Thin-Walled Structures 92 (2015) 146–159

Contents lists available at ScienceDirect

Thin-Walled Structures
journal homepage: www.elsevier.com/locate/tws

Cyclic performance of cold-formed steel shear walls sheathed

with double-layer wallboards on both sides
Jihong Ye n, Xingxing Wang, Hongyuan Jia, Mengyuan Zhao
Key Laboratory of Concrete and Prestressed Concrete Structures of the Ministry of Education, Southeast University, Nanjing, China

art ic l e i nf o a b s t r a c t

Article history: To satisfy the requirements of fire resistance and loading capacity of the walls in multi-story cold-formed
Received 31 March 2014 steel (CFS) structures, shear walls sheathed with double-layer wallboards on both sides were proposed.
Received in revised form Sheathing materials in these walls included gypsum wallboard (GWB), bolivian magnesium board (BMG)
15 January 2015
and calcium silicate board (CSB). Cyclic loading tests on six full-scale walls of this configuration were
Accepted 3 March 2015
Available online 21 March 2015
conducted, from which the shear performance of the walls could be obtained. Factors such as the
sheathing material, aspect ratio, stud section and stud spacing were considered. Another experimental
Keywords: study on the shear behavior of the screw connections was also performed to explore the potential
Cold-formed steel shear wall relationship between the walls and the screw connections in shear performance. The results showed
Cyclic test
that the peak strength of the walls sheathed with bolivian magnesium boards as the face layer
Shear strength
wallboards significantly exceeded the nominal value of the current standard. However, for the walls
Equivalent bracing model
Lateral stiffness sheathed with calcium silicate boards as the face layer wallboards, the tested walls exhibited brittleness
Drift angle limit damage with poor ductility after the peak strength. The equivalent-bracing model was used to calculate
the lateral stiffness of the walls, based on which a series of screw connection deformation limits and
shear-wall drift angle limits was suggested.
& 2015 Elsevier Ltd. All rights reserved.

1. Introduction However, there was no obvious difference in the shear performance

Cold-formed steel (CFS) shear walls have been widely used in
residential and small commercial buildings in the USA, Japan,
Australia, and Europe over the years because of their light weight,
easy installation, and other advantages, such as environmental
characteristics and recyclability [1–3]. In recent years, a growing
number of multi-story buildings have been framed with CFS wall
systems as the load-bearing structural components [4].
Because of the complex configuration of CFS shear walls, numer-
ous experimental studies have been conducted to determine their
shear performance. Studies on the influence of the sheathing
material were performed by Fülöp and Dubina [1], Tarpy and Girard
[5], Miller and Peköz [6], and Serrette et al. [7]. Lin et al. [2], Tarpy
and Girard [5], and Pan and Shan [8] conducted experimental studies
on CFS shear walls concerning the influence of the stud type and stud
spacing. These experimental investigations indicated that using
thicker and back-to-back end studs can prevent the end studs from
distorting before the wallboards reach their full shear strength.
n
Correspondence to: School of Civil Engineering, Southeast University, Nanjing
210096, China. þ86 2583795023.
E-mail address: yejihong@seu.edu.cn (J. Ye).
of the CFS shear walls when the stud spacing was decreased. Studies
performed by Shakibanasab et al. [3], Tarpy and Girard [5], Landolfo
et al. [9], and Yu [10] showed that the influence degree of the aspect
ratio depends on the ratio itself. In addition, the effect of the loading
mode was also studied by Fülöp and Dubina [1], Lin et al. [2],
Landolfo et al. [9], Balh et al. [11], and Nithyadharan and Kalyanara-
man [12]. These studies focused on the walls sheathed with single-
layer wallboards on one or two sides, which mainly applied to low-
rise residential buildings because of their low shear capacity. For
multi-story buildings, the loading capacity of the walls and the fire
resistance of the sheathings are more important. Experiments
performed by Fiorino et al. [13] and Nithyadharan and Kalyanaraman
[14] confirmed that the strength and stiffness of the screw connec-
tions between the CFS framing members and the sheathings govern
the CFS shear-wall behavior, and the failure of the screw connections
was finally reflected in the bearing failure of the wallboards. Conse-
quently, it appears notably important to improve the shear capacity of
CFS shear walls by using panels with higher strength, such as
wallboards, and to consider the fire resistance. Gypsum wallboard is
a common anti-fire sheathing for internal wallboards. However, the
contribution of gypsum wallboards to the shear capacity of CFS shear
walls is small because of its low strength [5–7]. Bolivian magnesium
board (BMG) and calcium silicate board (CSB) were adopted in this

http://dx.doi.org/10.1016/j.tws.2015.03.005
0263-8231/& 2015 Elsevier Ltd. All rights reserved.
 J. Ye et al. / Thin-Walled Structures 92 (2015) 146–159
paper because of their good fire resistance performance and high
strength [4]. Both types of boards were combined with gypsum
wallboards in the form of double-layer wallboards to sheath both
sides of the walls.
Specific properties of CFS shear walls, such as the shear capacity
and the lateral stiffness, should be determined to provide the basis
for the seismic design of such structures. Nominal shear strength for
wind and other in-plane loads for the shear walls were provided in
AISI-standard tables [15]. The available strength of the material of
the same capacity is cumulative for the walls with material of the
same type and nominal strength applied to opposite faces of the
same wall, and never more than 30% of the total demand on lateral
capacity shall be assumed to be provided by gypsum wallboards.
The equivalent-bracing model was proposed in the Japanese stan-
dard [16], which can be used to evaluate the lateral stiffness of the
CFS shear walls. However, these specifications and standards are
mainly for walls sheathed with single-layer wallboards on one or
two sides. For the walls sheathed with double-layer wallboards on
both sides, the applicability of the described methods in these
specifications and standards should be further investigated.
The PBSD (performance based seismic design) approach is based
on the coupling of multiple performance levels and ground motion
intensities, which generate performance objectives to be satisfied [17].
Different performance levels were defined in the USA, Europe, Japan,
and China: (1) IBC [18] adopted single-level performance criteria, and
two ground motion intensities (“design earthquake” and “maximum
considered earthquake”) were considered to satisfy the “no collapse”
requirement; (2) two-level performance criteria (“no collapse” and
“damage limitation”) under different ground motion intensities were
adopted in EC8 [19]; (3) BSL [20] adopted two-level performance
criteria, which can be concluded as “no damage in moderate earth-
quake” and “no collapse in severe earthquake”; and (4) three-level
Horizontal joint Vertical joints
10
540
300
150
3000
2440
10
14mm diameter bolt
3600 (4800)
Face layer
Base layer
Base layer
Face layer
Fig. 1. Details of specimen configurations.
147
performance criteria were proposed in GB 50011-2010 [21], which
were concordant with those used in Europe and Japan, and can
be concluded as “no damage in minor earthquake”, “repairability
in moderate earthquake” and “no collapse in severe earthquake”.
Although the application of the performance-based methodology has
long been verified for ordinary reinforced concrete and steel struc-
tures, its application to CFS steel construction remains largely
unexplored [9]. Consequently, it is necessary to divide different
damage levels of CFS shear walls under different ground motion
intensities to realize the PBSD in CFS structures.
This paper presents the test results of six full-scale shear walls
sheathed with double-layer wallboards on both sides. Moreover,
another experimental study on the shear behavior of screw
connections under monotonic tensile tests was performed to build
the relationship between the walls and the screw connections in
shear performance. Details of the specimens, test procedures, and
test results are presented. Based on the test results and consider-
ing the intimate relationship between the walls and the screw
connections, methods to evaluate the shear capacity and the
lateral stiffness of the walls are presented. Finally, a series of
screw connection deformation limits and shear-wall drift angle
limits is suggested based on the three-level performance criteria,
which designers could reference and benefit from in practice.
2. Experimental program
2.1. Test specimens
The experimental program was based on six full-scale wall tests
with different assemblies. The six configurations differed on the
sheathing material, aspect ratio, stud section and stud spacing. Fig. 1
Interior stud Top track
(C89/C140/double C89) (U91/U142)
Channel section
blocks
Base layer
Base layer
Face layer
End stud
(double C89/
double C140)
Bottom track
18mm diameter
bolt
screw stud
148 J. Ye et al. / Thin-Walled Structures 92 (2015) 146–159

