subject headings: Gypsum; Shear walls; Cold-formed steel.
Author keywords: Gypsum; Fastener spacing; Shear wall; Strength; Monotonic; Reversed cyclic; Blocking; Aspect ratio.

Introduction Present Provisions for Gypsum Board Shear Walls

Gypsum-sheathed walls are common elements in building Seismic design of CFS light-framed bearing wall structures
structures. These walls may be designed to provide lateral resis- sheathed with wood structural panels or steel sheets for lateral re-
tance or to be used as nonstructural partition walls. In commer- sistance can be accomplished using a maximum seismic response
cial buildings, gypsum partition walls are often detailed to be modification coefficient R [ASCE/Structural Engineering Institute
(SEI) 7-10 (ASCE 2010)] of 6.5, with an associated deflection am-
uncoupled from the primary lateral force-resisting system
plification factor C d ¼ 4:0 and an overstrength factor Ωo ¼ 3:0.
(PLFRS). This uncoupling is intended to limit interaction be-
Both R and C d are design seismic response parameters that attempt
tween the gypsum wall and the PLFRS. In contrast, gypsum to relate the linear elastic response of a structure at a defined design
walls in light-frame bearing wall structures (particularly residen- basis seismic event to a strength level design of the structure
tial structures) are rarely detailed to be uncoupled from the [NEHRP P-750 (FEMA 2009b)]—the code design base seismic
PLFRS, and no special attention is given to the design of these event is taken as two-thirds the maximum considered seismic event.
walls to accommodate the drifts expected in the PLFRS. It is A simple conceptual illustration of the relationship between R, C d ,
generally recognized, however, that gypsum walls play an and Ωo is shown in Fig. 1. Generally, C d is less than or equal to R,
important role in the overall seismic performance of light- and it is noted that for systems “where C d is substantially less than
frame structures [FEMA P695 (FEMA 2009a); Consortium of R, the system is considered to have damping greater than the nomi-
Universities for Research in Earthquake Engineering (CUREE) nal 5 percent of critical damping” [NEHRP P-750 (FEMA 2009b)].
2004a, b]. To complement existing U.S. design data and explore CFS light-frame wood structural panel and steel sheet shear wall
alternative panel attachment schedules, this paper presents test structures fall into the category of structures with damping greater
results of several full-scale cold-formed steel (CFS) framed than 5%. In addition to increased damping, which necessarily in-
gypsum board shear walls. Data are presented for blocked volves damage to elements in a structure, gypsum walls not detailed
to be uncoupled from the PLFRS may provide significant addi-
and unblocked assemblies, and the effects of intermediate (also
tional initial strength and stiffness to a structure.
referred to as field) fastener spacing, aspect ratios up to 2∶1, and
The beneficial strength and stiffness provided by gypsum walls
loading method (monotonic versus reversed cyclic test) are as secondary elements may be diminished as the seismic response
discussed. modification factor R decreases—a decrease in R implies a greater
ratio of strength provided by PLFRS elements to the strength avail-
1
Undergraduate Research Assistant, Dept. of Civil Engineering, Santa able from gypsum walls. For example, CFS framed structures that
Clara Univ., Santa Clara, CA 95053. utilize strap bracing (R ¼ 4:0 for bearing wall systems) for primary
2
Undergraduate Research Assistant, Dept. of Civil Engineering, Santa lateral resistance may not derive the same benefits in performance
Clara Univ., Santa Clara, CA 95053. as similarly framed structures with steel sheet or wood-panel
3
Associate Professor, Dept. of Civil Engineering, Santa Clara Univ., shear walls.
Santa Clara, CA 95053 (corresponding author). E-mail: rserrette@scu.edu
Where gypsum walls are used as the PLFRS, the available
Note. This manuscript was submitted on April 23, 2011; approved on
October 14, 2011; published online on June 15, 2012. Discussion period options for attachment of gypsum board in CFS framed structures
open until December 1, 2012; separate discussions must be submitted for are limited to the configurations listed in Table 1 for design in the
individual papers. This paper is part of the Journal of Structural Engineer- United States. These tabulated design values are applicable for
ing, Vol. 138, No. 7, July 1, 2012. ©ASCE, ISSN 0733-9445/2012/7-932– both seismic and wind design. In the United States, ASCE/SEI
941/$25.00. 7-10 limits the use of gypsum board shear walls to structures with

932 / JOURNAL OF STRUCTURAL ENGINEERING © ASCE / JULY 2012

J. Struct. Eng. 2012.138:932-941.

Downloaded from ascelibrary.org by New York University on 05/16/15. Copyright ASCE. For personal use only; all rights reserved.

