Zhang2020 PDF

You might also like

Download as pdf or txt
Download as pdf or txt
You are on page 1of 12

Journal of Constructional Steel Research 168 (2020) 105981

Contents lists available at ScienceDirect

Journal of Constructional Steel Research

Mechanical-property tests on assembled-type light steel modular house


Jun-Feng Zhang a, Jun-Jie Zhao a, Da-Yong Yang a, En-Feng Deng a,⁎, Huan Wang b, Shi-Yun Pang a,
Li-Ming Cai c, Shu-Cai Gao c
a
School of Civil Engineering, Zhengzhou University, Zhengzhou 450001, China
b
School of Mechanical Engineering, Zhengzhou University, Zhengzhou 450001, China
c
The Architectural design and research institute of Henan Province CO., Ltd, Zhengzhou 450014, China

a r t i c l e i n f o a b s t r a c t

Article history: The assembled-type light steel (ATLS) modular house—a modular building and an emerging type of eco-friendly
Received 10 August 2019 architecture—consists of a bottom frame and a top frame that are connected with columns and high-strength
Received in revised form 17 December 2019 bolts. It has the advantages of superior quality, a shorter construction time, and less resource wastage compared
Accepted 5 February 2020
with traditional construction. In this study, two full-scale models were tested to obtain the vertical bearing capac-
Available online xxxx
ity and the horizontal stiffness. A three-dimensional nonlinear finite-element (FE) model was established using
Keywords:
the general-purpose FE software ABAQUS and validated against the test results. The load–displacement and load–
Modular buildings stress curves were analysed and compared with the result of finite element analysis (FEA). The results indicated
Light steel modular house that the bottom frame beam of the ATLS modular house had a good deformation capacity and low rigidity. The
Vertical bearing capacity test horizontal stiffness of the ATLS modular unit was 1.21kN/mm and the results of FEA exhibited a minor difference
Horizontal stiffness test about 1.65% comparing with the test result. The buckling mode and load–displacement curves in the FE analysis
Mechanical properties exhibited good agreement with the test results.
© 2020 Elsevier Ltd. All rights reserved.

1. Introduction [16] proposed a bolted connection for connecting the cruciform gusset
plate and the adjacent units. Deng et al. [17] introduced an innovative
Modular steel construction—a construction technique that involves connection using socket-shaped tenons and a cruciform gusset plate
prefabrication in the factory and subsequent installation onsite—has to connect the module units. Chen et al. [18] proposed a new type of
been increasingly adopted by engineers owing to its superiority with re- design with beam-to-beam bolted connections. Additionally, finite-
gard to construction speed and quality. Recently, this new installation element analysis (FEA) has been adopted to study the mechanical prop-
method has been adopted in many areas, such as hotels, apartments, erties and failure modes of new types of connections [19–21]. Ding et al.
lodging houses, hospitals, offices, and other similar buildings with re- [22] investigated the effect of a corrugated steel plate shear wall with an
petitive units. Modular steel constructions are superior to traditional opening on the mechanical properties of a modular building. By
buildings with regard to the environmental impact, construction time, performing a hysteresis load test, Dai et al. [23] studied the strengthen-
energy consumption, light deadweight, and bearing capacity [1–4]. It ing effect of a steel strip in a modular building structure. Dhanapal et al.
has been verified that modular buildings satisfy global demands more [24] introduced a new type of innovative modular steel structure and
effectively than traditional constructions [5–9], and the applications of investigated the structural properties of typical angle joints with this
typical modular steel buildings are illustrated in Fig. 1(a) and (b). More- modular structure. However, most studies aimed to obtain the mechan-
over, the module development and performance evaluation of middle ical properties of connections and the strengthening effect of the wall-
and high-rise modular buildings have attracted the attention of many board; limited experimental studies have been performed on the
construction companies and scholars [10–12]. global mechanical properties of the module unit.
Lately, many studies have focused on modular steel constructions, As illustrated in Fig. 2, the assembled-type light steel (ATLS) modu-
including the structural performance under monotonic loads, hysteresis lar house is a modular construction with a light steel structure as its
loads, and seismic loads, as well as the requirements for the application framework and a sandwich board as its enclosure. Asymmetrical open
of the construction methods [9,13–15]. To improve the mechanical sections are adopted for the beams at the top and bottom of the column.
properties of the connections of modular steel construction, Park et al. All components of the ATLS modular house are manufactured in a fac-
tory and installed on a construction site. In contrast to traditional
modular-structure buildings, the frame and enclosure of the ATLS mod-
⁎ Corresponding author. ular house can be demolished repeatedly. Thus, the modular construc-
E-mail address: dengenfeng@zzu.edu.cn (E.-F. Deng). tion is convenient for transportation.

https://doi.org/10.1016/j.jcsr.2020.105981
0143-974X/© 2020 Elsevier Ltd. All rights reserved.
2 J.-F. Zhang et al. / Journal of Constructional Steel Research 168 (2020) 105981

Fig. 1. Modular construction.