Table 1
Test specimens.

Specimen number Section type of CFS members (mm) Sheathing material1 Stud spacing (mm) Fastener spacing2

Interior studs End studs Tracks

Specimen 1 C89 89  50  13  0.9 Double C89a U91 91  50  0.9 GWBc/CSBd 600 150/300
Specimen 2 C89 89  50  13  0.9 double C89a U91 91  50  0.9 GWBc/CSBd 400 150/300
Specimen 3 C89 89  50  13  0.9 Double C89a U91 91  50  0.9 GWBc/CSBd 600 150/300
Specimen 4 double C89a Double C89a U91 91  50  0.9 GWBc/BMGe 600 150/300
Specimen 5 C140 140  50  13  1.2 double C140b U142 142  50  1.2 GWBc/BMGe 600 mm 150/300
Specimen 6 C140 140  50  13  1.2 Double C140b U142 142  50  1.2 GWBc/BMGe 400 150/300

Note:
1
Base layer/face layer.
2
Perimeter/field, (mm/mm).
a, b
Coupled C sections consisted of two stud sections connected back-to-back using self-drilling wafer head screws spaced at 150 mm along the vertical.
c
12 mm gypsum wallboard.
d
12 mm calcium silicate board.
e
12 mm bolivian magnesium board.

Fig. 2. Experimental arrangement.

 J. Ye et al. / Thin-Walled Structures 92 (2015) 146–159
Fig. 3. Measuring-point arrangement.
shows the configuration details of the specimens. All specimens
were assembled with a rectangular geometry 3600 mm wide and
3000 mm high, except specimen 3, which was designed to provide
a different aspect ratio with a rectangular geometry 4800 mm wide
and 3000 mm high. Table 1 summarizes the different test speci-
mens in the program.
The steel frames included lipped channel section studs and
plain channel section tracks. Both studs and tracks were fabricated
from a galvanized steel sheet with a minimum yield strength of
345 MPa. The steel frames were assembled with 4.8 mm diameter
and 19 mm long self-drilling wafer head screws.
The specimens were sheathed with 3000  1200  12 mm3 gyp-
sum wallboards combined with either 2440  1220  12 mm3 bolivian
magnesium boards or 2440  1220  12 mm3 calcium silicate boards,
all of which were attached to the frames using 4.8 mm diameter and
45 mm long bugle head drywall screws as shown in Fig. 1. Because of
the different sizes of these sheathings, all specimens possessed vertical
and horizontal joints (Fig. 1a). For the horizontal joints, channel section
blocks were mounted inside of the frame at the corresponding
locations by screws. To observe the failure mode of the sheathings at
the joints, gypsum wallboards were installed as the base layer wall-
boards, whereas bolivian magnesium boards and calcium silicate
boards were installed as the face layer wallboards (Fig. 1b). For all
specimens, both the base layer wallboards and the face layer wall-
boards were vertically spaced. An edge distance of 20 mm was
maintained in these fields like specimen edges and horizontal joints;
another edge distance of 15 mm was maintained at the vertical joints
because of the narrow stud flange. In addition, a gap of 10 mm was left
between the frame and the sheathings to avoid loading on the
wallboards. All specimens had identical sheathing setups on both sides.
2.2. Test setup
Fig. 2 shows the overall test setup. Beam 1 was fixed on the floor
to simulate the rigid foundation. The bottom track of each specimen
was connected to beam 1 with 14 mm diameter shear bolts at the
center position between adjacent studs and 18 mm diameter anchor
rods at the hold-down locations. The top tracks of specimens 1, 2, 4,
5 and 6 were bolted to 3800 mm-long load beam 2, whereas the top
track of specimen 3 was bolted to 5000 mm-long load beam 3 with
14 mm diameter bolts. Both the vertical load, which was developed
by two hydraulic jacks, and the horizontal load, which was developed
149
by a hydraulic actuator, could be transmitted to the specimens via
beam 2 or beam 3. Two hydraulic jacks were embedded in a slide
rail to maintain the invariant vertical load position during the test
procedure (Fig. 2a). The hydraulic actuator provided the force
measurement, whereas the local displacements were measured
using other transducers, as later described. The specimens were
restrained against lateral displacement at two points on the upper
part, which acted as the sliding restraint (Fig. 2b).
The overall specimen response, such as lateral displacement, slip
and uplift, was measured and recorded using a series of displacement
transducers. Fig. 3 shows the details of the measuring-point arrange-
ment. D1 was used to measure the lateral displacement of the load
beam (beam 2 or beam 3); D2 was used to measure the lateral
displacement of the specimen on top; D3 and D4 were used to
measure the slip displacement between the specimen and
beam 1; D5 and D6 were used to measure the vertical displacement
of the specimen relative to beam 1; and D7 and D8 were used to
measure the vertical displacement of beam 1 relative to the founda-
tion. Moreover, the displacement of the hydraulic actuator was also
recorded.
2.3. Test procedure
The specimens were tested under cyclic load in addition to the
vertical load. First, an axial compression load was gradually
applied to all studs at a constant rate using two hydraulic jacks
(Fig. 2a). For the end studs and the interior studs, the applied loads
are 10 kN and 20 kN per stud, respectively, according to the
calculated results of a ground wall with a 600 mm stud spacing
of a representative three-story residence in Beijing, China, the
commonly size of which is 4800 mm wide of depth, 3600 mm
wide of the room and 3000 mm high of the story height. The
values of the applied loads are the standard values of the vertical
load for the studs, including the dead loads derived according to
the common residential construction and the live loads based on
the current load code. Both applied load values were kept constant
during the test. The tests were conducted in force-control mode
and displacement-control mode. Each specimen was tested under
stepped loading with a constant cyclic frequency of f ¼0.03 Hz. The
load capacity of each specimen was estimated before the experi-
ment according to previous experimental results and experience
at home and abroad. During the test, the force-control mode was
150 J. Ye et al. / Thin-Walled Structures 92 (2015) 146–159