Fig. 1. Linear elastic seismic demand, inelastic response, and overall system seismic performance factor relationships

Table 1. AISI S213-07 U.S. Nominal Shear Strengths for Gypsum Board wood-framed assemblies facilitates design that more closely aligns
Shear Walls with force demands. Such alignment can potentially produce
Nominal wall strength (kN∕m) for specified more economical and sustainable designs without sacrificing
fastener spacing at panel edges/field (mm∕mm) performance.
Comparing the values shown in Table 2, the following perfor-
Configuration 178∕178 102∕102 102∕305 203∕305
mance characteristics are evident:
12.7 mm gypsum 4.2 6.2 4.3 3.4 1. For unblocked walls, a reduction in wall stud spacing (corre-
board one side only sponds to an increase in numbers of fasteners used to attach the
Note: Maximum aspect ratio = 2∶1 (aspect ratio = ratio of wall height to wall sheathing) results in an increase in wall strength. This strength
length). All panel edges blocked. No. 6 drywall screw required. Fastener increase appears to be less significant for walls with closer
edge distance = 9.52 mm. Maximum stud spacing = 610 mm on center. supported panel edge fastener spacing;
Allowable stress design (ASD): Ω ¼ 2:0 (wind); Ω ¼ 2:5 (seismic). 2. Unblocked walls have capacities in the range of 80 to 86% that
Strength design: ϕ ¼ 0:65 (wind); ϕ ¼ 0:60 (seismic). of blocked walls; and
3. Data for the blocked 406 mm on center stud spacing walls sug-
gests a decrease in strength for walls with closer intermediate
a maximum height of 10.7 m in seismic design category (SDC) D fastener spacing. This decrease appears counterintuitive but
with seismic response modification factors R ¼ 2:0, C d ¼ 2:0, and may be the result of a direct use of test data.
Ωo ¼ 2:5. The relationship R ¼ C d suggests minimal secondary In contrast to the wood-framed wall data, the data in Table 1 for
component contribution to overall structural performance. The val- CFS framed walls shows (1) a sizeable increase in the nominal
ues in Table 1 apply to blocked walls. American Iron and Steel strength of blocked walls, with panel edge fastener spacing of
Institute (AISI) S213 (AISI 2009) requires a 65% reduction in 102 mm on center, as the intermediate fastener spacing is decreased
design strength for unblocked walls. from 305 to 102 mm on center; and (2) no strength increase for
In comparison with CFS framing, Table 2306.7 of the 2009 walls with closer framing spacing.
International Building Code (International Code Council (ICC),
Inc. 2009) includes several gypsum board shear wall assembly con- Test Program
figurations for lateral design of wood-framed structures. Some
of these configurations, specifically those that utilize 12.7-mm A total of 21 shear wall tests were conducted to investigate specific
gypsum board, are presented in Table 2. The greater range of aspects of gypsum board shear wall performance and provide some

Table 2. Allowable Shear Values for Gypsum-Sheathed Wood Framed Shear Walls
Allowable shear wall strength (kN∕m) for specified
fastener spacing at panel edges/field (mm∕mm)
Blocking
Configuration condition 178∕178 102∕102 102∕305 102∕406 152∕305 203∕305
12.7-mm gypsum board; one side only; studs at 406 mm on center Blocked 1.8 2.2 — 2.3 1.3 1.0
Unblocked 1.5 1.8 — — — 0.9
12.7-mm gypsum board; one side only; studs at 610 mm on center Blocked — — 2.3 — — —
Unblocked 1.1 1.6 — — — —
Note: Values acceptable when drywall screws are used. Reproduced from ICC 2009, Table 2306.7.

JOURNAL OF STRUCTURAL ENGINEERING © ASCE / JULY 2012 / 933

J. Struct. Eng. 2012.138:932-941.

 initial data to encourage an expansion of the present set of design reaction frame, as well as the top of the wall and the loading mecha-
values given in AISI S213. All walls utilized 12:7 × 1;220 × nism, to allow unrestrained movement of the gypsum panel relative
2;440 mm American Society for Testing and Materials Inter- to the CFS frame (that is, without bearing against the test frame)
national (ASTM) C1396 (ASTM 2009) regular core gypsum during testing. The top of the wall was attached to a modified 152 ×
wallboard (GWB) produced by Georgia Pacific (GP 2010). 152 × 3 mm hollow structural section (HSS) with large cut-outs in
The test walls were framed with Steel Stud Manufacturers one element to facilitate attachment of the wall to the loading
Association (SSMA 2010) 350S162-43 (227 MPa) stud and beam. The approximate moment of inertia and cross-sectional area
350T125-43 (227 MPa) track sections. All wall heights were of the modified HSS were 2:660 × 106 mm4 and 1;370 mm2 , re-
2,440 mm, and the wall width (out-to-out) was either 2,440 or spectively. Wall displacements were monitored using a series of
1,220 mm for 1∶1 and 2∶1 aspect ratio walls, respectively. The walls displacement transducers in accordance with the arrangement
were framed with studs at 610 mm on center, and back-to-back shown in Fig. 3.
studs were used at the wall ends (vertical boundaries/chords). As previously noted, both monotonic and reversed cyclic tests
The webs of the chord studs were connected by two lines of fas- were conducted in this test program. The loading rate in the mon-
teners. All framing connections were made with No. 8 × 13 mm otonic test was set at 0:50 mm∕s, and the reversed cyclic tests
Downloaded from ascelibrary.org by New York University on 05/16/15. Copyright ASCE. For personal use only; all rights reserved.