In this study, two identical full-size models were tested to study the strength (fy), and elongation (Elo). Additionally, the mechanical proper-
mechanical properties of the ATLS modular house under vertical and ties of the bolts were tested in a material-property test. The average ma-
horizontal loads. The load–displacement curves and load–stress curves terial properties are presented in Table 3. The ratio between the yield
were analysed, and the results of FEA were compared with the test re- strength and ultimate tensile strength (fy/fu) should be ≤0.85 and the
sults. The study provided useful information regarding the ultimate elongation (Elo) should be N20% according to the Tensile Test of Metallic
strength and stiffness of ALTS modular buildings, serving as a valuable Material (GB/T 228.1-2010) [25].
reference for further investigations.
3. Vertical bearing capacity test
2. Test specimens
3.1. Loading program and instrumentation
2.1. Geometric dimension
To induce the most unfavourable load conditions, the floor was di-
Two identical full-scale models were tested. As illustrated in Fig. 3 vided into 128 areas to apply the floor load about 3.19 kN/mm2, consid-
(a), a full-scale model with dimensions of 6055 mm × 2500 mm × ering the most unfavourable combination of dead load and live load (i.e.
2800 mm (length × width × height) was used. The joint between the 1.2 × dead load + 1.4 × live load) in accordance with the Load Code for
column and corner fitting is illustrated (a simplified 3D model) in the Design of Building Structures (GB 50009-2012) [26], as illustrated in
Fig. 3(b). The longitudinal beam and transverse beam were welded Fig. 4. The Then, the vertical load was applied to the top of the four col-
with corner fittings to form the top and bottom frames. The top and bot- umns simultaneously using hydraulic jacks in a professional laboratory.
tom frames were bolted by the column, and the column was bolted with To avoid impact damage of the corner fittings, there was a diaphragm
corner fittings. To apply a floor load conveniently, a calcium silicate between the corner fitting and the cylinder. Additionally, preloading
board was tiled on the bottom frame. The nominal thickness of all the was necessary before the nominal test. Step loading and continuous
components was 4 mm, and the measured dimensions of the compo- loading were employed during the vertical bearing capacity test. Every
nents are presented in Tables 1 and 2. Fig. 3(c) and (d) illustrate the step of the vertical load on the column, which was less than 20% of
bolt (M12-nut free bolt) and the arrangement of the bolts, respectively. the yield bearing capacity of the column, was imposed and sustained
The length of the bolt was 60mm and the cap of the bolt was 20mm. The for 10 min to inspect the components for signs of breakage, yielding,
columns were denoted as C1, C2, C3, and C4 (counterclockwise), as and buckling. Besides, the model was fixed on the laboratory bench by
shown in Fig. 3(e). The beams were denoted as B1 and B2. Fig. 3 the high strength bolts at the four corner fittings.
(f) and (g) illustrate the section profiles of the beams and columns, The ATLS modular unit was assembled to allow for the characteriza-
respectively. tion of the mechanical behaviour of the components and the unit. As il-
lustrated in Fig. 5, displacement transducers and strain rosettes were
used to record the displacement and the stress changes of the ATLS
2.2. Material properties modular unit. For example, “S1-C1-A” represents the monitoring point
at the middle height of column C1 for recording the stress, where “S1”
All the specimens were manufactured from Q345 steel with a nom- represents the strain gauge, “C1” represents the column, and “A” repre-
inal yield strength (fn) of 345 MPa. Three tensile coupons were tested to sents the position of the monitoring point. In this study, the stress was
determine the Young's modulus (E), ultimate tensile strength (fu), yield the maximum principle stress of the monitoring points. Additionally,
“C1-2” represents the displacement monitoring point at the top end of
the column, where “C1” represents the column and “2” represents the
position of the monitoring point.

3.2. Test phenomenon and observations

The deflection of the bottom frame beam increased with an increase


in the floor load. Additionally, a gap and excursion between the column
and the corner fitting was observed under the floor load. Fig. 6(a) illus-
trates the slight deflection (about 23.75mm) of the bottom frame beam
at the termination of the floor loading. Because of the bolted connection
Fig. 2. ATLS modular house. between the corner fitting and the column, the bolts would be stretched
J.-F. Zhang et al. / Journal of Constructional Steel Research 168 (2020) 105981 3

Welded Column

Bolted
Beam

Beam
Corner fitting

(a) Full-scale model (b) Joint between column and corner fitting

46

150
5-M12

46
75 75
210

(c) Bolt (d) The arrangement of the bolts


A D3
B C
D

D4
C3
C4 B2 t
H2
t D1
H

C1 C2

D2
B1
G

H1
E
F

(e) Components (f) Section of beam (g)Section of column


Fig. 3. ATLS modular house and details of the test specimens.

Table 1 when the beam produced a deflection. Thus, there was a gap between
Geometric size of the beams (units: mm). the column and the corner fitting when the floor load reached 2.93
Type L t H A B C D E F G kN/m2, as shown in Fig. 6(b). At this time, columns C2 and C3 produced
an excursion between the column and the corner fitting. With an in-
Nominal 5638.0 4.00 150.0 95.0 8.00 60.0 20.0 60.0 95.0 46.0
Measured B1 5637.5 3.86 148.5 93.8 8.15 58.5 19.8 58.3 94.5 45.5 crease in the floor load, the deformation of columns C2 and C3 became
Measured B2 5637.4 3.91 147.5 94.2 8.17 58.7 19.7 58.5 94.4 45.3 more visible. Furthermore, the deformation of columns C1 and C4
started to be observed.