Fig. 4. Failure modes of the wall specimens. (a) Tilting (T), (b) Pull-through (P), (c) Being sheared off (S), (d) Distortion (D) Cracks (F1) Strip (F2) Bulging (F3) Shedding (F4)
Crushing (F5), and (e) Sheathing failures.

replaced by the displacement-control mode when a turning point 3. Test results

of the load–displacement curve appeared. The relative displace-
ment that corresponded to the turning point was defined as the
elastic limit displacement Δel of the specimen. The displacement-
control mode followed the ECCS Recommendation [22], which
consisted of cycles of 1Δel, 2Δel, 3Δel… until failure or a significant
decrease of the load-bearing capacity occured.
3.1. Failure modes
Different observed failure modes from the cyclic tests can be
summarized as follows: (1) tilting of screws (T) (Fig. 4a);
(2) screws were pulled through the wallboard (P) (Fig. 4b);
 J. Ye et al. / Thin-Walled Structures 92 (2015) 146–159
The actual shear The measured
displacement displacement
Fig. 5. Shear–displacement model of CFS shear walls.
H
L+B+C
Fig. 6. Overturning displacement.
(3) screws were sheared off (S) (Fig. 4c); (4) stud distortion (D)
(Fig. 4d); and (5) sheathing failures: (i) cracks (F1), (ii) strip
deformation (F2), (iii) bulging (F3), (iv) shedding (F4), and
(v) crushing (F5) (Fig. 4e).
In the tests of specimens 1–3, the failure modes consisted of screws
being sheared off, stud distortion in the steel-frame failure (Sþ D), and
cracks, bulging, shedding, and crushing (F1þ F3þF4þ F5) in the
sheathing failure. In the tests of specimens 4–6, the failure modes
consisted of screw tilting, screws being pulled through the sheathing,
stud distortion in the steel frame failure (TþPþ D), and cracks, strip
deformation, bulging, shedding, and crushing (F1þ F2þF3þ F4þF5)
in the sheathing failure. It is worth emphasizing that all screw
connection deformations were observed either from the joints or
the specimen edges, and no obvious screw connection deformation
was found on the interior studs.
For specimens 1–3: at the beginning, the wallboards cracked at
the diagonal corners because of the relatively large internal forces
of the screw connections in these fields. When the load increased,
the wallboards below the horizontal joints had “rigid-body” rota-
tion; vertical cracks appeared on the wallboards above the hor-
izontal joints as a result of the squeezing action between the above
and below wallboards. When the load continued to increase, the
screws gradually sheared off, and the end stud bases distorted
because of the loss of the sheathing restraint, which was caused by
the failure of screw connections in these fields. When most of the
screws at the vertical joints were sheared off, the face layer
wallboards started bulging outward, and cracks appeared and
increased on the wallboards below the horizontal joints. When
all of the screws at the vertical joints were sheared off, the base
layer wallboards also started bulging outward, cracks appeared on
both the face layer wallboards and the base layer wallboards, and
bits of the face layer wallboards dropped from the frame. Finally, all
screws were sheared off, the wallboards were crushed at the
bottom of the specimen, and the specimen was damaged. After
the sheathing demolition, distortions were observed at the top and
bottom of the interior studs, but the deformation of screw holes on
both the face layer wallboards and the base layer wallboards was
151
Overturning
Slip displacement displacement
negligible. Compared to specimens 1 and 3, the steel-frame deforma-
tion of specimen 2 was smaller.
For specimens 4–6: the screws were primarily tilted and
sunken in the wallboards. When the load increased, the cracking
and the rotating mode of the wallboards formed and developed
identically to those of specimens 1–3. When the load continued
increasing, the screws gradually pulled through the face layer
wallboards because the screw holes slacked, and the end stud
bases distorted because of the same reason as that described for
specimens 1–3. When most screws at the vertical joints pulled
through the face layer wallboards, the face layer wallboards lost
the restraints and started bulging outward. When all screws at the
vertical joints pulled through the face layer wallboards, bits of the
face layer wallboards broke off because the cracks deepened and
dropped from the frame. Finally, all screws pulled through the face
layer wallboards, the wallboards were crushed at the bottom of
the specimen, and the specimen was damaged. After the sheathing
demolition, distortion at the top and bottom of the interior studs
was observed, and the screw holes on the base layer wallboards
had strip deformation. The steel frame deformation of specimens
4–6 decreased in sequence.
3.2. Load–displacement behavior
The measured displacement at the top of CFS shear walls
consisted of slip displacement, overturning displacement and
actual shear displacement. The expression to estimate the actual
shear displacement of CFS shear walls is summarized in (1)–(5).
Four terms in (1) define a relatively simple model (illustrated in
Fig. 5) for the behavior of CFS shear walls.
Δ ¼ Δ0  Δ1  Δφ ð1Þ
ðH=ðH AÞÞ UR2 þ R1
Δ0 ¼ ð2Þ
2
Δ1 ¼ R3  R4 ð3Þ
H
Δφ ¼ U Δa ð4Þ
LþBþC
Δa ¼ ðR6 R8 Þ  ðR5  R7 Þ ð5Þ
where Δ is the actual shear displacement of the CFS shear walls, Δ0
is the measured displacement at the top of the CFS shear walls, Δ1 is
the slip displacement of the CFS shear walls relative to the
foundation, Δφ is the overturning displacement (illustrated in
Fig. 6), H is the wall height, L is the wall length, A is the distance
between displacement transducers D1 and D2, B and C are the
horizontal distances between displacement transducers D5, D6 and
the specimen edges, respectively, and R1–R8 are measured values of
displacement transducers D1–D8, respectively.
Based on the above analysis and calculation, the load–displace-
ment curve of the specimen was obtained using the actual shear
displacement and the corresponding load, and the envelope curve
152 J. Ye et al. / Thin-Walled Structures 92 (2015) 146–159

