Вы находитесь на странице: 1из 13

Seismic Behavior of Cold-Formed Steel Frames with

Bolted Moment Connections


Ming Chen 1; Jia-Hao Huo 2; and Yi-Wen Xing 3
Downloaded from ascelibrary.org by Indian Institute Of Technology Delhi on 11/17/20. Copyright ASCE. For personal use only; all rights reserved.

Abstract: This paper investigates the seismic behavior of the cold-formed steel (CFS) frame with bolted moment connections by quasi-static
loading tests and numerical analysis. The quasi-static loading test takes the thickness of the gusset plate and the span of the frame as two key
parameters. Meanwhile, the effects of the column slenderness ratio, axial compression ratio of the column, height/thickness ratio of the column
web, and beam/column linear stiffness ratio are the primary focus of the finite-element method. By using these methods, this paper presents
seismic behavior and a simplified hysteretic model for the CFS frame. The results indicate that “failure of the beam end” and “failure of the
column base” are the two main failure modes. The CFS frame shows outstanding seismic performance, and the column slenderness ratio, beam/
column linear stiffness ratio, axial compression ratio of the column, and height/thickness ratio of the column web mainly affect it. The hys-
teretic model has been compared with the test results to demonstrate high reliability. The findings presented here provide future reference for
the design of such a cross-sectional frame. DOI: 10.1061/(ASCE)ST.1943-541X.0002538. © 2019 American Society of Civil Engineers.
Author keywords: Cold-formed steel (CFS) frame; Seismic behavior; Quasi-static loading test; Numerical analysis; Hysteretic model.

Introduction experienced severe vertical buckling, leading to out-of-plane bend-


ing deformation at the gusset plate.
The thin-walled cold-formed steel (TWCFS) structural system has Several studies have also investigated the seismic performance of
undergone a dramatic expansion in its range of applications over the bolted moment connections using gusset plates (Bagheri Sabbagh
past decade. It has its own unique advantages for use in low-rise and et al. 2012; Yu et al. 2005; Bučmys and Daniūnas 2015; Wrzesien
mid-rise buildings, including being lightweight, yielding simple and et al. 2012; Bučmys et al. 2018). Experimental results have been
fast construction, and having greater flexibility in manufacturing. used to validate different modeling techniques [e.g., detailed finite-
The seismic performance of TWCFS structures has been previ- element (FE) modeling, equivalent spring elements, and internal
ously investigated (e.g., Schafer et al. 2016; Dubina 2008; Moghimi force distributions], which can predict the stiffness and strength of
and Ronagh 2009; Fiorino et al. 2014; Lim and Nethercot 2004a; the connections and improve computational efficiency. According
Mohebbi et al. 2016; Xu et al. 2018). The majority of these studies to the AISI S400 standard (AISI 2015), CFS frame systems with
adopted shear wall panels as the main lateral loadbearing system, bolted moment connections are expected to withstand inelastic fric-
focused on the behavior of shear wall panels, and overlooked the tion and bearing deformation at the bolted-beam to column connec-
effects of TWCFS frames. However, moment-resisting TWCFS tions. These findings have shown that these semi-rigid joints have
frames can play an important role in seismic performance and good bearing deformation and ultimate lateral load, indicating that
potentially improve the seismic behavior of TWCFS structures. semi-rigid joints should be introduced to TWCFS frames.
In addition, compared with shear wall systems, moment-resisting The structural performance of TWCFS frames has been inves-
TWCFS frames provide higher flexibility for span arrangement and tigated, and studies have focused on different aspects. Ali and col-
future alterations. leagues (2010) tested ten TWCFS frames under lateral load, and
The structural performance of moment-resisting connections the beam and column members of the frames were formed by
(termed eaves and apex joints) with different bolt group sizes was single-lipped channel sections and connected back-to-back at the
investigated by Dundu and Kemp (2006). Other studies showed joint. They found that the stiffness and performance of the column
experimentally and numerically that the final failure mode of the base connections significantly affected the structural behavior of
joint zone was local buckling of the C section steel. To solve such a the frames. Sabbagh et al. (2010) performed an experimental inves-
failure, Lim and Nethercot (2003, 2004a) inserted two gusset plates tigation to study the seismic performances of two full-scale one-
at the eave joint and formed a new eave joint of a portal frame. story frames made of CFS members under lateral cyclic loads.
The results showed that the C shaped cross section steel web Their results indicated that the TWCFS moment-resisting frames
show good potential for aseismic performance, but the details of
1 the main structural elements require further study. Mojtabaei et al.
Professor, Civil Engineering School, Inner Mongolia Univ. of Science
and Technology, Baotou, Inner Mongolia 014010, China (corresponding (2018) investigated an innovative TWCFS frame under static mon-
author). Email: cmlx-1978@163.com otonic loading; the columns consisted of a box-shaped section,
2
Postgraduate, Civil Engineering School, Inner Mongolia Univ. of and the beam was formed from a back-to-back lipped channel.
Science and Technology, Baotou, Inner Mongolia 014010, China. The results show that the axial load ratio and the width-to-thickness
3
Postgraduate, Civil Engineering School, Inner Mongolia Univ. of
ratio of the columns have significant effects on the seismic perfor-
Science and Technology, Baotou, Inner Mongolia 014010, China.
Note. This manuscript was submitted on December 15, 2018; ap-
mance of this frame, and with appropriate design of the main struc-
proved on July 23, 2019; published online on December 17, 2019. Dis- tural elements, good seismic performance can be obtained from
cussion period open until May 17, 2020; separate discussions must be this frame.
submitted for individual papers. This paper is part of the Journal of According to the aforementioned studies, TWCFS moment-
Structural Engineering, © ASCE, ISSN 0733-9445. resisting frames provided superior structural performance than

© ASCE 04019212-1 J. Struct. Eng.