Table 2
Geometric size of the columns (units: mm).

Type L t H1 H2 D1 D2 D3 D4

Nominal 2450.0 4.00 210.0 152.0 25.0 30.0 30.0 25.0


Measured C1 2447.5 3.88 208.1 151.4 25.7 30.1 30.8 26.1
Measured C2 2448.7 3.87 209.6 151.3 25.7 30.3 30.7 25.7
Measured C3 2447.5 3.91 208.2 151.6 27.5 31.1 30.4 26.7
Measured C4 2448.5 3.97 206.1 151.5 25.7 30.3 30.1 26.2

Table 3
Average measured material properties.

Type E (GPa) fy (MPa) fu (MPa) fy/fu Elo. (%)

Specimens 208 365 510 0.72 34


Bolts 206 640 800 0.8 /
Fig. 4. Test setup.
4 J.-F. Zhang et al. / Journal of Constructional Steel Research 168 (2020) 105981

75
(C4-2)C1-2 C2-2 (C3-2)
(C4-1) C1-1 C2-1(C3-1)

1325
2800
(C4-4)C1-4 C2-4(C3-4)
(C4-3)C1-3 C2-3(C3-3)

1325
B1(B2)
(C4-6)C1-6 C2-6(C3-6)
3027.5 3027.5
(C4-5)C1-5 C2-5 (C3-5)

75
6055
(a) Arrangement of displacement transducers

1400
2800
A H

1240
E (R) O G(N)
B (Q) I (M)

160
50 2767.5 2767.5 50
6055

(b) Strain rosettes


25 25
30

60

1 1
3
30

2 2
3 2 1
150

3 10 10 10
30

3
60

2 1
30 152 210 30

(c) Strain rosettes of beam section (d) Strain rosettes of column section
Fig. 5. Measurement system.

The vertical load was applied to the column as the floor loading was exhibit lateral flexural-torsional buckling. Evident flexural-torsional
terminated. Both the gap between the column and the bottom frame buckling of columns C1, C2, and C3 was observed, as illustrated in
beam and the deflection of the bottom frame decreased with an in- Fig. 7(a)–(c), when the vertical load on the column reached 182.89
crease in the vertical load on the column. Additionally, with the increase kN. However, for column C4, visible flexural-torsional buckling occurred
in the vertical load on the column, the four columns gradually started to when the vertical load on the column reached 115.46 kN, as illustrated

Fig. 6. Experimental phenomenon under floor load.


J.-F. Zhang et al. / Journal of Constructional Steel Research 168 (2020) 105981 5

Fig. 7. Column deformation under the vertical load on the column.

Fig. 8. Residual plastic deformation.

in Fig. 7(d). Ultimately, the vertical bearing capacity test was stopped load. When the floor load reached 2.5 kN/m 2 i.e. 1.0 × dead load
owing to the excessive deformation of the four columns. Fig. 8(a) and + 1.0 × live load, the deflection of the bottom frame beam was
(b) illustrate the residual plastic deformation of the four columns and 18.36mm approximately 1/330 of the span of the bottom frame
the corner fitting, respectively. beam, meeting the requirement of JGJ 227-2011 [27]. The gap be-
tween the column and the corner fitting would reduce under the
3.3. Test results and discussion vertical load on the column (note: the floor load remained as
3.19kN/mm2) and intensified the constrains at the beam ends, caus-
The load–displacement curves of the bottom frame beam are il- ing a decrease of the flexural deformation at the midspan of the
lustrated in Fig. 9. As illustrated in Fig. 9(a), the deformation of beam, as shown in Fig. 9(b). When the vertical load on the column
the bottom frame beam was in the elastic stage under the floor reached 211.68 kN, the deflection of bottom frame beam was

3.5 250
3.0 200
Load (kN/m2)

2.5
Load (kN)

2.0 150
1.5 100
1.0
50
0.5
0.0 0
0 10 20 30 0 10 20 30
Displacement (mm) Displacement (mm)
(a) Load-displacement curve under floor (b) Load-displacement curve under
load floor load and vertical load on
the column

Fig. 9. Load-displacement curve at the midspan of the bottom frame beam.


6 J.-F. Zhang et al. / Journal of Constructional Steel Research 168 (2020) 105981

4.0 250
200
3.0

Load (kN/m2)

Load (kN)
150
2.0
100
1.0 50
0.0 0
0 50 100 150 0 50 100 150
Stress(N/mm2) Stress(N/mm2)
(b) Load-displacement under floor load
(a) Load-stress curve under floor load
and vertical load on the column

Fig. 10. Load-stress curve at point S1-B1-O of the bottom frame beam.