formed by the peak points of each first load step circle was defined
as the skeleton load–displacement curve of the specimen.
Fig. 7 shows the typical load–displacement curves and the
corresponding envelope curves of specimen 1 (Fig. 7a) and speci-
men 5 (Fig. 7b) under cyclic load, respectively. During the initiation
of loading, both specimens 1 and 5 were in the elastic range, and
the curves appeared linear. When the load increased, both speci-
mens entered the elastic–plastic deformation phase because cracks
appeared and the screws tilted, and the curves developed from
linear to fusi-form. When the load reached a certain value, the
cracks deepened, and the end stud bases distorted. Meanwhile, the
screw connections of specimen 5 appeared to have larger deforma-
tion, which caused obvious turning points of the curves that
implied that the specimens yielded, and the curves developed from
fusi-form to bow-form. After the specimens yielded, the pinching of
the curves became more apparent because of the opening and
closing of the screw holes, and the curves appeared completely
bow-form. It is notably important to distinguish that for specimen 1,
when the load reached the maximum value, there was no obvious
variation in the area of the hysteretic curve in successive cycles of
the same amplitude, and specimen 1 lost its bearing capacity as
soon as the screws were sheared off and the brittleness damage
appeared. For specimen 5, when the load reached the maximum
value, the shear strength and the area of the hysteretic curve
gradually decreased in successive cycles of the same amplitude
because the screws gradually pulled through the face layer wall-
boards, which indicated the strength degradation of the specimen;
the no-load slipping of the screw connections led to a Z-form curve.
In other words, the hysteretic behavior of specimen 5 was char-
acterized by notably significant pinching and strength degradation.
specimens are indicated in Fig. 9, where Fp is the peak strength,
Δp is the relative displacement that corresponded to Fp, Fs is the
unit shear strength, Fe ¼0.4Fp is the conventional elastic strength
limit, Δe is the relative displacement that corresponded to Fe, Fy is
the yield strength, Δy is the relative displacement that corre-
sponded to Fy, Δu is the relative displacement that corresponded to
0.85Fp beyond the peak load and μ ¼ Δu/Δy is the ductility
coefficient.
Two methods were introduced to determine the yield limit of the
specimens. In Method I, the conventional yield limit was evaluated
based on the equivalent elasto-plastic energy absorption [23]. The
secant line was chosen so that the hatched parts in Fig. 10(a) had
identical areas. The relative displacement that corresponded to
the intersection point of the secant line to a line of Fp, tangent to
the experimental curve, was defined as the yield displacement (Δy-I),
and the relative load that corresponded to the yield displacement
(Δy-I) was defined as the yield strength (Fy-I). In Method II, which
was proposed by Han [24], the point of load that corresponded to
0.6Fp was defined as the elasticity limit, the relative displacement
that corresponded to the intersection point of the secant line
through the elasticity limit to a line of Fp, tangent to the experi-
mental curve, was defined as the yield displacement (Δy-II), and the
relative load that corresponded to the yield displacement (Δy-II) was
defined as the yield strength (Fy-II).
1.2
1.0
0.8
0.6

Fig. 8 summarizes the ratio of the measured envelope specimen 0.4

strength to the peak specimen strength against the displacement ratio 0.2
of the specimen (the ratio of the top of the specimen displacement to
F/Fp

0.0
the specimen height) for each specimen tested. The curve becomes
-0.2
non-linear at approximately 40–50% of the peak load. After the peak Specimen 1
load was attained, for specimens 1–3, there was a short or no descent -0.4 Specimen 2
part of the curve, which indicated brittleness damage; on the contrary, -0.6 Specimen 3
Specimen 4
for specimens 4–6, the carried load gradually reduced with the increase -0.8 Specimen 5
in displacement, exhibiting considerable ductility, before failure. Specimen 6
-1.0

-1.2
3.3. Analysis of the test results
-0.03 -0.02 -0.01 0.00 0.01 0.02 0.03

Δ/H
The parameters (Fp, Fs, Fe, Δe, Fy, Δy and Δu) that can be used to
characterize the typical load–displacement behavior of the Fig. 8. Envelope curves of the specimens.

125 150

125
100
100
75
75
50
50
Force (kN)

Force (kN)

25
25
0
0
-25
-25
-50
-50
-75 -75

-100 -100

-125 -125
-60 -40 -20 0 20 40 60 -175 -150 -125 -100 -75 -50 -25 0 25 50 75 100 125 150 175

Displacement (mm) Displacement (mm)