J. Struct. Eng., 2020, 146(3): 04019212


conventional TWCFS shear wall panels. However, additional stud- and the seismic behavior of the CFS frame is clarified. The hyster-
ies are required to investigate the nonlinear seismic performance etic model is established, and relevant suggestions for CFS frame
of TWCFS moment-resisting frames based on different types of design are proposed.
connections, sections, and design parameters. Therefore, our team
proposed a “nonbox” section consisting of back-to-back CFS, gus-
set plates, and high-strength bolts, and the gusset plate is located Test Program
between back-to-back CFS and connected by bolts. The mechani-
cal properties of the compression member and joint base of this
CFS Frame Design
section have been researched.
This paper analyzes the seismic behavior of a cold-formed steel Five 1:3 CFS scale models (numbered S-1, S-2, S-3, S-4, and S-5)
(CFS) frame whose beam-column joints, column bases, and the were designed to study the effects of the thickness of the gusset plate
position of beam and column comprise this novel section. These and span of the CFS frame. The beam, column, and gusset plate of
Downloaded from ascelibrary.org by Indian Institute Of Technology Delhi on 11/17/20. Copyright ASCE. For personal use only; all rights reserved.

regions are strengthened by this cross section, facilitating the con- CFS frame adopted Q235B steel [GB 50017 (Chinese Standard
nection of beam-column members and ensuring that the frame has 2017)]. The frame section was formed by C shaped cross section
better mechanical performance. In the quasi-static loading test, the steel back-to-back, and the web was connected to the gusset plate
thickness of the gusset plate and span of the frame (beam/column by a grade 8.8 M20 friction-type high-strength preloaded bolt
linear stiffness ratio; hereafter, this is abbreviated as linear stiffness whose pretension force was 125 kN. The beam and column of C
ratio) are used as key parameters. The effects of slenderness ratio, shaped cross section were chosen from the appendix of reference
axial compression ratio, height/thickness ratio of the column web [GB 50018 (Code of China 2002)] with dimensions of C140 ×
(hereafter, this will be abbreviated as height/thickness ratio), and 60 × 20 × 2.5 mm and C180 × 70 × 20 × 2.0 mm, respectively.
linear stiffness ratio are investigated using the FE method. The hys- The precise dimensions of the CFS frame are shown in Table 1
teresis curve and backbone curve of the CFS frame are obtained, and Fig. 1.
The material properties of the C shaped cross section steel and
gusset plate were tested according to ASTM A370 (ASTM 2014).
Table 1. Details of the CFS frames The yield stress, ultimate stress, fracture strain, and modulus of elas-
No. tplate (mm) Span (mm) Height (mm) K ticity measured for the tested specimens are presented in Table 2.
S-1 8 1,400 1,200 0.481
S-2 10 1,400 1,200 0.481 Loading Setup and Protocol
S-3 12 1,400 1,200 0.481
S-4 8 2,400 1,200 0.281 The tests were conducted in the Civil Engineering Laboratory of
S-5 8 2,600 1,200 0.259 the Inner Mongolia University of Science and Technology. The ex-
Note: tplate = thickness of the gusset plate; and K = linear stiffness ratio, perimental test setup used in this study is shown schematically in
meaning K b =K c , where K b and K c are linear stiffness of the beam and Fig. 2. A 500-kN hydraulic actuator operating in displacement con-
column, respectively, representing the flexure stiffness of the beam and trol was utilized to apply horizontal displacement at the top of
column unit length. the column. The CFS frame was fixed to the test platform using a

Fig. 1. Geometry of the CFS frame (unit: meters): (a) geometry of the frame; (b) Gusset plates A; (c) Gusset plates C; (d) Gusset plates B;
(e) A-A section; and (f) Gusset plates D.

© ASCE 04019212-2 J. Struct. Eng.

J. Struct. Eng., 2020, 146(3): 04019212


Table 2. Material properties ensure load transfer of the CFS frame without changing the column
Yield Ultimate and beam size.
strength strength E-modulus Poisson Elongation Two damage modes are presented in Figs. 3 and 4.
Material (MPa) (MPa) (GPa) ratio (%)
Load-Displacement Hysteresis Curve
C shaped cross section 246.82 394.42 2.12 0.33 12.51 Fig. 5 shows the load-displacement hysteresis curve for each CFS
steel frame. The results show that the hysteresis curves of the five spec-
Gusset plate of 8 mm 387.91 504.58 1.97 0.33 28.65
imens are essentially consistent. The area inside the hysteresis loop
Gusset plate of 10 mm 300.18 446.71 1.96 0.31 27.53
Gusset plate of 12 mm 268.33 384.53 1.92 0.32 26.85 is large and of good symmetry for each CFS frame, indicating good
energy dissipation capacity and ductility.
During the initial stage of the test, the CFS frames were essen-
tially in the elastic stage. The load and displacement in the hysteresis
Downloaded from ascelibrary.org by Indian Institute Of Technology Delhi on 11/17/20. Copyright ASCE. For personal use only; all rights reserved.

concrete ground beam and bolts. At the load-end of the frame, a curves grew linearly, indicating that the stiffness of the CFS frame
hydraulic jack was set to apply the cyclic load. At the middle point was constant. The curves show good symmetry. As cyclic loading
of the CFS frame column, two H-type steel beams were set to increased, the proximal flange of the load-end appeared to buckle,
prevent out-of-plane instability occurring on the CFS frame. the stiffness of the CFS frame obviously degraded and residual
The loading protocol was designed according to reference deformation gradually increased. When the specimen entered the
[ASTM E2126 (ASTM 2011)] (Table 3). The test was halted when yielding stage, the plastic failure at the web of the proximal beam
the load value fell below 80% of the ultimate load. gradually accumulated, the stiffness decreased more rapidly, and the
hysteresis loop had a full shape. After reaching the ultimate load, the
stiffness of the CFS frame gradually declined. As loading continued,
Test Results and Analysis
the slip between high-strength bolts and C section steel increased,
and there was some slippage of the bolts. Because of the serious
Failure Models loosening of bolts at the distal joint zone, the hysteresis curve of
There are two types of CFS frame-failure modes. The failure mode S-3 appears approximately perpendicular to the horizontal axis of
of the S-1, S-2, and S-3 frames is termed “failure of the beam end.” the curve. For S-4 and S-5, because the transfer efficiency of bend-
In the early stages of loading, obvious deformation of the frames ing moment is so bad that the stiffness of the frame decreases sig-
was not apparent. When loaded to yield load, the C-type steel at nificantly, the slip is obvious in the hysteresis curves, and pinching
the beam end was first to yield, and the plastic hinge was formed. is more severe than for S-1, S-2, and S-3.
The plastic hinge area was outside the beam-column joint region.
As the loading continued, the plastic area at the end of the beam Backbone Curve
continued to expand, resulting in decreased stiffness of the beam The points of the maximum values of the loads in each cycle of
and redistribution of the internal force of the CFS frame. The bend- the hysteresis curve are extracted and connected in turn to obtain the
ing moment of the column ends increased too significantly to form backbone curve. As shown in Fig. 6, the backbone curves of the five
a plastic hinge. At this point, the CFS frames reached ultimate load CFS frames are “S” shaped, indicating that the CFS frame under-
capacity. During the increase of cyclic loading, the plastic zone at goes elastic, elastic-plastic, and ultimate plastic failure under cyclic
the column end expanded continuously and formed a plastic hinge, loading.
and the CFS frames eventually suffered damage. From a comparison of the backbone curves for S-1, S-2, and
The failure mode of the S-4 and S-5 frames is termed “failure of S-3, the ultimate load capacity of the CFS frame is shown to be
the column base.” That is, the buckle was concentrated on proximal essentially unchanged with increasing thickness of the gusset plate.
beam ends and column bases instead of distal beam ends and col- This consistency is because the load capacity of the CFS frame
umn bases. Because the span of the CFS frame was so large, the is mainly determined by the C shaped cross section dimensions.
linear stiffness ratio was reduced, blocking the transmission of the Therefore, the effect of gusset plate thickness is not significant pro-
bending moment and leading to one side of the column taking all of vided the joint is stronger than the component. However, the thick-
the load. The CFS frame system failed to perform its function and ness of the gusset plate has a significant effect on the initial stiffness
thus did not conform to the expected force characteristics. There- of the frame; when thickness increases from 8 to 12 mm, initial
fore, when the required span is large, extra supports can be added to stiffness increases by 28.86%.