3.5 250
3.0 E1
200 E2
2.5 E3
Load (kN/m2)

Load (kN)
2.0 E1 150 N1
E2 N2
1.5 E3 100 N3
1.0 N1
N2 50
0.5
N3
0.0 0
0 50 100 150 200 0 100 200 300
Stress (N/mm2) Stress (N/mm2)

(b) Load-stress curves under vertical


(a) Load-stress curves under floor load
load on column

Fig. 11. Load-stress curves at points of E and N of the bottom frame beam.

250 250

200 200
Load (kN)

Load (kN)

150 150

100 C1-3 100 C3-3


C1-4 C3-4
50 C2-3 50 C4-3
C2-4 C4-4
0 0
-10 0 10 20 30 -20 0 20 40
Displacement (mm) Displacement (mm)
(a) Load-displacement curves of (b) Load-displacement curves of
columns C1 and C2 columns C3 and C4

Fig. 12. Load-displacement curves of columns C1–C4 under the vertical load on the columns.

250 250

200 200
Load (kN)

150
Load (kN)

150 I1
A1 I2
100 A2 100 I3
A3
H1 B1
50 50 B2
H2
H3 B3
0 0
0 100 200 300 0 100 200 300 400
Stress (N/mm2) Stress (N/mm2)
(a) Load-stress curves at the middle (b) Load-stress curves at the bottom end
height points of column C1 and C2 points of C1 and C2

Fig. 13. Load-stress curves at key points under the vertical load on the columns.
J.-F. Zhang et al. / Journal of Constructional Steel Research 168 (2020) 105981 7

75
(C4-1)C1-1 C2-1 (C3-1)

1325
(C4-2)C1-2 C2-2 (C3-2)

2800
(C4-3)C1-3 C2-3 (C3-3)

1325
(C4-4)C1-4 C2-4 (C3-4)

6055

75
(a) Arrangement of displacement transducers

D(J) F(K)

2790
A(R) C(Q) G(L) H(N)

50 50
6055

(b) Arrangement of strain rosettes at key points


30

25 25
60

1 1
30

2 2 3
3 2 1
150

3 10 10 10
30

3
2 1
60

30 152 210 30

(c) Beam section (d) Column section

Fig. 14. Measurement system.

11.64 mm, which was approximately 51% smaller than the maxi- was approximately 56.73 N/mm2, which was reduced by approximately
mum deflection under the floor load. 53% compared with the stress under the floor load.
Additionally, the load–stress curves at point S1-B1-O of the bottom Furthermore, the load–stress curves for points E and N at the end of
frame beam were obtained. Fig. 10(a) illustrates the load–stress curve the bottom frame beam are illustrated in Fig. 11. As illustrated in Fig. 11
of the bottom frame beam under the floor load. The results indicate (a), the stress at points E and N remained in the elastic stage and
that the bottom frame beam exhibited elastic properties under the the maximum stress occurred at point S2-B1-E (approximately 125.16
floor load. The maximum stress of the bottom frame beam under the N/mm2). Fig. 11(b) illustrates the stress at points E and N under the
floor load was approximately 120.49 N/mm2. Similar to the load– vertical load on the column (note: the vertical load remained as
displacement curve of the bottom frame beam, Fig. 10(b) reveals the re- 3.19kN/mms2). Because of the gap between the column and the corner
duction of the deflection of the bottom frame beam under the vertical fitting, the vertical load on the column couldn’t transmit to the beam
load on the column (note: the floor load remained as 3.19kN/mm2). end regularly between the initial stage and the later period, leading to
When the vertical load on the column reached 211.68 kN, the stress the rapid shift, as shown in Fig. 11(b). In contrast to the load–stress
curve in Fig. 10(b), the stresses at points E and N increased with an
increase under the vertical load on the column. The maximum stress
occurred at point S2-B1-E (approximately 262.54 N/mm2), which was
significantly lower than the yield stress of the steel, when the vertical
load on the column reached 211.68 kN. Moreover, the stresses at points
G and R were similar to those at points E and N.
Therefore, it was concluded that the cold-formed thin-wall bottom
frame beam had low rigidity, which caused evident deflection under
the floor load. The vertical load on the column reduced the deformation
of the bottom frame beam, indicating the beneficial constraint of the
column to the bottom frame beam.
In addition to the analysis of the properties of the beams, the prop-
erties of the columns are considered in studying the properties of the
ATLS modular unit. Fig. 12 illustrates the load–displacement curves of
the middle height of columns C1–C4 under a vertical load on the
Fig. 15. Test setup. column. The curves exclude the influence of the deformation at both
8 J.-F. Zhang et al. / Journal of Constructional Steel Research 168 (2020) 105981

Fig. 16. Deformation between columns and corner fittings.

12 12
10 10
8 8
Load (kN)

Load (kN)
6 6
4 4
2 C2-1 C2-2
2
C3-1 C3-2
0 0
0 10 20 30 0 5 10 15
Displacement (mm) Displacement (mm)
(a) Load-displacement curves at points (b) Load-displacement curves at points
C2-1 and C3-1 C2-2 and C3-2

Fig. 17. Load-displacement curves of columns C2 and C3 under the horizontal load.