Fig. 7. Typical load–displacement curves of the specimens. (a) Specimen 1 and (b) Specimen 5.
 J. Ye et al. / Thin-Walled Structures 92 (2015) 146–159
Table 2 summarizes the experimental results of the specimens,
where the average value in two opposite directions is adopted
based on the first envelope curve and μI and μII are the ductility
coefficients of the specimens based on Δy-I and Δy-II, respectively.
As shown in Table 2, the yield strength and the peak strength of
specimen 1 are higher than those of specimen 4. By comparing
specimen 1 with specimen 3, no improvement is observed on the
unit shear strength by increasing the wall length. The comparison
results between specimens 5 and 6 indicate that specimens with
smaller stud spacing have no obvious improvement in the peak
strength and only a slight improvement in the yield strength.
Comparison of the results between specimens 4 and 5 shows that
the strength of shear walls with a similar steel consumption can be
considerably improved by properly changing the stud section type.
Furthermore, specimen 3 has no effective ductility coefficient
because of the sudden destruction after the peak load, and the
ductility coefficient of specimens 1 and 2 are lower than those of
specimens 4–6.
4. Screw connection test
A series of screw connection tests was performed to investigate
the shear behavior of the screw connections between the steel frame
and the sheathings. The screw connection specimens were prepared
using steel plates and boards with identical material properties as the
steel frame and the sheathings of the wall specimens. A typical screw
connection specimen is shown in Fig. 11; the wallboards were cut
into blocks of 200 mm (l)  120 mm (w), which were attached to
steel plates using 4.8 mm diameter screws. Thirty sets of specimens
were tested, and an outline description of the tests is provided in
Tables 3 and 4.
In addition, two other sets of specimens with double-layer boards
were designed to reveal the relationship between the walls and the
Fig. 9. Parameters of the load–displacement curves.
Fig. 10. Alternative definitions of the yield limit for cyclic loading. (a) Method I and (b) Method II.
153
screw connections in the failure mode. Because the test data were
not sufficient to quantitatively analyze the shear performance of
screw connections with double-layer boards, only the failure modes
of these specimens were described in this paper.
All tests were finished by performing displacement-controlled
tensile tests at a loading velocity of 0.03 mm/s. At least two
identical specimens were tested for each set of specimen to
determine the behavior. The mean value was used if the strength
and displacement parameters of the two specimens were within
15% of each other. Otherwise, the mean values were based on the
results of at least three identical specimens.
For the gypsum wallboard and the bolivian magnesium board:
the screw connections experienced tilting of the screw (Fig. 12a),
bearing against the board (Fig. 12b) and edge-tearing failure of the
board (Fig. 12c); these screw connections actually failed because of
the compression failure of boards without deformation in screws. For
the calcium silicate board: the screw connections failed because of
the mutually perpendicular cracks surrounding the screw (Fig. 12d).
As shown in Fig. 13, the failure modes of the screw connections with
double-layer boards were notably similar to those of the correspond-
ing wall specimens. Two aspects can be concluded: the screws were
sheared off (Fig. 13a) with negligible deformation of the screw holes
on the gypsum wallboard and the calcium silicate board (Fig. 13b);
the screw hole slacked on the face layer board (Fig. 13c) with the
strip deformation of the screw hole on the base layer board (Fig. 13d),
both of which were exactly the same regardless of the type of board
(gypsum wallboard or bolivian magnesium board) used as the face
layer board (Fig. 13c and d).
Typical load–deformation curves of the screw connections are
illustrated in Fig. 14: for the gypsum wallboard and the bolivian
magnesium board, the curves of the screw connections have
similar trends (Fig. 14a); for the calcium silicate board, the screw
connections abruptly disrupted (Fig. 14b). The test results are
summarized in Tables 3 and 4, where only specimens with loading
direction parallel to the entire wallboard width are listed because
of the conservative results in this direction. The parameters (Pp, dp,
Pe, de, Py and dy) that can be used to characterize the typical load–
deformation behavior of the screw connections were provided
using two methods: for the gypsum wallboard and the bolivian
magnesium board, the parameters were determined using the
same methods as in Fig. 10; for the calcium silicate board, Pαp ¼ αPp
(α o1) was presented to replace Py because the screw connections
abruptly disrupted, dαp is the deformation corresponding to Pαp as
shown in Fig. 15, and the other four parameters were determined
using the same methods as those used in screw connections for
the gypsum wallboard or the bolivian magnesium board. As shown
in Tables 3 and 4, increasing the steel plate thickness has
negligible effect on the strength but an obvious effect on the
154 J. Ye et al. / Thin-Walled Structures 92 (2015) 146–159

Table 2
Experimental results.

Specimen number Fp (kN) Fs (kN) Fe (kN) Fy (kN) Δe (mm) Δy (mm) Δu (mm) μI μII

Fy-I Fy-II Δy-I Δy-II

Specimen 1 93.1 25.9 36.3 73.2 70.4 3.71 21.02 19.24 59.68 2.84 3.10
Specimen 2a 79.1 22.0 32.2 63.6 61.1 3.51 14.04 12.22 36.37 2.59 2.98
Specimen 3 123.0 25.6 49.2 84.6 78.6 4.04 17.89 15.02 – – –
Specimen 4 72.6 20.2 29.0 59.0 51.3 2.74 13.45 9.82 67.80 5.04 6.90
Specimen 5 94.0 26.1 37.6 73.6 68.2 3.66 17.55 14.53 53.60 3.05 3.69
Specimen 6 94.0 26.1 37.6 75.5 74.0 3.42 14.75 13.48 44.80 3.04 3.32

a
Specimen failed with an unusual failure mode caused by assembling error.

Fig. 11. Screw connection test setup.

Table 3 5. Prediction of the shear performance of CFS shear walls

Test results of screw connections (for GWB and BMG).

Specimen Pp Pe Py-I Py-II dp de dy-I dy-II

Previous studies and cyclic test results described in this paper
number (N) (N) (N) (N) (mm) (mm) (mm) (mm) showed that CFS shear walls exhibited notably complex and highly
non-linear behavior. It is uneconomical to repeatedly obtain the
G12/H-15/0.9 442 177 355 313 1.55 0.11 0.63 0.47 shear performance of CFS shear walls by experiment. Because the
G12/H-20/0.9 522 208 415 371 2.93 0.19 1.06 0.77
shear behaviors of CFS shear walls and screw connections are closely
G12/H-25/0.9 625 250 515 487 2.69 0.40 1.24 1.08
G12/H-15/1.2 456 182 379 388 1.51 0.31 0.97 0.92 related, it is important to develop a convenient method to evaluate
G12/H-20/1.2 579 231 492 489 1.95 0.36 1.11 1.02 specific properties, such as the lateral stiffness and the shear capacity
G12/H-25/1.2 693 277 581 564 2.43 0.42 1.33 1.17 of CFS shear walls, via constitutive screw connection tests.
B12/H-15/0.9 538 215 448 440 1.93 0.42 1.33 1.28
As illustrated in part 4, the failure modes of the screw connections
B12/H-20/0.9 636 254 533 519 2.35 0.37 1.21 1.10
B12/H-25/0.9 687 275 581 575 3.67 0.74 2.15 1.90
were similar to those observed from the corresponding wall speci-
B12/H-15/1.2 426 170 353 351 1.25 0.26 0.78 0.70 mens. Hence, the following work was performed to study the shear
B12/H-20/1.2 616 246 518 515 2.09 0.34 1.01 0.92 behavior of the walls sheathed with double-layer wallboards on both
B12/H-25/1.2 797 319 644 595 2.72 0.25 1.10 0.90 sides: (1) evaluation of the shear capacity of the walls based on the
shear strength of the screw connections; (2) evaluation of the lateral
Note: G and B label gypsum wallboard and bolivian magnesium board, respectively;
12 is the thickness of the board (mm); H implies that the loading direction is stiffness of the walls based on the shear strength and the deforma-
parallel to the entire wallboard width; 15, 20 and 25 are the edge distances (mm); tion of the screw connections; and (3) definition of the shear-wall
0.9 and 1.2 are the thicknesses of the steel plate (mm). drift angle limit and the corresponding deformation limit of the
screw connections based on three-level performance criteria, includ-
ing “no damage in minor earthquake”, “repairability in moderate
Table 4 earthquake” and “no collapse in severe earthquake”.
Test results of screw connections (for CSB).
5.1. Evaluation of the shear capacity of the walls
Specimen number Pp (N) Pe (N) dp (mm) de (mm)

C12/H-15/0.9 918 367 0.80 0.20

The shear capacity of the walls was evaluated based on the
C12/H-20/0.9 1385 554 1.16 0.45 following assumptions: (1) the shear capacity of CFS shear wall is
C12/H-25/0.9 1908 763 1.77 0.58 assumed to be the accumulation of the shear strength of the screw
connections for the face layer wallboard at either the up edges or
Note: C labels the calcium silicate board.
the down edges of the wall and (2) considering the effect of the
gypsum wallboard, an increase (30% of the accumulative value) is
deformability of the screw connections, and the peak strength of considered. Table 5 summarizes the shear strength of the walls.
the screw connections can be significantly increased by increasing For specimens 1 and 3, the calculated results (Fpc) are approxi-
the edge distance to the screws. mately equal to the test results (Fpt) because of specimens that lost the
 J. Ye et al. / Thin-Walled Structures 92 (2015) 146–159 155

Fig. 12. Failure modes of screw connections (single-layer board).