modified truss head (MTH) self-drilling screws. The overall fram- were performed using the test protocol defined in ASTM E2126
ing details for both the 1∶1 and 2∶1 walls are shown in Fig. 2. Method B (ASTM 2010a). The cycling rate was set constant at
For blocked walls, the gypsum panels were installed parallel to 10 mm∕s. For both monotonic and reversed cyclic loading, the
framing such that all edges were supported by either a stud or maximum wall displacement was set at 101.2 mm.
track flange. Panels were installed perpendicular to framing for
the unblocked walls—creating a midheight unblocked horizontal
joint. The fastener schedule for attachment of gypsum panels Test Results
was defined as se ∕si , where se = the panel edge fastener spacing,
and si = the panel intermediate or field fastener spacing. In all tests, Material Properties
No. 6 × 32 mm bugle head (BH) drywall drilling screws were
used to attach the gypsum board to the framing. At all panel edges, The material properties of the CFS framing members were deter-
the fasteners were installed at a constant 9.5-mm edge distance. mined in accordance with ASTM A370 (ASTM 2010b), and the
Details of each wall included in the test program are summarized results are summarized in Table 4. Ry and Ru in Table 4 define
in Table 3—in most cases, two walls of each configuration were the ratio of the measured yield and tensile strengths to their respec-
tested. tive minimum specified values. Both ratios are less than the values
recommended in AISI S213 for the steel used. The thickness values
reported in Table 4 are the base metal thickness and both values are
Test Setup within 0.92% of the designated thickness (1.09 mm).
The 12.7-mm gypsum panels used on the tests were purchased
The 27-mm gypsum walls included in this study were relatively at a local hardware store. The average dry mass of the gypsum
light and all walls were tested in a horizontal plane. The walls were panel was determined from 305 × 305 mm sections cut from the
installed in a reaction test frame and anchored to the base of the center of panel. The average mass density of the 12.7-mm gypsum
frame at each end with Simpson Strong-Tie (SST) S/HDU9 panel was 0:00743 g∕mm2 .
hold-downs (SST 2010) for overturning resistance and two or three
Shear Wall Response
19-mm-high strength (379-MPa minimum yield strength) shear an-
chor rods, as shown in Fig. 1. Aluminum plates, 90 mm wide and In general, wall failure (defined by the wall reaching its peak
19 mm thick, were installed between the bottom of the wall and the strength) in both the monotonic and reversed cyclic tests resulted

Fig. 2. Overall framing details

934 / JOURNAL OF STRUCTURAL ENGINEERING © ASCE / JULY 2012

J. Struct. Eng. 2012.138:932-941.

 Table 3. Wall Details
Test seriesa,b,f,g,h Test number Wall length (mm) Block conditionc Loading protocol Aspect ratio Fastener scheduled (mm∕mm)
GWB.4-4 A 2440 Blocked Monotonic 1∶1 102∕102
GWB.7-7 A 2440 Blocked Monotonic 1∶1 178∕178
GWB.4-6 A 2440 Blocked Monotonic 1∶1 102∕152
B 2440 Blocked Monotonic 1∶1 102∕152
GWB.4-6_S A 1220 Blocked Monotonic 2∶1 102∕152
B 1220 Blocked Monotonic 2∶1 102∕152
GWB.4-6_N A 2440 Unblocked Monotonic 1∶1 102∕152
B 2440 Unblocked Monotonic 1∶1 102∕152
GWB.4-6_C A 2440 Blocked Reversed cyclice 1∶1 102∕152
GWB.4-12 A 2440 Blocked Monotonic 1∶1 102∕305
Downloaded from ascelibrary.org by New York University on 05/16/15. Copyright ASCE. For personal use only; all rights reserved.