12 12
10 10
8 8
Load (kN)

Load (kN)

6 6

4 D1 4
F1
D2 2 F2
2 D3 F3
0 0
0 50(N/mm2) 100 0 50 100 150
Stress Stress (N/mm2)
(a) Load-stress curves at point D of (b) Load-stress curves at point F of
column C1 column C2

Fig. 18. Load-stress curves at key points of columns C1 and C2 under the horizontal load.

ends of the columns. Fig. 12(a) and (b) indicate that the displacement of (b), the stress state at bottom end point of columns C1 and C2 was sim-
the four columns suddenly increased when the vertical load on the col- ilar to that in Fig. 13(a). Additionally, the stress states at the middle
umn reached approximately 101.30 kN, and the load continued to in- height points and bottom end points of columns C3 and C4 were similar
crease slightly with an increase in the displacement of the columns. As to the stress states of columns C1 and C2, respectively. The vertical bear-
illustrated in Fig. 12(a), the two perpendicular flanges of column C1 ing capacity test was stopped owing to the excessive deformation of the
gradually produced opposite deformation under the vertical load on columns. The stress of all four columns was lower than the yield stress
the column. Columns C2–C4 exhibited almost the same deformation of the steel, according to the analysis results illustrated in Fig. 13.
mode as that of C1. These findings indicate that the asymmetric section
columns easily produced flexural-torsional buckling under the vertical 4. Horizontal stiffness test
load on the column and exhibited the typical primary buckling mode.
Fig. 13(a) illustrates the load–stress curves at the middle height 4.1. Loading program and instrumentation
points of columns C1 and C2 under the vertical load on the column.
The maximum stress occurred at point S3-C1-A (approximately The horizontal loading program was similar to the vertical bearing
266.77 N/mm2) when the vertical load on the column reached 211.68 capacity test. Step loading and continuous loading were employed dur-
kN. In contrast to column C1, the maximum stress of column C2 was ap- ing the horizontal stiffness test. Every step of the horizontal load on the
proximately 59.88 N/mm2 (occurring at point S3-C2-H), which was sig- top end of the column, which was less than 20% of the yield bearing ca-
nificantly lower than the yield stress of the steel. As illustrated in Fig. 13 pacity of the column, was imposed and sustained for 10 min to inspect
J.-F. Zhang et al. / Journal of Constructional Steel Research 168 (2020) 105981 9

12 12
10 10
8 8

Load (kN)
Load (kN)
6 6
4 4 G1
C1
2 C2 2 G2
C3 G3
0 0
0 20 40 60 0 50 100 150
Stress (N/mm2) Stress (N/mm2)
(a) Load-stress curves at point C (b) Load-stress curves at point G

Fig. 19. Load-stress curves at key points of beam B1 under the horizontal load.

the components for signs of breakage, yielding, and buckling. A vertical changed from the linear elastic phase to the nonlinear elastic phase with
load on the roof and a floor load were applied to simulate the most ad- an increase in the horizontal load. The deformation of C1 and C4 was
verse load conditions before horizontal loading, in accordance with GB similar to that of C2 and C3. The horizontal stiffness of the ATLS modular
50009-2012 [26]. Differently, the vertical load was applied at one time unit during the linear elastic phase is defined as follows:
(approximately 42.6 kN, which is equivalent to the weight of two
ATLS modular units). The horizontal stiffness test was stopped owing 2p
K¼ ð1Þ
to the excessive deformation of the ATLS modular unit. Besides, transla- Δ
tional degrees of freedom at the bottom of the corner fitting were
p represents the horizontal load and Δ represents the average displace-
restrained.
ment at points C2-1 and C3-1.
Similar to the vertical bearing capacity test, displacement transduc-
According to Fig. 17(a), the horizontal stiffness of the ATLS modular
ers and strain rosettes were adopted to monitor the mechanical proper-
unit was calculated as K = 1.21 kN/mm using formula (1). The horizon-
ties of the ATLS modular unit. Fig. 14 illustrates the measurement
tal displacement of the structure should not exceed H/300 (about
system for the horizontal stiffness test. The arrangement of the displace-
9.33mm) under wind load, in accordance with the Technical Specifica-
ment transducers differed only slightly from that in the vertical bearing
tion for Low-Rise Cold-Formed Thin-Walled Steel Buildings (JGJ 227-
capacity test, as illustrated in Fig. 14(a).
2011) [27], where H represents the height of the building. When the
horizontal displacement of the ATLS modular unit reached 9.33 mm,
4.2. Test phenomenon and observation the horizontal load was 5.83 kN and the deformation of the ATLS mod-
ular unit was still in the elastic stage.
The deformation of the ATLS modular unit was almost identical to Fig. 18(a) illustrates the load–stress curve at point D of column C1
that in the vertical bearing capacity test, at the initial phase. As illus- under the horizontal load. When the load reached 9.62 kN, the stress
trated in Fig. 16(a), a gap appeared between the column and the corner at S2-C1-D was approximately 85.40 N/mm2, and it was larger than
fitting under the floor load, but it disappeared with an increase in the other regions on the column. Besides, as the load reached 9.17 kN, the
vertical load on the top of the ATLS modular unit. Subsequently, a hori- stress at S1-C2-F was approximately 142.31 N/mm2 as illustrated in
zontal load was applied to the top of the columns through two hydraulic Fig. 18(b). The stresses of C4 and C3 were similar to those of C1 and
jacks synchronously, as illustrated in Fig. 15. Fig. 16(b) illustrates the ex- C2, respectively. Analysing the load–stress curves of the columns re-
cursion between the column and the corner fitting under the vealed that the longer flange suffered more stress than the shorter
horizontal load. flange. This suggests that the longer flange of the column was more
prone to instability failure.
4.3. Test results and discussion In addition to analysis of the stress state of the column, Fig. 19 illus-
trates the load–stress curves at points C and G of the bottom frame
Fig. 17(a) illustrates the load–displacement curves at points C2-1 beam under the horizontal load. As illustrated in Fig. 19(a), the stress
and C3-1 under the horizontal load. Similarly, Fig. 17(b) illustrates the at point C decreased at the initial phase, and then increased with an in-
load–displacement curves at points C2-2 and C3-2 under the horizontal crease in the horizontal load. When the load reached 9.17 kN, the stress
load. According to the results, the deformation of the ATLS modular unit at S1-C-B1 was 53.36 N/mm2, where “S” represents the strain rosette,