Fig. 13. Failure modes of the screw connections (double-layer boards).

156 J. Ye et al. / Thin-Walled Structures 92 (2015) 146–159

700
2000
650
600 1800
550 1600
500
1400
450
Force (N)

Force (N)
400 1200
350 1000
300
800
250
200 600
150 400 C12/H-15/0.9
100 C12/H-20/0.9
G12/H-/0.9: 15; 20; 25 200
50 C12/H-25/0.9
B12/H-/0.9: 15; 20; 25
0 0
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0 6.5 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 2.4 2.6 2.8 3.0
Displacement (mm) Displacement (mm)
Fig. 14. Typical load–deformation curves of the screw connections. (a) For GWB and BMG and (b) For CSB.

wall was only 30%. The calculated results (Fpc) increase to 62.5 kN,
64.5 kN, and 64.5 kN, when the shear strength of the screw
connections for both layer wallboards are accumulated without
strength reduction in the gypsum wallboards; the ratios of the
calculated results (Fpc) to the test results (Fpt) are 0.86, 0.69, and
0.69 with a considerable safety assurance.

5.2. Evaluation of the lateral shear stiffness

An equivalent bracing model was adopted to predict the lateral

stiffness on different levels to define the shear-wall drift angle limit.
Fig. 15. Definition of the yield limit of the screw connections (for CSB).
As shown in Fig. 17, the model is based on the following assumptions:
(1) the steel frame is assumed as bar elements that were mutually
hinged around the wall, and the bases of the end studs are assumed as
Table 5 fixed ends in consideration of the hold-downs; (2) the wallboards are
Shear strength of the walls. equivalent to diagonal braces considering only the axial load; (3) the
equivalent braces are assumed to be the primary members to resist
Specimen Fpt Fpc γp F400 F300 F100 γ400 γ300 γ100 the lateral load; and (4) the axial rigidity of the braces are cumulative.
number (kN) (kN) (kN) (kN) (kN)
According to material mechanics theory, brace AD and brace BC
Specimen 1 93.1 97.2 1.04 45.8 49.7 89.2 0.49 0.53 0.96 have identical absolute values of internal force. For example, for brace
Specimen 2n 79.1 97.2 1.23 49.3 52.5 71.9 0.62 0.66 0.91 AD, the lateral stiffness of the CFS shear walls sheathed with single-
Specimen 3 123.0 129.6 1.05 61.0 68.3 101.7 0.58 0.64 0.96 layer wallboards on two sides is summarized in (6)–(13) based on the
Specimen 4 72.6 44.6 0.61 47.9 52.1 71.4 0.66 0.72 0.98
above assumptions and various elastic theories, structural mechanics
Specimen 5 94.0 43.2 0.46 53.3 59.9 87.7 0.57 0.64 0.93
Specimen 6 94.0 43.2 0.46 53.7 61.4 90.6 0.57 0.65 0.96 theories and force-balance principles.

Note: Fpt is the peak strength obtained from test results, whereas Fpc is obtained Δ0 Eb Ab
from the calculated results; F400, F300, F100 are the shear strength to the point of Ft ¼ ð6Þ
Lb
drift angle corresponding to 1/400, 1/300, 1/100, respectively; γp, γ400, γ300, γ100 are
the ratios of Fpc, F400, F300, F100 to Fpt, respectively.
M AB ¼ ic ð2θB  6Δ=HÞ ð7Þ

shear strength as soon as the screws were sheared off (Fig. 16a and b). M CD ¼ ic ð2θD  6Δ=HÞ ð8Þ
For specimens 4–6, the screws were originally perpendicular to the
wallboards, and the compression area of the wallboards surroun- M BA ¼ M DC ¼ 0 ð9Þ
ding the screws became large when screws tilted, which indirectly
improved the shear strength of the specimens until the screws pulled M BA þ M AB
Q AB ¼ ð10Þ
through the face layer wallboards (Fig. 16a and c). Thus, the calculated H
results (Fpc) of specimens 4–6 are lower than the test results (Fpt),
particularly for the latter two specimens, which might be because the M DC þ M CD
Q CD ¼ ð11Þ
thicker studs held the tails of the screws and delayed the screws from H
being pulled out.
As shown in Table 5, because the calculated results (Fpc) of Q AB þ Q CD þ P  2F t cos θ ¼ 0 ð12Þ
specimens 4–6 are nearly half of the test results (Fpt), it would be
unreasonable to evaluate the shear capacity considering that the P 6ic 2L2 Eb Ab
K¼ ¼ þ ð13Þ
contribution of the gypsum wallboards to the shear capacity of the Δ H 2
ðL2 þ H 2 Þ1:5
 J. Ye et al. / Thin-Walled Structures 92 (2015) 146–159 157

The face layer board

The face layer board
The base layer board The base layer board
Screw
Screw

Stud
Stud

The face layer board

The base layer board
Screw

Stud

Fig. 16. Damage process of the screw connections.