B 2440 Blocked Monotonic 1∶1 102∕305

GWB.4-12_S A 1220 Blocked Monotonic 2∶1 102∕305
B 1220 Blocked Monotonic 2∶1 102∕305
GWB.4-12_N A 2440 Unblocked Monotonic 1∶1 102∕305
B 2440 Unblocked Monotonic 1∶1 102∕305
GWB.4-12_C A 2440 Blocked Reversed cyclice 1∶1 102∕305
GWB.6-6 A 2440 Blocked Monotonic 1∶1 152∕152
B 2440 Blocked Monotonic 1∶1 152∕152
GWB.6-12 A 2440 Blocked Monotonic 1∶1 152∕305
B 2440 Blocked Monotonic 1∶1 152∕305
GWB.8-12 A 2440 Blocked Monotonic 1∶1 203∕305
a
All wall heights = 2,440 mm (out-to-out).
b
All walls constructed with 12.7-mm GWB.
c
Panels installed parallel to framing for blocked walls. Panels installed perpendicular to framing for unblocked walls.
d
mm∕mm = fastener panel edge spacing/intermediate (interior or field) fastener spacing.
e
Reversed cyclic testing in accordance with ASTM E2126-10, Method B.
f
Wall framing: SSMA 350S162-43 (227 MPa) stud sections and SSMA 350T125-43 (227 MPa) track sections.
g
See Fig. 1 for other wall details.
h
Specimen identification: [GWB.X-X][A∕B][S∕N∕C]. GWB: gypsum wallboard. X-X: panel edge fastener spacing/25.4—intermediate fastener spacing/25.4.
A∕B: test number in the series (A or B). S∕N∕C: wall configuration (S ¼ 2∶1 aspect ratio, or N = equals; unblocked, or C = reversed cyclic test).

Fig. 3. Displacement transducer layout

from a wedge-like breakout of the gypsum board core at the panel Table 5 summarizes peak response parameters for the tested
edges. Prior to breakout, fastener tilt and bearing deformation in the walls. Displacement values for two tests, GWB.4-12AN and
gypsum board was evident. The unblocked walls experienced sig- GWB.6-12B, are not included in Table 5 because of an initial in-
nificantly more damage at the fastener locations, particularly along operability of a transducer. The values listed in the table for the
the nonpapered edges. Fig. 4 shows typical damage/behavior at reversed cyclic tests (GWB: ×
X C) are based on envelope curves
panel edges. and an averaging of the “positive/push” and “negative/pull”

JOURNAL OF STRUCTURAL ENGINEERING © ASCE / JULY 2012 / 935

J. Struct. Eng. 2012.138:932-941.

 Table 4. CFS Stud and Track Properties
Yield strength Tensile strength
Section Average F y (MPa) Ry Average F u (MPa) Ru % elongation Thickness (mm)
350S162-43 294 1.29 355 1.14 39.9 1.10
350T125-43 295 1.30 358 1.15 37.4 1.08
Note: Tested in accordance with ASTM A370. Coupons parallel to the member axis. Average values based on three tests. % elongation measured over
a 50.8-mm gauge length.

Comparison of Recent Tests Results with Prior Work

Prior work at Santa Clara University (Nguyen et al. 1996; Morgan
Downloaded from ascelibrary.org by New York University on 05/16/15. Copyright ASCE. For personal use only; all rights reserved.

et al. 2002) facilitates a comparison with some of the tests in the

present test program. In 1996 Nguyen et al. tested 2; 440 ×
2; 440 mm blocked 12.7-mm gypsum walls with se ∕si fastener
spacing of 102 mm∕102 mm and 178 mm∕178 mm. The results
from these tests formed the basis of the values in Table 1.
GWB.4-4 and GWB.7-7 tests in the present test program are similar
to the tests by Nguyen et al. with the following exceptions:
(1) the present tests used 1.09-mm (43-mil) framing versus
0.84-mm (33-mil) in the earlier work, and (2) panels were attached
to one side of the wall and oriented parallel to framing in the present
work versus perpendicular to framing and on both sides of the wall
in the earlier work. In addition, blocked horizontal joints in the
earlier work used a 38.2-mm (33-mil) flat strap in accordance with
the requirements of AISI S213. Comparing average peak strength
values for gypsum on one side only, the peak strength of the tests in
the earlier work are 31–35% more that the values from the present
work. It is believed that this difference resulted primarily from the
blocking and gypsum panel orientation. Since only single tests
were conducted in the present program, additional testing should
be completed to verify the noted difference and, if necessary,
the appropriate values in Table 1 should be accordingly modified
to provide lower bound strengths.
The research work by Morgan et al. (2002) included monotonic
testing of 12.7-mm gypsum blocked and unblocked 2;440 ×
2;440 mm shear walls with se ∕si equal to 102 mm∕305 mm.
The blocked walls in the Morgan tests were constructed with panels
Fig. 4. Typical panel damage after peak strength parallel to framing. Similar tests, GWB.4-6 and GWB.4-6_N, were
included in the present test program with one difference—the
framing thickness. The prior tests utilized 0.84-mm (33-mil) des-
responses. To compare the overall response of tested walls, the ignated framing while the present tests utilized 1.09-mm (43-mil)
average values for paired tests were computed and normalized with framing. In CFS design, an increase in framing thickness tends to
respect to specific tests. The normalized values are shown in further restrict screw tilt and can result in an increase in shear wall
Table 6. Where two tests were conducted but paired values were strength with an associated decrease in peak strength displacement.
not available (tests GWB.4-12_N and GWB.6-12), the available Fig. 6 compares the measured responses for the blocked walls
data were used as an approximate of the average. and Fig. 7 compares the responses for the unblocked walls from
Comparing the peak strength displacements ΔVpeak , Fig. 5 Morgan et al. and the present tests. As shown in Fig. 6, the recently
shows that approximately 75% of the ΔVpeak values were in the tested walls exhibited higher peak capacities and marginally lower
20- to 32-mm range. The median ΔVpeak value was 26.4 mm. These initial stiffness than the walls in the previous work. The peak
values, as well as the load-displacements curves, indicate that the strength displacements of the 1.09-mm-thickness framed walls
peak strengths of the walls were all approached as the lateral were also marginally less than those of the 0.84-mm-thickness
displacement approached 25.4 mm. framed walls.
The scenario for the unblocked tests is quite different in terms of
peak strength. As shown in Fig. 7, the peak capacities for the walls
Discussion of Test Results with 1.09-mm-thickness framing were significantly higher than
those of the walls with 0.84-mm-thickness framing.
Although limited in overall scope, results from the test program These results appear to be intuitively correct given that thicker
show interesting and informative patterns and provide useful data studs tend to provide more restrain to screw tilt. The comparisons
for design consideration. The following sections discuss compar- also suggest that the effect of framing thickness may be more sig-
isons to prior work and the effects of intermediate fastener spacing, nificant in unblocked walls compared with blocked walls. More
blocking, aspect ratio, and loading pattern. data are needed to reliably quantify, in a general manner, the impact