Fig. 20. FE model.


10 J.-F. Zhang et al. / Journal of Constructional Steel Research 168 (2020) 105981

Fig. 21. Comparison of the results under the floor load between the test and the FEA.

the first “1” represents the serial number of the strain rosette, “C” repre- connecting two points and springs connecting points to ground. The
sents the number of key points, and “B1” represents beam 1. In contrast bolts between the column and the corner fitting were replaced by
to the stress at point C, the stress at point G always increased with an springs-connect two points with a stiffness of 1.5 × 105 kN/mm accord-
increase in the horizontal load. As illustrated in Fig. 19(b), when the ing the Eq. (2).
load reached 9.17 kN, the stress at S1-G-B1 was approximately 119.23
N/mm2. The stresses at points C and G were both lower than the yield As
stress of the steel. However, the horizontal stiffness test had to be kb ¼ 1:6 ð2Þ
Lb
stopped owing to the excessive horizontal deformation of the ATLS
modular unit.
where Lb is the bolt elongation length, taken as the grip length (total
thickness of material and washers), plus half the sum of the height of
5. Finite-element (FE) simulations the bolt head and the height of the nut. As is cross-sectional area of
the screw and Kb is the spring stiffness of the bolt. [30,31]. Addition-
5.1. General ally, face-to-face contact between the end plate of the column and
the end plate of the corner fitting was established [32]. The mesh
FE simulations were conducted using the commercial software pack- sizes for the corner fitting, beam, and column were 20, 50, and 50
age ABAQUS (Version 6.14) [28]. Two identical FE full-scale models mm, respectively.
were established to simulate the vertical bearing capacity test and the
horizontal stiffness test. The measured geometries of the components 5.3. Simulation of vertical bearing capacity test
are presented in Tables 1 and 2. The material properties in the FEA
were the same as those in the test. The accuracy of the FE model The loading program and the boundary condition were the
(FEM) was verified against the test results. same as the program used in the vertical bearing capacity test.
Figs. 21 and 22 illustrate a comparison of the results between
5.2. FEM the vertical bearing capacity test and the FEA. As illustrated in
Fig. 21(a), the gap between the column and the corner fitting in
Fig. 20 illustrates a representative FEM developed in this study. The the FEA was almost equal to the gap in the test. Fig. 22
main components in the two tests were modelled, including the steel (b) illustrates a comparison of the deflection of the bottom
column, steel beam, and corner fitting. The shell element with a reduced frame beam under the floor load. The two curves were highly con-
integration point (S4R) in the ABAQUS element library was employed to sistent. When the floor load reached 3.19 kN/mm2, the beam de-
simulate the steel column, steel beam, and corner fitting [29]. The bilin- flection of the bottom frame was approximately 23.75 and
ear elastoplastic model of the steel and Mises yield criterion were 23.4 mm in the test and FEA, respectively. Analysis of Fig. 21 indi-
employed in the FEM. A new beam (beam system) was formed by com- cates that the spring element accurately simulated the bolt con-
bining the stringers with the beams. Four corner fittings were tied to the straint between the column and the bottom frame beam.
new beam via a surface-to-surface tie constraint to form the top and Fig. 22 illustrates the deformation of the four columns under the ver-
bottom frames. Then, by assembling the top and bottom frames and tical load on the columns, in the FEA. The four columns exhibited almost
the four columns, an ATLS modular unit model was established. the same buckling modes that were observed in the vertical bearing ca-
ABAQUS has a convenient spring element library, including springs pacity test. The same phenomenon as the test was captured. When the

Fig. 22. Deformation of the four columns in the FEA and the test.
J.-F. Zhang et al. / Journal of Constructional Steel Research 168 (2020) 105981 11

12 12
10 10
8 8

Load (kN)