i-layer wallboard, which was obtained from the test; ti is the thickness
Δ of the i-layer wallboard; Pp is the peak strength of the screw
P B B' D D'
connection; d is the deformation of the screw connection; and ns is
θ θ the number of screws along the vertical edge on one side of the wall.
E bA b As shown in Eq. (14), the axial rigidity of the braces (EbiAbi) is
Δ'
H ic the crucial influencing factor on the evaluation of the lateral
stiffness of the walls with the same configuration, which directly
Lb reflects in the deformation of the screw connections (d). The cyclic
test results, which were described in part 3, also confirmed that
A the nonlinear behavior of the walls is lumped into the deformation
L C of the screw connections. Hence, it would be significant to divide
different deformation levels of screw connections to calculate the
Fig. 17. Equivalent bracing model.
lateral stiffness based on which the shear-wall drift angle limit
would be defined. In this study, three deformation levels were
If the contribution of both layer wallboards for the lateral defined as follows: (1) elastic deformation limit (δe), (2) yield
stiffness is considered, Eq. (13) can be evolved to Eq. (14) deformation limit (δy), and (3) ultimate deformation limit (δu).
P2 Table 6 summarizes the lateral shear stiffness of the walls. As
P 6ic 2L2 i¼1 Ebi Abi shown in Table 6, the elastic stiffness values KEc for both walls
K¼ ¼ þ ð14Þ
Δ H 2
ðL2 þ H 2 Þ1:5 sheathed with gypsum wallboards and calcium silicate boards and
those sheathed with gypsum wallboards and bolivian magnesium
1 ðH 2 þ L2 Þ1:5 boards are consistent with the test results, and the absolute value of
Ebi Abi ¼ U ð15Þ the relative errors is within 14%. For the yield stiffness of the walls
2 ðHL=Gi t i Þ þ ð2d=P p ns ÞðH 2 þHLÞ
sheathed with gypsum wallboards and bolivian magnesium boards,
where Ft is the tension of brace AD; Δ0 is the elongation value of brace the calculated results KYc-I and KYc-II were separately compared with
AD; EbiAbi is the axial rigidity of the brace, which was simplified from K300 and K400 to determine the reasonable drift angle limit. The
the i-layer wallboards on one side of the wall and obtained using Eq. absolute value of the relative errors (κ3) include the errors (κ3-I)
(15) [16]; Lb is the length of the brace; MAB, MBA, MCD and MDC are the between KYc-I and K300 and the errors (κ3-II) between KYc-II and K300,
bending moments of the end studs; ic is the line rigidity of the end whereas the absolute value of the relative errors (κ4) include the
studs; θB and θD are the rotation of the end studs on top; QAB and QCD errors (κ4-I) between KYc-I and K400 and the errors (κ4-II) between
are the shear forces of the end stud base; Δ is the actual shear KYc-II and K400, all of which are less than 20%.
displacement of the wall; H is the wall height; L is the wall length; P is It is important to realize that for the walls sheathed with gypsum
the lateral load; θ is the angle between brace AD and the top of the wallboards and calcium silicate boards, the yield stiffness was invalid
wall; K is the lateral stiffness of the wall; Gi is the shear modulus of the because the screw connections abruptly disrupted. The relationship
158 J. Ye et al. / Thin-Walled Structures 92 (2015) 146–159

Table 6
Evaluation of the lateral shear stiffness.

Specimen number KE (N/mm) K1000 κ1 κ2 K300 K400 KYc (N/mm) κ3 κ4

KEt KEc KYc-I KYc-II κ3-I κ3-II κ4-I κ4-II

Specimen 1 9794 10707 10052 0.08 0.03 4967 6107 – – – – – –

Specimen 2n 9163 10707 9720 0.14 0.06 5246 6577 – – – – – –
Specimen 3 12190 13970 12736 0.13 0.04 6826 8136 – – – – – –
Specimen 4 10569 10264 10666 0.03 0.01 5210 6391 5947 6260 0.14 0.20 0.07 0.02
Specimen 5 10986 10654 11739 0.03 0.06 6142 7111 6412 6770 0.04 0.10 0.10 0.05
Specimen 6 10287 10654 10630 0.03 0.03 5987 7160 6412 6770 0.07 0.13 0.10 0.05

Note: KEt ¼ 0.4Fp/Δe labels the elastic stiffness devised from the test results, whereas KEc labels the calculated results from Eq. (14) based on the elastic deformation (de) of the
screw connections; KYc-I and KYc-II label the calculated results of the yield stiffness from Eq. (14) based on the yield deformations (dy-I) and (dy-II) of the screw connections,
respectively; K1000, K300 and K400 are the secant stiffness values to the point of drift angle that correspond to 1/1000, 1/300 and 1/400, respectively; κ1–κ4 are the absolute
values of the relative errors between KEt and KEc, KEt and K1000, K300 and KYc, and K400 and KYc, respectively.
n
Specimen 2 failed with an unusual failure mode caused by assembling error, as noted in table 2.