Fig. 2. Experimental test setup: (a) photograph; and (b) sketch.

© ASCE 04019212-3 J. Struct. Eng.

J. Struct. Eng., 2020, 146(3): 04019212


Table 3. Test loading protocol (Shamim et al. 2013)], and CFS strap braced walls [1.23–1.51
Loading Displacement (Fiorino et al. 2016)], the CFS frame indicates superior ductility.
Protocol (kN) (mm) Cycles According to the data for S-1, S-2, and S-3, energy consump-
tion is increased by increasing the thickness of the gusset plate.
Force-controlled 5=−5 — 1
10/−10 — 1
Consequently, the thickness of the gusset plate affects energy
::: — 1 consumption.
F y = − Fy Δy 1
Displacement-controlled — 1.25Δy = − 1.25Δy 3 FE Model
— 1.50Δy = − 1.50Δy 3
— 1.75Δy = − 1.75Δy 3
— ::: 3 Model Design
Downloaded from ascelibrary.org by Indian Institute Of Technology Delhi on 11/17/20. Copyright ASCE. For personal use only; all rights reserved.

Note: Fy and Δy = yield load and yield displacement of the CFS frame. According to the normal column height (slenderness ratio), beam
span (linear stiffness ratio), and section parameters of the CFS
frame, an analysis specimen “BASE” was constructed. According
By comparing the backbone curves of S-1, S-4, and S-5, the re- to the test results, the load capacity of the CFS frame was not sig-
sults show that when the linear stiffness ratio reduces from 0.481 nificantly affected by the thickness of the gusset plate. Therefore,
to 0.281 and to 0.259, the ultimate load capacity reduces by 9.68% based on the “BASE” specimen and considering the axial com-
and 17.59%, respectively, and the initial stiffness reduces by 29.12% pression ratio, column slenderness ratio, linear stiffness ratio, and
and 39.62%, respectively. These results show that the linear stiffness height/thickness ratio, a series of specimens was designed. The ef-
ratio has a significant effect on the ultimate load capacity of the CFS fects of these parameters on the mechanical behaviors of the CFS
frame. frames were investigated, and the parameters of the CFS frames are
listed in Table 5.
Ductility and Energy Dissipation
Ductility and energy dissipation are two important indexes to evalu-
ate seismic behavior. The calculation results for the ductility coef- Elements and Meshing
ficient (μ) and energy of the input earthquake (PE ) (Mojtabaei et al. The sixteen models of the CFS frame described previously were
2018) of the CFS frame are presented in Table 4. The energy dis- modeled using ANSYS (Fig. 7); all models were modeled using
sipation behaviors of S-4 and S-5 cannot be considered because three elements, SHELL181, COMBIN39, and BEAM188. The con-
their failure modes are not in accordance with the mechanism of stitutive relationships of the C shaped cross section steel and gusset
the CFS frame. plate are simplified as the multilinear kinematic hardening model,
The ductility coefficients of three CFS frames (S-1, S-2, and considering the Bauschinger effect. The BEAM188 element was uti-
S-3) are all above 2.47. Compared with the ductility coefficients lized to replace the screw. The COMBIN39 element was used to sim-
of the CFS frame structure covered by calcium silicate board ulate the slippage and shear failure of the bolts (Lim and Nethercot
[1.80–2.21 (Lin et al. 2014)], the staggered truss structure 2004b) (Fig. 8). The constitutive relationships of the materials
[1.75 (Zhou et al. 2009)], steel sheathed shear walls [2.04–4.86 were determined by the experimental raw material performances.

Fig. 3. Failure mode of “failure of the beam end”: (a) distal beam; (b) distal column base; (c) proximal column base; and (d) global failure.
The loading end of the specimen was regarded as the proximal end and the nonloading end was regarded as the distal end.

Fig. 4. Failure mode of “failure of the column base”: (a) distal column and beam; (b) proximal column base; (c) proximal beam; and (d) global failure.
The loading end of the specimen was regarded as the proximal end and the nonloading end was regarded as the distal end.

© ASCE 04019212-4 J. Struct. Eng.

J. Struct. Eng., 2020, 146(3): 04019212


Downloaded from ascelibrary.org by Indian Institute Of Technology Delhi on 11/17/20. Copyright ASCE. For personal use only; all rights reserved.

Fig. 5. Load-drift hysteresis curves: (a) S-1 frame; (b) S-2 frame; (c) S-3 frame; (d) S-4 frame; and (e) S-5 frame.

After debugging, the mesh sizes of the gusset plate and C shaped Constraint Conditions and Loading Scheme
cross section steel grid were 30 and 18–22 mm, respectively; the The CFS frame column base plates achieved rigid connection by
length-to-thickness ratio of the unit was controlled at 1/6–1/11; constraining all degrees of freedom. At the beam end, column end,
and the size of the grid at the bolt hole was controlled within 5 mm. middle of the beam span, and middle of the column, the out-of-plane

© ASCE 04019212-5 J. Struct. Eng.

J. Struct. Eng., 2020, 146(3): 04019212


FE Model Validation
An ANSYS FE model with the same size as the test specimen S-3
was established. According to the method of reference (Fiorino
et al. 2018), the load-displacement backbone curves in Fig. 9 and
data in Table 6 obtained from the tests and the numerical analysis
were compared. The model provided good results until reaching
maximum strength, whereas the failure displacement and the energy
of the input earthquake absorbed by structures have some differen-
ces. The main reason is that there are initial imperfections in the
specimen, and the P-Δ effect occurs after the maximum strength,
leading to the rapid decrease of the load. A comparison of the failure
Downloaded from ascelibrary.org by Indian Institute Of Technology Delhi on 11/17/20. Copyright ASCE. For personal use only; all rights reserved.

modes in Fig. 10 obtained from the tests and the numerical analysis
similarly shows that the deformation and strain of the numerical
analysis agreed well with the test results, verifying the FE model.