Load (kN)
6 6
C2-1 Test C2-2 Test
4 C3-1 Test 4
C3-2 Test
C2-1 FEA C2-2 FEA
2 C3-1 FEA 2
C3-2 FEA
0 0
0.0 10.0 20.0 30.0 0.0 5.0 10.0 15.0
Displacement (mm) Displacement (mm)
(a) Load- displacement curves at top (b) Load- displacement curves at
end points of columns C2 and C3 middle height points of columns C2 and
C3

Fig. 23. Comparison of the deformation between the Test and FEA.

vertical load was applied to the column, the deflection of the beam 9.33mm), the deformation of the ATLS modular unit was still in
would decrease too. This suggests that the FEA accurately simulated elastic stage, satisfying the requirements of JGJ 227-2011.
the phenomenon observed in the vertical bearing capacity test. (4) The horizontal stiffness of the ATLS modular unit was 1.19
kN/mm in the FEA, which exhibited a small error of approxi-
5.4. Simulation of horizontal stiffness test mately 1.65% compared with the test result. Thus, the results
of the FEA were reliable.
For this simulation, both the loading program and the boundary con-
dition were the same as the program used in the horizontal stiffness Declaration of Competing Interest
test. The load and corresponding displacement of the columns obtained
via the experiment and the FEA are compared in Fig. 23. Fig. 23 None.
(a) illustrates the load–displacement curves at points C2-1 and C3-1,
for both the horizontal stiffness test and the FEA. In the FEA, the hori- Acknowledgements
zontal stiffness of the ATLS modular unit was 1.19 kN/mm according
to formula (1). Compared with the stiffness in the horizontal stiffness The research described in this paper was financially supported by
test, it exhibited a small error of approximately 1.65%. Fig. 23 the Key Research Projects of Henan Higher Education Institutions
(b) illustrates the deformation at points C2-2 and C3-2, indicating a de- (Grant NOs. 16A560010 and 20A560001), the Natural Science Founda-
formation mode similar to that in Fig. 23(a). The comparison indicated tion of China (Grant NO. 51908511) and the China Postdoctoral Science
that in the elastic stage, the deformation of the ATLS modular unit was Foundation (Grant NO. 2019M662532).
almost the same between the test and the FEA, and the horizontal stiff-
ness of the ATLS modular unit in the FEA exhibited close agreement References
with the test results.
[1] E.M. Generalova, V.P. Generalov, A.A. Kuznetsova, Modular buildings in modern con-
struction, Procedia Eng. 153 (2016) 167–172.
6. Conclusions [2] R.M. Lawson, O. Ray, G. Chris, Design in modular construction, CRC Press, Boca
Raton, FL, 2014.
[3] R.M. Lawson, O. Ray, B. Rory, Application of modular construction in high-rise build-
An experimental investigation of the mechanical properties of an ings, J. Archit. Eng. 18 (2) (2012) 148–154.
ATLS modular house was performed. Two identical full-scale modules [4] A. Fathieh, O. Mercould, Seismic evaluation of modular steel buildings, Eng. Struct.
122 (2016) 83–92.
were used in a vertical bearing capacity test and a horizontal stiffness
[5] L. Seungjae, P. Jaeseong, S. Sudeok, K. Changhoon, Seismic performance evaluation
test. The following conclusions are drawn. of the ceiling-bracket-type modular joint with various bracket paratransducers, J.
Constr. Steel Res. 150 (2018) 298–325.
(1) In the vertical bearing capacity test, the ATLS modular unit failed [6] A.W. Lacey, W.S. Chen, H. Hao, K.M. Bi, Structural response of modular buildings –
An overview, J. Build. Eng. 16 (2018) 45–56.
owing to the excessive flexural-torsional buckling of the four col- [7] F.E. Boafo, J.H. Kim, J.T. Kim, Performance of modular prefabricated architecture:
umns. The columns exhibited the typical primary buckling mode. case study-based review and future pathways, Sustainability. 8 (2016) 558.
Additionally, FEA results exhibited good agreement with the re- [8] J. Kim, J. Lee, A basic study on the application of modular construction, J. Korean
Hous Assoc. 25 (2014) 39–46[CrossRef].
sults of the vertical bearing capacity test under floor load, and [9] E.F. Deng, L. Zong, Y. Ding, X.M. Dai, N. Lou, Y. Chen, Monotonic and cyclic response
the spring element accurately simulated the constraint in the of bolted connections with welded cover plate for modular steel construction, Eng.
bolt connection region. Struct. 167 (2018) 407–419.
[10] R. Sanches, O. Mercould, B. Roberts, Experimental investigations of vertical
(2) The bottom frame beam of the ATLS modular house produced post-tensioned connection for modular steel structures, Eng. Struct. 175 (2018)
18.36mm deflection under the standard load combination, 776–789.
which was approximately 1/330 of the span of the bottom [11] X.M. Dai, L. Zong, Y. Ding, Z.X. Li, Experimental study on seismic behavior of a novel
plug-in self-lock joint for modular steel construction, Eng. Struct. 181 (2019)
frame beam. The deformation capacity of the ATLS modular 143–164.
house satisfied with the requirement of JGJ 227-2011. However, [12] H.K. Park, J.H. Ock, Unit modular in-fill construction method for high-rise buildings,
the bottom frame remained in the elastic stage. This indicates the KSCE J. Civ. Eng. 20 (4) (2015) 1201–1210.
[13] C.D. Annan, M.A. Youssef, M.H. El Naggar, Experimental evaluation of the seismic
good deformation capacity and weak rigidity of the bottom
performance of modular steel-braced frames, Eng. Struct. 31 (2009) 1435–1446
frame beam of the ATLS modular house. [CrossRef].
(3) The horizontal stiffness of the ATLS modular unit was 1.21 kN/mm. [14] W. Ferdous, B. Yu, D.T. Ngo, A. Manalo, P. Mendis, New advancements, challenges
According to the observation and analysis, the ATLS modular unit and opportunities of multi-storey modular buildings – A state-of-the-art review,
Eng. Struct. 183 (2019) 883–893.
had a weak horizontal stiffness under a horizontal load. However, [15] K. Giriunas, H. Sezen, R.B. Dupaix, Evaluation, modelling, and analysis of shipping
when the horizontal displacement reached to H/300 (about container building structures, Eng. Struct. 43 (2012) 48–57.
12 J.-F. Zhang et al. / Journal of Constructional Steel Research 168 (2020) 105981