between the deformation and the force of the screw connections was
linear; hence, the shear strengths to the point of drift angle that
corresponded to 1/400 (F400) and 1/300 (F300) were compared to the
peak shear strength (Fp) of the walls to determine the reasonable
yield deformation of the screw connections. Both F400 and F300 are
nearly half of the peak shear strength (Fp), as shown in Table 5;
therefore, the yield deformation limit (δy) should not exceed dαp
(α ¼0.5). However, the equivalent bracing model is not suitable for
calculating the lateral stiffness after the wall yield. Therefore, F100
was compared with the peak shear strength (Fp) of the walls, where
F100 is the shear strength to the point of drift angle that corresponded
to 1/100, which was proposed to define the drift angle limit for
seismic design in rare earthquake [16]. Table 5 shows that F100 is
notably close to the peak strength of the walls; thus, the ultimate
deformation limit (δu) should not exceed dp.
5.3. Performance criteria
By relating the deformation of the screw connections with the
lateral stiffness of the walls, the three-level performance criteria
(i.e., no damage in minor earthquake; repairability in moderate
earthquake; no collapse in severe earthquake) could be satisfied
from the viewpoint of the drift angle limit with the corresponding
screw deformation limit, as follows:
(1) In case of serviceability level wind loading and minor earth-
quake, CFS shear walls should be designed in the elastic stage
with a drift angle limit of 1/1000, and the deformation of
screw connections should not exceed the elastic deformation
limit (δe), which corresponds to 0.4Pp.
(2) In case of moderate earthquake, the yield deformation in the
screw connections is permitted but should not exceed the
yield deformation limit (δy). It would be safer for CFS struc-
tures if 1/400 is adopted as the drift angle limit instead of
1/300. Definitions of the yield deformation limit (δy) of the
screw connections should be considered from two aspects. For
sheathings with abrupt disruption in the failure mode of the
corresponding screw connections, dαp should be considered as
(δy), where α is the ratio between the shear strength to the
point of drift angle that corresponds to 1/400 (F400) and the
peak strength (Fp) of the walls. For example, for the calcium
silicate board, d0.5p should be considered as (δy). For sheath-
ings with apparent symptoms (e.g., titling of screws and
slackness of screw holes) in the failure mode of the corre-
sponding screw connections, (δy) could be determined accord-
ing to the methods described in Fig. 10. Method II would be the
better choice because of the conservative results of the yield
deformation limit (δy) compared to method I, and the struc-
ture would be safer.
(3) In case of severe earthquake, sheathings with abrupt disrup-
tion in the failure mode of the corresponding screw connec-
tions should not be considered as wallboards. For sheathings
with apparent symptoms in the failure mode of the corre-
sponding screw connections, if 1/100 is taken as the shear-wall
drift angle limit, the peak deformation (dp) corresponding to
(Pp) should be considered as the ultimate deformation limit
(δu) of the screw connections to prevent structure collapse.
6. Conclusions
The conclusions based on the results of this study are summar-
ized as follows:
(1) The walls sheathed with double-layer wallboards on both
sides had significantly improved shear capacity. For the screw
connections in fields such as the edges and the joints, which
represented the most sensitive parts of the walls, the failure
modes were either the screws being sheared off or pulled
through the wallboards.
(2) Differences in sheathing materials appear to significantly influ-
ence the shear behavior of the screw connections. The screw
connections of the calcium silicate board had abrupt disruption
in the failure mode. The walls sheathed with calcium silicate
boards as the face layer wallboards abruptly disrupted and had
lower ductility than the walls sheathed with bolivian magne-
sium boards as the face layer wallboards.
(3) The shear strength of the walls can be significantly improved
by changing the interior stud section under the condition of
identical steel consumption, but no meaningful difference was
observed when the stud spacing was changed from 400 mm to
600 mm at the center of a field fastener spacing of 300 mm.
The improvement on the unit shear strength by increasing the
wall length is negligible.
(4) For the walls with a significant difference in strength between
the face layer wallboards and the base layer wallboards, such as
the calcium silicate board and the gypsum wallboard, the shear
capacity can be increased by 30% because of the contribution of
the gypsum wallboards. For the walls with a similarity in
strength between the face layer wallboards and the base layer
wallboards, such as the bolivian magnesium board and the
gypsum wallboard, the shear capacity can be evaluated by
accumulating the total strength of the screw connections for
both layer wallboards.
(5) For the walls sheathed with gypsum wallboards and bolivian
magnesium boards, the three-level performance criteria for
seismic design can be satisfied depending on the drift angle
limits, such as 1/1000, 1/400 and 1/100, which correspond to
 J. Ye et al. / Thin-Walled Structures 92 (2015) 146–159
the screw connection deformation limits δe, δy and δu, respec-
tively. It would be safe to determine the yield deformation
limit of screw connections for gypsum wallboard or bolivian
magnesium board according to Method II.
(6) The walls sheathed with gypsum wallboards and calcium
silicate boards can only be used in areas of low seismicity. The
shear-wall drift angle limit for minor and moderate earthquakes
are 1/1000 and 1/400, which correspond to the screw connection
deformation limits δe and δy, respectively. The yield deformation
limit (δy) of the screw connections for the calcium silicate board
can be defined based on the ratio between the shear strength to
the point of drift angle corresponding to 1/400 (F400) and the
peak strength (Fp) of the walls.
Acknowledgments
This research is sponsored by the National Key Technology
Research and Development Program of the Ministry of Science and
Technology of China (2011BAJ08B04).
References
[1] Fülöp LA, Dubina D. Performance of wall-stud cold-formed shear panels under
monotonic and cyclic loading Part I: experimental research. Thin-Walled
Struct 2004;42(2):321–38.
[2] Lin SH, Pan CL, Hsu WT. Monotonic and cyclic loading tests for cold-formed
steel wall frames sheathed with calcium silicate board. Thin-Walled Struct
2014;74:49–58.
[3] Shakibanasab A, Attari Nader KA, Mehdi S. A statistical and experimental
investigation into the accuracy of capacity reduction factor for cold-formed
steel shear walls with steel sheathing. Thin-Walled Struct 2014;77:56–66.
[4] Chen W, Ye JH, Bai Y, Zhao XL. Full-scale fire experiments on load-bearing
cold-formed steel walls lined with different panels. J Constr Steel Res
2012;79:242–54.
159
[5] Tarpy T.S., Girard J.D. Shear resistance of steel-stud wall panels. In: Proceed-
ings of the 6th International Specialty Conference on Cold-Formed Steel
Structures. St. Louis; 1982. p. 449–65.
[6] Miller T, Peköz T. Behaviour of gypsum-sheathed cold-formed steel wall studs.
J Struct Eng 1994;120(5):1644–50.
[7] Serrette RL, Encalada J, Juadines M, Nguyen H. Static racking behavior of
plywood, OSB, gypsum, and fiberboard walls with metal framing. J Struct Eng
1997;123(8):1079–86.
[8] Pan CL, Shan MY. Monotonic shear tests of cold-formed steel wall frames with
sheathing. Thin-Walled Struct 2011;49(2):363–70.
[9] Landolfo R, Fiorino L, Della Corte G. Seismic behaviour of sheathed cold-
formed structures: physical tests. J Struct Eng 2006;132(4):570–81.
[10] Yu C. Shear resistance of cold-formed steel framed shear walls with 0.686 mm,
0.762 mm, and 0.838 mm steel sheet sheathing. Eng Struct 2010;32:1522–9.
[11] Balh N, DaBreo J, Ong-Tone C, El-Saloussy K, Yu C, Rogers CA. Design of steel
sheathed cold-formed steel framed shear walls. Thin-Walled Struct 2014;75:76–86.
[12] Nithyadharan M, Kalyanaraman V. Behaviour of cold-formed steel shear wall
panels under monotonic and reversed cyclic loading. Thin-Walled Struct
2012;60:12–23.
[13] Fiorino L, Della Corte G, Landolfo R. Experimental tests on typical screw
connections for cold-formed steel housing. Eng Struct 2007;29:1761–73.
[14] Nithyadharan M, Kalyanaraman V. Experimental study of screw connections in
CFS-calcium silicate board wall panels. Thin-Walled Struct 2011;49:724–31.
[15] American Iron and Steel Institute (AISI). North American standard for cold-
formed steel framing-lateral design 2007 edition with supplement No. 1.
Washington, DC: American Iron and Steel Institute; 2007.
[16] Japan Iron and Steel Federation. Design guide for light gauge steel buildings.
Tokyo: Gihodo Shuppan Co., Ltd.; 2002 (in Japanese).
[17] Bozorgnia Y, Bertero V. Earthquake engineering – from engineering seismology
to performance based engineering. International Code Council. U.S.A: CRC Press;
2004.
[18] International Building Code [S]. U.S.A: International Code Council, Inc.; 2009.
[19] Eurocode 8: Design of structures for earthquake resistance – Part I: general
rules, seismic actions and rules for buildings. Bruxelles: European Committee
for Standardization; 2004.
[20] BSL. Building Standard Law. Japan; 1987.
[21] GB 50011-2010. Code for seismic design of building. Peking: China Architec-
ture & Building Press; 2010 (in Chinese).
[22] ECCS. Recommended testing procedure for assessing the behaviour of struc-
tural steel elements under cyclic loads; 1985.
[23] Park R. Evaluation of ductility of structures and structural assemblages from
laboratory testing. Bull N Z Natl Soc Earthq Eng 1989;22(3):155–66.
[24] Han LH, Yang YF, Tao Z. Concrete-filled thin-walled steel SHS and RHS beam-
columns subjected to cyclic loading. Thin-Walled Struct 2003;41:801–33.