Results of Numerical Analysis


Fig. 6. Load-drift backbone curves of the frames.

Load-Displacement Hysteretic Curve and


Backbone Curve
Table 4. μ and PE of frames
According to the two failure mode types, hysteretic curves were
No. S-1 S-2 S-3 S-4 S-5 selected, as shown in Figs. 11 and 12. The results show that the
Δy =mm 10.34 9.44 9.39 9.75 10.07 hysteresis loops for “failure of the beam end” are large and have
Δu =mm 25.51 23.49 23.21 34.48 39.02 no significant pinch phenomenon, and the areas inside the hyster-
μ 2.47 2.49 2.47 3.54 3.65 etic loops are large and have good symmetry, indicating good en-
PE =kJ 1.43 1.44 1.46 1.66 1.63 ergy dissipation capacity. When the hysteresis curve reaches the
Note: Δy = yield displacement of the CFS frame; Δu = failure displacement load limit, the slope of the curve decreases, the degree of reduction
of the CFS frame; μ = ductility coefficient, meaning Δu =Δy ; and PE = is slow, and the stiffness of the CFS frame gradually declines.
energy of the input earthquake absorbed by structures, In CFS frames, PE Although the hysteretic curves of “failure of the column base” have
can be calculated based on the area under the lateral load-displacement a full shape, the ultimate load capacity is lower than that of “failure
curve until the post-peak lateral load of the system drops by 20%. of the beam end” and decreases sharply after reaching the ultimate
load, showing poor energy dissipation capacity. All specimens were
damaged as per “failure of the beam end,” except NC3 (n ¼ 0.3),
Table 5. Parameters of FE models NC4 (n ¼ 0.4), LB3 (K ¼ 0.312), and LB4 (K ¼ 0.281).
tC Height (λ) H c−w n Span (K) Fig. 13 and Table 7 show the backbone curves and eigenvalues
No. (mm) (mm) (mm) (N=mm2 ) (mm) for different parameters.
BASE 3 3,000 (23.22) 250 0.2 4,800 (0.351)
Fig. 13(a) shows a comparison of the backbone curves for differ-
HC1 3 2,700 (20.96) 250 0.2 4,800 (0.316) ent slenderness ratios. Improving the slenderness ratio reduced the
HC2 3 3,300 (25.62) 250 0.2 4,800 (0.389) initial stiffness, yielding load, and ultimate load of the CFS frame
HC3 3 3,600 (27.95) 250 0.2 4,800 (0.421) and significantly improved the yielding displacement, ultimate dis-
NC1 3 3,000 (23.22) 250 0 4,800 (0.351) placement, and failure displacement. This is because the P-Δ effect
NC2 3 3,000 (23.22) 250 0.1 4,800 (0.351) of the CFS frame with a larger slenderness ratio is more noticeable
NC3 3 3,000 (23.22) 250 0.3 4,800 (0.351) under cyclic loading. The load capacity and stiffness of the elastic
NC4 3 3,000 (23.22) 250 0.4 4,800 (0.351) stage and elastic-plastic stage are much lower than for those with
CSC1 3 3,000 (23.22) 220 0.2 4,800 (0.351) smaller slenderness ratios.
CSC2 3 3,000 (23.22) 280 0.2 4,800 (0.351)
According to Fig. 13(b), when the axial compression ratio in-
CSC3 3 3,000 (23.22) 300 0.2 4,800 (0.351)
CSC4 2.5 3,000 (23.22) 250 0.2 4,800 (0.351)
creases from 0 to 0.2, there is little effect on the ultimate load capac-
LB1 3 3,000 (23.22) 250 0.2 3,600 (0.468) ity and deformation of the CFS frame. When the axial compression
LB2 3 3,000 (23.22) 250 0.2 4,200 (0.401) ratio increases to 0.3 and 0.4, respectively, there is a significant
LB3 3 3,000 (23.22) 250 0.2 5,400 (0.312) effect on ultimate load capacity and deformation. Consequently, the
LB4 3 3,000 (23.22) 250 0.2 6,000 (0.281) axial compression ratio should be maintained within a specified
Note: tC = thickness of C shaped cross section steel; H c−w = height
range. The effect of axial compression ratio is not significant for
of column web; height of the beam web is 200 mm, and thickness of initial stiffness at the initial stage of loading; after the CFS frame
the gusset plate is 10 mm; n = axial compression ratio; λ = column yields, the axial compression has a significant influence on stiff-
slenderness ratio; and K = linear stiffness ratio. ness degradation, and the speed of stiffness degradation of the CFS
frame accelerates. This result is because the frame column, which
has a larger axial compression ratio, has a significant P-Δ effect
degree of freedom was limited to prevent the out-of-plane instability when the CFS frame generates lateral displacement.
of the C shaped cross section steel. Analysis of the skeletal curves of the height/thickness ratio
Cyclic loading of the FE model was mainly divided into two shows that if the height of the column web increases and the thick-
stages where the first stage was to apply a pretightening force to ness remains unchanged, then the bending stiffness of the column
the bolts and the second stage was to apply cyclic displacement to increases, leading to increased frame load capacity and stiffness.
the loading plate. The load capacity and displacement of the frame decrease as the

© ASCE 04019212-6 J. Struct. Eng.

J. Struct. Eng., 2020, 146(3): 04019212


Downloaded from ascelibrary.org by Indian Institute Of Technology Delhi on 11/17/20. Copyright ASCE. For personal use only; all rights reserved.

Fig. 7. FE model.

Table 6. Comparison of analysis and test data


Parameter Value
Specimen s-3
Ana
Ke 1.05
Test
Fig. 8. Bolted connection. Ana
FP 1.09
Test
Ana
du 1.21
Test
Ana
PE 1.39
Test
Note: ke = initial stiffness; FP = ultimate load capacity; du = failure
displacement; and PE = the energy of the input earthquake absorbed by
structures.

In addition, as the linear stiffness ratio of the CFS frame increases,


the speed of stiffness degradation also accelerates.

Ductility and Energy Dissipation


Table 8 presents the energy of the input earthquake absorbed by the
CFS frame, which ranges from 3.17 to 6.42, and the ductility co-
efficients range from 2.37 to 3.50. The results show that the CFS
frame has good ductility and energy dissipation capacity, as well as
Fig. 9. Comparison of S-3 load-displacement backbone curve good deformability.
verification.