936 / JOURNAL OF STRUCTURAL ENGINEERING © ASCE / JULY 2012

J. Struct. Eng. 2012.138:932-941.

 Table 5. Summary of Test Results
Test series Test number Aspect ratio V peak (kN∕m) ΔVpeak (mm) Average V peak (kN∕m) Average ΔVpeak (mm)
GWB.4-4 A 1∶1 4.6 43.6 4.6 43.6
GWB.7-7 A 1∶1 3.2 33.5 3.2 33.5
GWB.4-6 A 1∶1 4.4 25.3 4.5 28.4
B 1∶1 4.5 31.4
GWB.4-6_S A 2∶1 4.2 26.4 4.1 29.4
B 2∶1 4.0 32.5
GWB.4-6_N A 1∶1 2.7 42.6 2.6 39.5
B 1∶1 2.4 36.5
GWB.4-6_C A 1∶1 5.1 20.3 5.1 20.3
GWB.4-12 A 1∶1 4.8 23.4 4.7 23.3
Downloaded from ascelibrary.org by New York University on 05/16/15. Copyright ASCE. For personal use only; all rights reserved.

B 1∶1 4.5 23.3

GWB.4-12_S A 2∶1 4.6 26.4 4.2 26.4
B 2∶1 3.8 26.4
GWB.4-12_N Aa 1∶1 1.8 — 1.8 29.4
B 1∶1 1.8 29.4
GWB.4-12_C A 1∶1 4.5 20.8 4.5 20.8
GWB.6-6 A 1∶1 3.2 28.4 3.3 27.4
B 1∶1 3.3 26.3
GWB.6-12 A 1∶1 3.0 20.3 3.0 20.3
Ba 1∶1 3.0 —
GWB.8-12 A 1∶1 2.5 26.4 2.5 26.4
a
Displacement transducer error (initial displacement values not used).

Table 6. Normalized Average Test Results

Wall description Normalized values (R)
Test type Block condition Loading protocol Aspect ratio N Vpeak N Δvpeak
a,f
GWB.4-4 Blocked Monotonic 1∶1 1.00 1.00
GWB.4-6a Blocked Monotonic 1∶1 0.97 0.65
GWB.4-12a Blocked Monotonic 1∶1 1.01 0.53
GWB.4-6_Sa Blocked Monotonic 2∶1 0.89 0.67
GWB.4-12_Sa Blocked Monotonic 2∶1 0.91 0.60
GWB.4-6_Ca,f Blocked Reversed cyclic 1∶1 1.10 0.46
GWB.4-12_Ca,f Blocked Reversed cyclic 1∶1 0.98 0.48
GWB.4-6_Nc Unblocked Monotonic 1∶1 0.57 1.39
GWB.4-12_Nd,e Unblocked Monotonic 1∶1 0.40 1.26
GWB.7-7b,f Blocked Monotonic 1∶1 1.00 1.00
GWB.6-6b Blocked Monotonic 1∶1 1.00 0.82
GWB.6-12b,e Blocked Monotonic 1∶1 0.92 0.60
GWB.8-12b Blocked Monotonic 1∶1 0.77 0.79
a
Normalized with respect to test type (GWB.4-4).
b
One test conducted.
c
Normalized with respect to test type (GWB.4-6).
d
Normalized with respect to test type (GWB.4-12).
e
Two tests conducted but displacement values are taken from test.
f
Normalized with respect to test type (GWB.7-7).