[16] K.S. Park, J. Moon, S.S. Lee, K.W. Bae, W.R. Charles, Embedded steel column-to- [24] X.C. Liu, F.Y. Cui, X.X. Zhan, C. Yu, Z.Q. Jiang, Seismic performance of bolted connec-
foundation connection for a modular structural system, Eng. Struct. 110 (2016) tion of H-beam to HSS-column with web end-plate, J. Constr. Steel Res. 156 (2019)
244–257. 167–181.
[17] E.F. Deng, J.B. Yan, Y. Ding, L. Zong, Z.X. Li, X.M. Dai, Analytical and numerical studies [25] GB/T 228.1-2010, Metallic Materials Tensile Testing at Ambient Temperature, China
on steel columns with novel connections in modular construction, Int. J. Steel Struct. Standard Press, Beijing, China, 2010 [in Chinese].
17 (4) (2017) 1613–1626. [26] GB 50009-2012, Load code for the design of building structures, Ministry of Housing
[18] Z.H. Chen, J.D. Liu, Y.J. Yu, Experimental study on interior connections in modular and Urban-Rural Development of the People’s Republic of China, 2012 , [in Chinese].
steel buildings, Eng. Struct. 147 (2017) 625–638. [27] JGJ 227-2011, Technical specification for low-rise cold-formed thin-walled steel
[19] A.A. Naserabad, M.R. Ghasemi, N. Shabakhty, H.G. Arab, Evaluation of Three Support buildings, Ministry of Housing and Urban-Rural Development of the People’s Re-
Shapes on Behavior of New Bolted Connection BBCC in Modularized Prefabricated public of China. Beijing, China, 2011 , [in Chinese].
Steel Structures, Int. J. Steel Struct. 18 (5) (2018) 1639–1653. [28] Dassault Systémes, Abaqus 6.14 Documentation, Simulia Systems, Providence, RI,
[20] S. Srisangeerthanan, M.J. Hashemi, P. Rajeev, E. Gad, S. Fernando, Numerical study 2014.
on the effects of diaphragm stiffness and strength on the seismic response of [29] Y.J. Yu, Z.H. Chen, Rigidity of corrugated plate sidewalls and its effect on the modular
multi-story modular buildings, Eng. Struct. 163 (2018) 25–37. structural design, Eng. Struct. 175 (2018) 191–200.
[21] Z.H. Chen, J.D. Liu, Y.J. Yu, C.H. Zhou, R.J. Yan, Experimental study of an innovative [30] Eurocode 3 1993-1-8:2005, Design of steel structures—Part 1–8: Design of joints,
modular steel building connection, J. Constr. Steel Res. 139 (2017) 69–82. British Standards Institution, London, 2005.
[22] Y. Ding, E.F. Deng, L. Zong, X.M. Dai, L. Lou, Y. Chen, Cyclic tests on corrugated steel [31] Z.L. Li, K. Soga, F. Wang, P. Wright, K. Tsuno, Behaviour of cast-iron tunnel segmental
plate shear walls with openings in modularized-constructions, J. Constr. Steel Res. joint from the 3D FE analyses and development of a new bolt-spring model, TUNN
138 (2017) 675–691. UNDERGR SP TECH. 41 (2014) 176–192.
[23] X.M. Dai, Y. Ding, L. Zong, E.F. Deng, L. Lou, Y. Chen, Experimental study on seismic [32] X.C. Liu, S.H. Pu, A.L. Zhang, X.X. Zhan, Performance analysis and design of bolted
behavior of steel strip reinforced CSPSWs in modular building structures, J. Constr. connections in modularized prefabricated steel structures, J. Constr. Steel Res. 133
Steel Res. 151 (2018) 228–237. (2017) 360–373.

You might also like