Research on the Hysteretic Model

thickness of the web decreases and height remains constant, caus- Orthogonal Test Design
ing the bending stiffness of the column to decrease. Considering the four parameters of axial compression ratio, n;
Through observing the backbone curves of five specimens with slenderness ratio, λ; height/thickness ratio, D; and linear stiffness
different linear stiffness ratios, the load capacity and displacement at ratio, K; nine FE models of CFS frames (FEM1–FEM9) were es-
all stages decrease as the linear stiffness ratio decreases. Meanwhile, tablished by orthogonal test design methods and analyzed by the
changing the linear stiffness ratio has a significant effect on the stiff- FE method. Other parameters remained unchanged, and the param-
ness degradation of the CFS frame. Before the CFS frame yields, the eters of models are presented in Table 9.
linear stiffness ratio has a greater effect on the initial stiffness, and as Table 10 presents the ultimate load capacity and initial stiffness
the linear stiffness ratio increases, the initial stiffness also increases. extracted from the results of FE analysis.

© ASCE 04019212-7 J. Struct. Eng.

J. Struct. Eng., 2020, 146(3): 04019212


Downloaded from ascelibrary.org by Indian Institute Of Technology Delhi on 11/17/20. Copyright ASCE. For personal use only; all rights reserved.

Fig. 10. Failure mode verification: (a) proximal column base; and (b) distal column base.

Fig. 11. Load-drift hysteresis curve of “failure of the beam end.”


Fig. 12. Load-drift hysteresis curve of “failure of the column base.”

Regression Analysis of Load Capacity and 0


where Pmax = frame ultimate load capacity; n = axial compression
Stiffness Parameters ratio; λ = slenderness ratio; K = line stiffness ratio; and D = height/
To establish the hysteretic model, K 2 =K 1 (ratio of strengthened thickness ratio.
The thickness of C shaped cross section steel has a certain effect
stiffness to elastic stiffness), K 3 =K 1 (ratio of negative stiffness
on the load capacity that we did not consider in the orthogonal ex-
to elastic stiffness), Py =Pmax (ratio of yield load to ultimate load),
perimental analysis. Therefore, the horizontal limit load should be
and Pmax were determined, as shown in Table 11. The relationship
reduced accordingly. The reducing function is gained by regression
between the four parameters and these results from Table 11 was
analysis. We assume that
established by regression analysis.
0
The fitting equation for ultimate load capacity referring to Pmax ¼ fðtÞPmax ¼ fðtÞψðn; λ; D; KÞ ð4Þ
the function model in reference (Wang et al. 2015) is expressed as
follows: where fðtÞ ¼ α1 tα2 .
With regression analysis, we obtain α1 ¼ 0.098 and α2 ¼ 2.146.
0
Pmax ¼ ψðn; λ; K; DÞ ð1Þ Referring to the fitting equation (Qi et al. 2010) and using
Matlab software to perform multiple regression analysis on orthog-
onal data, the fitting equations for K 2 =K 1, K 3 =K 1 , Py =Pmax , and
where ψðn; λ; D; KÞ ¼ α1 þ α2 n þ α3 λ þ α4 D þ α5 K ð2Þ Pu =Pmax are obtained, as shown in Eqs. (5)–(8)

The confidence of multivariate nonlinear regression fitting on K 2 =K 1 ¼ 4.351n3.288 − 2.343λ−0.153


orthogonal data is 95%, thus þ 0.082D0.609 þ 0.207K −0.281 ð5Þ

ψðn; λ; D; KÞ ¼ 246.516 − 59.382n − 2.859λ K 3 =K 1 ¼ −ð0.252n−0.157 þ 0.260λ0.192


− 0.235D − 34.242K ð3Þ þ 0.216D−0.038 − 0.271K −0.183 Þ ð6Þ

© ASCE 04019212-8 J. Struct. Eng.

J. Struct. Eng., 2020, 146(3): 04019212


Downloaded from ascelibrary.org by Indian Institute Of Technology Delhi on 11/17/20. Copyright ASCE. For personal use only; all rights reserved.

Fig. 13. Load-drift backbone curves for different parameters: (a) different slenderness ratio; (b) different axial compression ratio; (c) different height/
thickness ratio; and (d) different linear stiffness ratio.

Table 7. Characteristics of backbone curves


Yield point Limit point Failure point Table 8. μ and PE of FE models
No. Δ=mm P=kN Δ=mm P=kN Δ=mm P=kN No. PE =kJ μ
HC1 19.29 83.48 30.00 94.27 60.77 80.13 HC1 5.66 3.22
BASE 21.09 70.84 45.00 82.30 67.12 69.95 BASE 3.88 3.18
HC2 27.27 66.23 45.00 73.59 66.75 62.56 HC2 4.38 2.45
HC3 31.67 64.58 60.00 65.00 75.09 61.80 HC3 4.63 2.37
NC1 25.42 77.42 45.00 88.18 82.11 75.51 NC1 6.42 3.21
NC2 24.65 76.96 45.00 86.96 79.53 74.09 NC2 6.33 3.38
NC3 18.92 70.03 30.00 76.59 56.40 65.32 NC3 3.80 3.10
NC4 16.96 58.40 30.00 68.34 47.37 58.09 NC4 3.17 2.79
CSC1 27.13 66.49 60.00 74.17 81.56 63.04 CSC1 4.33 3.01
CSC2 18.21 77.92 30.00 91.45 60.15 77.73 CSC2 4.61 3.27
CSC3 18.01 79.16 30.00 97.88 56.68 81.20 CSC3 5.55 3.50
CSC4 22.22 69.06 46.35 73.87 62.61 62.79 CSC4 4.26 2.69
LB1 19.65 83.03 27.00 91.89 57.83 79.56 LB1 4.96 2.94
LB2 20.97 78.09 28.5 85.62 62.47 72.78 LB2 4.88 2.97
LB3 22.14 67.35 45.35 78.67 68.95 69.07 LB3 4.68 3.11
LB4 23.71 64.92 46.00 72.57 65.01 61.25 LB4 4.36 2.74

© ASCE 04019212-9 J. Struct. Eng.

J. Struct. Eng., 2020, 146(3): 04019212


Table 9. Parameters of orthogonal test models
No. n λ D K
FEM1 0.1 37.75 250/3 0.35
FEM2 0.1 39.85 280/3 0.26
FEM3 0.1 42.51 300/3 0.22
FEM4 0.2 37.22 280/3 0.21
FEM5 0.2 37.65 300/3 0.25
FEM6 0.2 45.64 250/3 0.38
FEM7 0.3 35.25 300/3 0.20
FEM8 0.3 42.97 250/3 0.31
FEM9 0.3 43.50 280/3 0.27
Downloaded from ascelibrary.org by Indian Institute Of Technology Delhi on 11/17/20. Copyright ASCE. For personal use only; all rights reserved.