of framing thickness on the strength and displacement of gypsum center panel edge fastener spacing (se ), respectively, and different
shear walls. intermediate fastener spacing (si ). For the 102 mm on center panel
edge fastener spacing, Fig. 8 shows that varying the intermediate
Effect of Intermediate Fastener Spacing fastener schedule from 102 to 152 to 305 mm on center had no
Figs. 8 and 9 compare the effect of intermediate fastener spacing for significant effect on the blocked wall performance up to the peak
the blocked 1∶1 aspect ratio walls with fixed 102 and 152 mm on strength. This trend is also evident in Table 6 where the normalized

JOURNAL OF STRUCTURAL ENGINEERING © ASCE / JULY 2012 / 937

J. Struct. Eng. 2012.138:932-941.

Downloaded from ascelibrary.org by New York University on 05/16/15. Copyright ASCE. For personal use only; all rights reserved.

Fig. 5. Distribution of peak strength displacements

Fig. 6. Comparison of prior GWB.4-6 blocked gypsum wall tests with

present tests
Fig. 8. Monotonic response of GWB.4-4, GWB.4-6, and GWB.4-12
blocked 1∶1 aspect ratio gypsum walls

Fig. 7. Comparison of prior work GWB.4-6_N (unblocked) gypsum

wall tests with present tests

Fig. 9. Monotonic response of GWB.6-6 and GWB.6-12 blocked 1∶1

strength ratios for the GWB.4-6 and GWB.4-12 test series with aspect ratio gypsum walls
respect to the GWB.4-4 test series range between 0.97 and 1.01.
Beyond the peak strength, the GWB.4-6 and GWB.4-12 walls ex-
hibited similar responses, but the GWB.4-4 wall exhibited a much 2:99 kN∕m for si ¼ 305 mm. Additionally, the results show that
greater toughness. the walls with the closer intermediate fastener spacing exhibited
As the fastener panel edge spacing se is increased from 102 to overall tougher responses.
152 mm on center, Fig. 9 shows an increase in strength for closer A comparison of the responses of the GWB.4-6 and GWB.4-12
spaced intermediate fasteners—3:3 kN∕m for si ¼ 152 mm versus unblocked 1∶1 aspect ratio walls, Fig. 10, shows a substantial

938 / JOURNAL OF STRUCTURAL ENGINEERING © ASCE / JULY 2012

J. Struct. Eng. 2012.138:932-941.

 Fig. 10. Monotonic response of GWB.4-6 and GWB.4-12 unblocked
1∶1 aspect ratio gypsum walls
Downloaded from ascelibrary.org by New York University on 05/16/15. Copyright ASCE. For personal use only; all rights reserved.

increase (approximately 40%) in peak strength resulting from the

denser (152 versus 305 mm) intermediate fastener spacing. This
increase may be because of the fact that unblocked walls have rel-
atively low stiffness and strength, thus, even a moderate change in
the number of fasteners at intermediate points tends to influence the
wall strength. These results, though based on limited testing,
suggest that intermediate fasteners can play an important role in
improving unblocked shear wall performance.
Collectively, comparing the overall responses of (1) the GWB.4-4, Fig. 11. Comparison of blocked (GWB.4-6 and GWB.4-12) and
GWB.4-6, and GWB.4-12 blocked walls; (2) the GWB.4-6 and unblocked (GWB.4-6_N and GWB.4-12_N) wall performance
GWB.4-12 unblocked walls; and (3) the GWB. 6-6 and GWB.6-12
blocked walls, the following preliminary observations can be made:
1. Unblocked and low strength blocked walls appear to derive
greater peak strength increases from closer intermediate fasten-
er spacing;
2. Peak strength increases for blocked walls with closer inter-
mediate fastener spacing may be a function of the panel edge
fastener spacing. As the edge fastener spacing is decreased,
there may be little to no benefit from practical decreases in
intermediate fastener spacing; and
3. Closer intermediate fastener spacing tends to increase the
postpeak strength toughness of both blocked and unblocked
gypsum walls.