Fig. 14. P-Δ backbone model for the CFS frame.

Table 10. Characteristics of orthogonal test models


No. Ultimate load capacity (kN) Initial stiffness (kN=mm) 8
> K Δ 0 ≤ Δ ≤ Δy
FEM1 101.46 5.72 < 1
P ¼ fðΔÞ ¼ Py þ K 2 ðΔ − Δy Þ Δy ≤ Δ ≤ Δmax ð9Þ
FEM2 94.48 4.82 >
:
FEM3 88.74 4.29 Pmax þ K 3 ðΔ − Δmax Þ Δmax ≤ Δ ≤ Δu
FEM4 99.65 5.78
FEM5 93.73 5.28
According to Eq. (10)
FEM6 72.94 2.91
FEM7 100.41 6.96 P ¼ KeΔ ð10Þ
FEM8 73.97 3.72
FEM9 73.69 3.67
where K e = the initial stiffness of the frame.
Based on the simplified second-order elastic analysis methods,
the initial stiffness function of the CFS frame, K e , in this paper can
be calculated by
Table 11. Calculation results
   6ic N 2
No. K 2 =K 1 K 3 =K 1 Py =Pmax Pmax 12ic 6N  H þ 10
K e ¼ 2fðic ; ib ; N; HÞ ¼ 2 2
 2 ð11Þ
FEM1 0.14 −0.09 0.78 101.46 H 5H 4ic þ6ib  2NH
15
FEM2 0.29 −0.07 0.77 94.48 qffiffiffiffiffiffi qffiffiffiffiffiffiffiffiffi
FEM3 0.36 −0.07 0.73 88.74 If u ¼ H EINc ¼ nNicu H, then
FEM4 0.30 −0.10 0.87 99.65
FEM5 0.37 −0.10 0.84 93.73 N NH
FEM6 0.24 −0.15 0.75 72.94 u2 ¼ H 2 ¼ ð12Þ
EI c ic
FEM7 0.33 −0.10 0.72 100.41
FEM8 0.33 −0.14 0.75 73.97
FEM9 0.34 −0.15 0.75 73.69 By substituting Eq. (12) into Eq. (11), the simplified equation
can be expressed as follows:
3EI c
Ke ¼ ψ1 ðu; KÞ ð13Þ
Py =Pmax ¼ 0.051n−0.033 þ 1.215λ−0.357 H3

þ 0.111D0.280 þ 0.061K 3.981 ð7Þ where

48 − 6.4u2 þ 0.16u4 þ 72ð1 − 0.01u2 Þ


Pu =Pmax ¼ 0.85 ð8Þ ψ1 ðu; KÞ ¼ ð14Þ
6 þ 9K − 0.2u2
where K 1 , K 2 , and K 3 = elastic stiffness, strengthened stiffness, and
With reference to the results of reference (Shi 2006), the initial
negative stiffness of the CFS frame, respectively.
frame stiffness is greater than the elastic stiffness when the frame
Py , Pmax , and Pu = yield load, horizontal limit load, and failing
undergoes lateral displacement. Therefore, K 1 ¼ K e =α, where η
load of the CFS frame, respectively.
represents the reduction coefficient obtained by regression analysis,
and α ¼ 0.769, meaning that K 1 ¼ 0.769K e . In addition, since the
Hysteretic Models of Backbone Curve influence of the thickness of C shaped cross section steel on the
load capacity is not considered, the reduction function of the thick-
In this paper, a multilinear model is used to simulate the P-Δ back- ness of C shaped cross section steel obtained through regression
bone curve of the CFS frame (Guo and Yang 2004). A simplified analysis can be written as fðtÞ ¼ 0.444t0.746 .
model of the backbone curve is shown in Fig. 14. Line OA is the The ultimate elastic stiffness of the CFS frame can be expressed
elastic stage, line AB is the elastic-plastic stage, and line BC is the as follows:
plastic deformation stage. Point A corresponds to the yield load,
Py ; point B corresponds to the ultimate load, Pmax ; and point C K 1 ¼ αfðtÞK e ð15Þ
corresponds to the failure load, 0.85Pmax . K 1 is the elastic stiffness,
K 2 is the strengthened stiffness, and K 3 is the negative stiffness. where fðtÞ ¼ 0.444t0.746 , α¼0.769.
A simplified equation for the three-line P-Δ backbone curve of As discussed, once the five parameters K 1 , Pmax , Py =Pmax ,
the CFS frame can be expressed as follows: K 2 =K 1 , and K 3 =K 1 are determined, the seven parameters K 2 , K 3 ,

© ASCE 04019212-10 J. Struct. Eng.

J. Struct. Eng., 2020, 146(3): 04019212


Fig. 15. Calculation model of load-off stiffness.
Downloaded from ascelibrary.org by Indian Institute Of Technology Delhi on 11/17/20. Copyright ASCE. For personal use only; all rights reserved.

Fig. 17. Resilience model of the hysteresis curve.

Fig. 16. Load-off stiffness fitting.

Py , Pu , Δy , Δmax , and Δu can be derived from the five parameters,


and the P-Δ backbone curve of the CFS frame can be determined.

Hysteretic Model of Hysteresis Curve Fig. 18. Comparison of simplified and test load-displacement back-
Prior to the CFS frame reaching the yield stress point, the loading bone curve model.
stiffness and unloading stiffness are consistent with the elastic
stiffness, K 1 , without degradation. When the CFS frame enters the
elastic-plastic stage, the loading stiffness decreases to K 2 , while the
Hysteretic Model Validation of Backbone Curve and
load-off stiffness K 4 is not consistent with the elastic stiffness K 1
Hysteresis Curve
and decreases with the increased number of cycles and loading
displacement. Fig. 18 and Table 12 present a comparison of the simplified and test
The load-off stiffness calculation model (Watson et al. 1994; backbone curves. They both indicate the eigenvalue error is within
Watson and Park 1994) is shown in Fig. 15. The method for deter- 10% for all stages, except failure displacement. This indicates that
mination is as follows: The peak load of the hysteresis loop is point the simplified backbone curve model reflects the entire process of
A, and the intersection of the unloading curve and the Δ axis is load-displacement change for the CFS frame.
point B. The slope of line AB is the load-off stiffness. The relation- A comparison of the simplified and the test hysteresis curves is
ship of K 4 =K 1 (load-off stiffness/elastic stiffness) and Δj =Δy shown in Fig. 19. Before the CFS frame yields, the relationship
(post-yield displacement/yield displacement) is acquired by regres- between load and displacement is linear; the loading stiffness is
sion analysis, as shown in Eq. (16) the same as the unloading stiffness, and the area of the hysteresis
loop is zero. When the CFS frame enters the elastic-plastic stage,
 −0.299
Δj the positive loading and negative loading stiffness decrease and
K 4 =K 1 ¼ ð16Þ the area of the hysteresis loop increases. After entering the failure
Δy
stage, the load capacity decreases. The presented analysis illustrates
that the simplified hysteresis model essentially matches the test
where Δj = displacement after yield of the CFS frame; and Δy =
hysteresis curve.
yield displacement of the CFS frame. Fig. 16 shows the fitting
curve for the load-off stiffness.
Fig. 17 shows the hysteretic model of the hysteresis curve of Conclusion
the CFS frame under quasi-static load. All key parameters can be
obtained, and the hysteresis curve of the CFS frame can be defined In order to investigate the seismic behavior of the CFS frame, rel-
by Eqs. (1)–(16). evant tests and numerical analyses were performed on 21 frames

© ASCE 04019212-11 J. Struct. Eng.

J. Struct. Eng., 2020, 146(3): 04019212


Table 12. Comparison of test and calculation results
Yield Yield Ultimate Ultimate Failure Failure
load, Py displacement, Δy load, Pmax displacement, Δmax load, Pu displacement, Δu
Comparison result (kN) (mm) (kN) (mm) (kN) (mm)
Test result 46.69 10.34 58.57 19.43 49.78 25.51
Calculation result 51.70 9.65 61.54 21.29 52.31 35.58
T.R./C.R. 0.90 1.07 0.95 0.91 0.95 0.72
Note: T.R. = test result; and C.R. = calculation result.

the CFS frame, and is in good agreement with test results. This is
Downloaded from ascelibrary.org by Indian Institute Of Technology Delhi on 11/17/20. Copyright ASCE. For personal use only; all rights reserved.

an important foundation for seismic analysis of the CFS frame.

Acknowledgments
This work was supported by the China National Natural Science
Foundation Program (Project Nos. 51368043 and 51768055), the
Inner Mongolia Natural Science Foundation Program (Project No.
2017MSLH0723), and the Excellent Youth Fund of Inner Mongolia
University of Science and Technology (Project No. 2016YQL10).

References
Fig. 19. Comparison of simplified and test load-displacement hyster- AISI (American Iron and Steel Institute). 2015. North American standard
esis curves. for seismic design of cold-formed steel structural systems. AISI S400.
Washington, DC: AISI.
Ali, B. A., S. Saad, and M. H. Osman. 2010. “Cold-formed steel frame with
bolted moment connections.” Int. J. Civ. Struct. Eng. 1 (3): 534–544.
with varying parameters, including the thickness of gusset plate,
https://doi.org/10.6088/ijcser.00202010043.
axial compression ratio, column slenderness ratio, beam/column ASTM. 2011. Standard test methods for cyclic (reversed) load test
linear stiffness ratio, and height/thickness ratio of the column web. for shear resistance of walls for buildings. ASTM E2126. West
The seismic behavior was summarized by the hysteresis curve and Conshohocken, PA: ASTM.
the backbone curve of the CFS frame, and the hysteretic models ASTM. 2014. Standard testing methods and definitions for mechanical
for the CFS frame were obtained. testing of steel products. ASTM A370-13. West Conshohocken, PA:
According to the presented analyses, the conclusions and design ASTM.
recommendations are given as follows: Bagheri Sabbagh, A., M. Petkovski, K. Pilakoutas, and R. Mirghaderi. 2012.
1. There are two failure modes for the CFS frame: “failure of the “Experimental work on cold-formed steel elements for earthquake resil-
ient moment frame buildings.” Eng. Struct. 42 (Sep): 371–386. https://
beam end” follows the “strong column weak beam” seismic de-
doi.org/10.1016/j.engstruct.2012.04.025.
sign principles. In contrast, “failure of the column base” does not Bučmys, Ž., and A. Daniūnas. 2015. “Analytical and experimental in-
conform to design principles. The beam/column linear stiffness vestigation of cold-formed steel beam-to-column bolted gusset-plate
ratio and the axial compression ratio are control parameters. In joints.” J. Civ. Eng. Manage. 21 (8): 1061–1069. https://doi.org/10
order to avoid “failure of the column base,” the axial compres- .3846/13923730.2015.1084039.
sion ratio should be less than 0.3 and the beam/column linear Bučmys, Ž., A. Daniūnas, J.-P. Jaspart, and J.-F. Demonceau. 2018. “A
stiffness ratio of the CFS frame should be greater than 0.32. component method for cold-formed steel beam-to-column bolted gusset
2. Varying the analysis parameters has different degrees of influ- plate joints.” Thin Walled Struct. 123 (Feb): 520–527. https://doi.org/10
ence on the seismic behavior of the CFS frame. When the web .1016/j.tws.2016.10.022.
thickness of the column decreases, the CFS frame ductility, Chinese Standard. 2017. Standard for design of steel structure.
[In Chinese.] GB 50017. Beijing: China Architecture & Building Press.
equivalent viscous damping coefficient, and stiffness decrease.
Code of China. 2002. Technical code of cold-formed thin-wall steel struc-
Improving the web height of the column reduces the CFS frame tures. [In Chinese.] GB 50018. Beijing: China Planning Press.
ductility, equivalent viscous damping coefficient, and stiffness. Dubina, D. 2008. “Behavior and performance of cold-formed steel-framed
When the beam/column linear stiffness ratio of the CFS frame houses under seismic action.” J. Constr. Steel Res. 64 (7–8): 896–913.
decreases, the initial stiffness decreases and the energy dissipa- https://doi.org/10.1016/j.jcsr.2008.01.029.
tion capacity first increases and then decreases. Dundu, M., and A. R. Kemp. 2006. “Strength requirement of single cold-
3. The trend of stiffness degradation is similar for every frame, and formed channels connected back-to-back.” J. Constr. Steel Res. 62 (3):
the stiffness degradation range is essentially the same. The duc- 250–261. https://doi.org/10.1016/j.jcsr.2005.07.006.
tility coefficient, μ, is between 2.45 and 4.97, and the equivalent Fiorino, L., O. Iuorio, and R. Landolfo. 2014. “Designing CFS structures:
viscous damping coefficient, he , is between 0.28 and 0.36. The new school BFS in Naples.” Thin Wallled Struct. 78 (Jan): 37–47.
https://doi.org/10.1016/j.tws.2013.12.008.
Compared with reinforced concrete structures and H-shaped
Fiorino, L., S. Shakeel, V. Macillo, and R. Landolfo. 2018. “Seismic
steel frame structures, the CFS frame has excellent ductility and response of CFS shear walls sheathed with nailed gypsum panels:
energy dissipation capabilities. Numerical modeling.” Thin Wallled Struct. 122 (Jan): 359–370.
4. Considering the influence of slenderness ratio, height/thickness https://doi.org/10.1016/j.tws.2017.10.028.
ratio of the column web, axial compression ratio, and beam/ Fiorino, L., M. T. Terracciano, and R. Landolfo. 2016. “Experimental in-
column linear stiffness ratio, a hysteretic model is established for vestigation of seismic behaviour of low dissipative CFS strap-braced

© ASCE 04019212-12 J. Struct. Eng.