Effect of Blocking
The beneficial effects of blocking are well recognized and reflected
in the present building standards. AISI S213 indicates that the
capacity of unblocked gypsum shear walls shall be taken as
35% of the blocked wall capacity. Results from this test program,
where panels were installed parallel to framing in blocked walls,
suggests that higher capacities (that is, greater than 35%) are pos-
sible for unblocked walls with closer intermediate fastener spacing.
Fig. 11 compares the response of unblocked shear walls
(GWB.4-6_N and GWB.4-12_N) with those of corresponding
blocked walls (GWB.4-6 and GWB.4-12), and numerical values
of this comparison are given in Table 6. The GWB.4-6 blocked
and unblocked test results show that the unblocked walls developed Fig. 12. Comparison of blocked (GWB.4-4 and GWB.7-7) and
peak strengths that far exceeded 35% (57 versus 35%) of the cor- unblocked (GWB.4-4_N and GWB.7-7_N) wall performance
responding blocked wall capacities. The GWB.4-12 blocked and
unblocked test results show that the unblocked walls developed
peak strengths closer to 35% (40 versus 35%) of the corresponding blocked wall tests with gypsum panels installed parallel to framing
blocked wall capacities. suggest unblocked wall strengths greater than 35% of the corre-
Including data from Nguyen et al. (1996) and Morgan et al. sponding blocked wall values. This difference in unblocked wall
(2002) with the present test data, comparisons of blocked and un- strength compared with blocked wall strength from the earlier work
blocked GWB.4-4 and GWB.7-7 gypsum shear wall behavior are and the present work may be attributed to the effect of a horizontal
shown in Fig. 12. The Nguyen and Morgan tests support the AISI joint in the blocked walls in the earlier tests.
S213 35% reduction in blocked wall strength to obtain unblocked Based on the data presented, it appears that reductions in
shear strengths. However, both the present GWB.4-4 and GWB.7-7 blocked wall strength to estimate unblocked wall strength may

JOURNAL OF STRUCTURAL ENGINEERING © ASCE / JULY 2012 / 939

J. Struct. Eng. 2012.138:932-941.

 be dependent on panel orientation and blocking details. As a lower
bound estimate of unblocked wall strength from blocked wall
(regardless of panel orientation) values, the recent test data supports
the 35% recommendation in AISI S213.
Effect of Aspect Ratio and Vertical Panel Joints
Present standards (AISI S213) adopt the position that the strength
and expected overall performance of walls with aspect ratios no
greater that 2∶1 will be similar or at least meet some minimum per-
formance expectation. Given this upper limit (2∶1) on aspect ratio
with no expected reduction in wall performance, testing is often
conducted using 2∶1 aspect ratio walls and a single structural panel.
Some industry groups and companies (ICC Evaluation Service, Inc.
2010) have expressed concern about a single panel test with regard
Downloaded from ascelibrary.org by New York University on 05/16/15. Copyright ASCE. For personal use only; all rights reserved.

to overall wall performance when a panel joint aligns over a

single stud flange. To address both the 2∶1 and less aspect ratio
assumption and vertical panel joint issue, Fig. 13 compares the re-
sponse of GWB.4-6 and GWB.4-12 blocked walls with 1∶1 aspect
ratio and a vertical panel joint with 2∶1 aspect ratio (no panel joint)
walls. The test results show no significant difference in overall
performance of the 1∶1 and 2∶1 walls, and that the panel joint,
aligned over a single stud flange, had no impact on strength or
performance.
Fig. 14. Comparison of monotonic and reverse cyclic envelope
Effect of Loading
response
Fig. 14 compares the monotonic responses of the GWB.4-6
and GWB.4-12 walls with their respective reverse cyclic test
counterparts—GWB.4-6_C and GWB.4-12_C. The response curves
for the reversed cyclic tests are the “push” and “pull” envelope roughly the same displacement in the GWB.4-12_C test (compared
curves derived from the hysteresis response. There is little to no dif- with the GWB.4-12 walls). These limited results suggest that walls
ference in initial response of the walls under monotonic and reversed with closer field fastener spacing may exhibit higher strength under
cyclic loading. However, the reversed cyclic test peak strength val-
reversed cyclic testing compared with monotonic testing and higher
ues were either higher (GWB.4-6_C compared with GWB.4-6) than
strength may be associated with lower peak strength displacements.
or very similar to (GWB.4-12_C compared with GWB.4-12) the
Thus, on the basis of these tests, it appears that monotonic testing
monotonic test values. Fig. 14 also shows that the reversed cyclic
loading peak wall strength occurred at a smaller displacement in may be sufficient to evaluate the performance of gypsum shear
the GWB.4-6_C test (compared with the GWB.4-6 walls) but at walls for both seismic and wind design—the present practice in
AISI S213.

Conclusion

This paper presents results from a full-scale monotonic and re-

versed cyclic test program of gypsum-sheathed CFS framed shear
walls. The test program was designed to investigate the effects of
blocking, intermediate fastener spacing, aspect ratio, and loading
pattern. The tested configurations also provide alternative means
of construction to present design standards that can potentially
facilitate more economical designs.
On the basis of the discussion presented, several inferences can
be made with regard to the lateral load performance of gypsum
shear walls:
1. Additional work is needed to determine the effect of panel
orientation (parallel or perpendicular to framing) in blocked
gypsum walls;
2. Depending on whether blocked wall strengths are determined
from tests with panels oriented parallel or perpendicular
to framing, the present recommendation that the capacity
of an unblocked gypsum shear walls be taken as 35% of
the corresponding blocked wall capacity appears to be
conservative;
3. The capacity of unblocked walls may be significantly en-
hanced by reducing the intermediate fastener spacing. The
Fig. 13. Comparison of blocked 1∶1 and 2∶1 aspect ratio walls
same strength benefit may not be available in blocked walls;