J. Struct. Eng., 2020, 146(3): 04019212


stud walls.” J. Constr. Steel Res. 127 (Dec): 92–107. https://doi.org/10 J. Constr. Steel Res. 66 (3): 470–479. https://doi.org/10.1016/j.jcsr
.1016/j.jcsr.2016.07.027. .2009.10.007.
Guo, Z. X., and Y. Yang. 2004. “State-of-the-art of restoring force models Schafer, B. W., et al. 2016. “Seismic response and engineering of cold-
for RC structures.” [In Chinese.] World Earthquake Eng. 20 (4): 47–51. formed steel framed buildings.” Structures 8 (Part 2): 197–212. https://
Lim, J. B. P., and D. A. Nethercot. 2003. “Ultimate strength of bolted doi.org/10.1016/j.istruc.2016.05.009.
moment-connections between cold-formed steel members.” Thin Walled Shamim, I., J. DaBreo, and C. A. Rogers. 2013. “Dynamic testing of single-
Struct. 41 (11): 1019–1039. https://doi.org/10.1016/S0263-8231(03) and double-story steel-sheathed cold-formed steel-framed shear walls.”
00045-4. J. Struct. Eng. 139 (5): 807–817. https://doi.org/10.1061/(ASCE)ST
Lim, J. B. P., and D. A. Nethercot. 2004a. “Finite element idealization of a .1943-541X.0000594.
cold-formed steel portal frame.” J. Struct. Eng. 130 (1): 78–94. https:// Shi, W. L. 2006. Experimental and theoretical study on semi-rigid beam-to-
doi.org/10.1061/(ASCE)0733-9445(2004)130:1(78). column composite joints with flush end-plate connection. Shanghai,
Lim, J. B. P., and D. A. Nethercot. 2004b. “Stiffness prediction for bolted China: Tongji Univ.
moment-connections between cold-formed steel members.” J. Constr. Wang, Z. S., M. Z. Su, H. Ma, and X. Z. Pan. 2015. “Study on load-
Downloaded from ascelibrary.org by Indian Institute Of Technology Delhi on 11/17/20. Copyright ASCE. For personal use only; all rights reserved.

Steel Res. 60 (1): 85–107. https://doi.org/10.1016/S0143-974X(03) displacement restoring force model of single-story single-bay ling-
00105-6. weight portal frame with variable cross-sections.” [In Chinese.] Build.
Lin, S. H., C. L. Pan, and W. T. Hsu. 2014. “Monotonic and cyclic loading Struct. 44 (15): 12–16. https://doi.org/10.19701/j.jzjg.2015.15.003.
experiments for cold-formed steel wall frames sheathed with calcium
Watson, S., and R. Park. 1994. “Simulated seismic load experiments on
silicate board.” Thin Walled Struct. 74 (Jan): 49–58. https://doi.org/10
reinforced concrete columns.” J. Struct. Eng. 120 (6): 1825–1849.
.1016/j.tws.2013.09.011.
https://doi.org/10.1061/(ASCE)0733-9445(1994)120:6(1825).
Moghimi, H., and H. R. Ronagh. 2009. “Performance of light-gauge cold-
Watson, S., F. A. Zahn, and R. Park. 1994. “Confined reinforcement for
formed steel strapbraced stud walls subjected to cyclic loading.” Eng.
Struct. 31 (1): 69–83. https://doi.org/10.1016/j.engstruct.2008.07.016. concrete columns.” J. Struct. Eng. 120 (6): 1798–1824. https://doi.org
Mohebbi, S., S. R. Mirghaderi, F. Farahbod, A. B. Sabbagh, and S. Torabian. /10.1061/(ASCE)0733-9445(1994)120:6(1798).
2016. “Experiments on seismic behavior of steel sheathed cold-formed Wrzesien, A., J. B. P. Lim, and D. A. Nethercot. 2012. “Optimum joint
steel shear walls cladded by gypsum and fiber cement boards.” Thin detailed for a general cold formed steel portal frame.” Adv. Struct. Eng.
Walled Struct. 104 (Jul): 238–247. https://doi.org/10.1016/j.tws.2016 15 (9): 1623–1639. https://doi.org/10.1260/1369-4332(2016).15.9.1623.
.03.015. Xu, Z., Z. Chen, B. H. Osman, and S. Yang. 2018. “Seismic performance
Mojtabaei, S. M., M. Z. Kabir, I. Hajirasouliha, and M. Kargar. 2018. of high-strength lightweight foamed concrete-filled cold-formed steel
“Analytical and experimental study on the seismic performance of shear walls.” J. Constr. Steel Res. 143 (Apr): 148–161. https://doi
cold-formed steel frames.” J. Constr. Steel Res. 143 (Apr): 18–31. .org/10.1016/j.jcsr.2017.12.027.
https://doi.org/10.1016/j.jcsr.2017.12.013. Yu, W. K., K. F. Chung, and M. F. Wong. 2005. “Analysis of bolted moment
Qi, J. J., L. Z. Jiang, C. Z. Zhang, and Z. W. Yu. 2010. “Effects of interface connections in cold formed steel beam–column sub-frames.” J. Constr.
slip, vertical uplift and shear deformation on dynamic behavior of steel- Steel Res. 61 (9): 1332–1352. https://doi.org/10.1016/j.jcsr.2005.03.001.
concrete composite continuous beams.” [In Chinese.] J. Cent. South Zhou, X. H., Y. J. He, L. Xu, and Q. S. Zhou. 2009. “Experimental study
Univ. (Sci. Technol.) 41 (6): 2334–2343. and numerical analyses on seismic behaviors sheathed with calcium
Sabbagh, A. B., R. Mirghaderi, M. Petkovski, and K. Pilakoutas. 2010. “An silicate board.” Thin Walled Struct. 47 (11): 1343–1353. https://doi.org
integrated thin-walled steel skeleton structure (two full scale tests).” /10.1016/j.tws.2009.03.007.

© ASCE 04019212-13 J. Struct. Eng.

J. Struct. Eng., 2020, 146(3): 04019212

Вам также может понравиться