940 / JOURNAL OF STRUCTURAL ENGINEERING © ASCE / JULY 2012

J. Struct. Eng. 2012.138:932-941.

 4. Reduced intermediate fastener spacing generally enhances the References
postpeak response of a wall resulting in increased wall
toughness; AISI. (2009). North American standard for cold-formed steel framing—
5. Wall strength is not affected by aspect ratio—2∶1 versus 1∶1 or lateral design, S213-07, Washington, DC.
supported vertical panel joints over a single stud flange; ASCE. (2010). Minimum design loads for buildings and other structures,
6. Monotonic testing may be adequate to determine both seismic ASCE/SEI 7-10, Reston, VA.
and wind design strengths. Additional research is encour- ASTM. (2009). Standard specification for gypsum board, C1396, West
aged; and Conshohocken, PA.
7. Three unpublished alternatives gypsum shear wall ASTM. (2010a). Standard test methods for cyclic (reversed) load test for
configurations—GWB.4-6, GWB.6-12, and GWB.6-6— shear resistance of vertical elements of the lateral force resisting
systems for buildings, E2126, West Conshohocken, PA.
were introduced.
ASTM. (2010b). Standard test methods and definitions for mechanical test-
ing of steel products, A370, West Conshohocken, PA.
Acknowledgments Consortium of Universities for Research in Earthquake Engineering
Downloaded from ascelibrary.org by New York University on 05/16/15. Copyright ASCE. For personal use only; all rights reserved.

(CUREE). (2004a). Recommendations for earthquake resistance in

This research program was made possible by a generous grant from design and construction of woodframe buildings—Part I: Recommen-
Santa Clara University’s School of Engineering and material dona- dations, CUREE Publication No. W-30a, Richmond, CA.
tions from several companies, including California Expanded Consortium of Universities for Research in Earthquake Engineering
Metal Company (CEMCO), Grabber Construction Products, and (CUREE). (2004b). Recommendations for earthquake resistance in
the Simpson Strong-Tie Company. The writers would like to thank design and construction of woodframe buildings—Part I: Topical
discussion, CUREE Publication No. W-30b, Richmond, CA.
Matt Eiler, P.E. of Anderson-Peyton Structural Engineers, for his
FEMA. (2009a). Quantification of building seismic performance factors,
comments on the paper.
P695, Washington, DC.
FEMA. (2009b). NEHRP recommended seismic provisions for new build-
Notation ings and other structures—Part 2: Commentary to ASCE/SEI 7-05,
P-7502, Washington, DC.
The following symbols are used in this paper: Georgia-Pacific Gypsum (GP). (2010). “Toughrock specification sheet.”
Cd = seismic deflection amplification factor (ASCE/SEI Atlanta, GA, 〈www.gpgypsum.com〉.
7-10); ICC Evaluation Service, Inc. (2010). “Proposed revisions to the acceptance
F u = tensile strength of cold-formed steel framing; criteria for power-driven pins for shear wall assemblies with
F y = yield strength of cold-formed steel framing; cold-formed steel framing and wood structural panels.” AC 230,
N Vpeak = normalized measured peak wall strength; Birmingham, AL, 〈www.icc-es.org/Criteria_Development/1010-pre/
Responses/14_AC230.pdf〉.
N ΔVpeak = normalized measured deflection at peak wall strength;
International Code Council, Inc. (ICC). (2009). 2009 international building
R = seismic response modification factor (ASCE/SEI
code (IBC), Country Club Hills, IL.
7-10);
Morgan, K., et al. (2002). “Performance of cold-formed steel-framed shear
Ru = ratio of average measure tensile strength to specified walls: Alternative configurations.” Final Report: LGSRG-06-02, Dept.
minimum tensile strength; of Civil Engineering, Santa Clara Univ., Santa Clara, CA.
Ry = ratio of average measure yield strength to specified Nguyen, H., et al. (1996). “Shear wall values for light weight steel
minimum yield strength; framing.” Report No. LGSRG-3-96, Dept. of Civil Engineering, Santa
se = panel edge (boundary) fastener spacing; Clara Univ., Santa Clara, CA.
si = panel intermediate fastener spacing; Simpson Strong-Tie Company (SST). (2010). “Cold-formed steel connec-
V peak = peak wall strength; tors.” C-CFS10, Pleasanton, CA, 〈www.strongtie.com〉.
ΔVpeak = deflection at peak wall strength; and Steel Studs Manufacturers Association (SSMA). (2010). “Product technical
Ωo = seismic overstrength factor (ASCE/SEI 7-10). information.” Chicago, 〈www.ssma.com〉.

JOURNAL OF STRUCTURAL ENGINEERING © ASCE / JULY 2012 / 941

J. Struct. Eng. 2012.138:932-941.