You are on page 1of 16

Engineering Structures 106 (2016) 80–95

Contents lists available at ScienceDirect

Engineering Structures
journal homepage: www.elsevier.com/locate/engstruct

Experimental and theoretical studies on the stability of steel tube–coupler


scaffolds with different connection joints
Hongbo Liu a, Li Jia a, Suolin Wen b, Qun Liu c, Gang Wang a, Zhihua Chen a,⇑
a
State Key Laboratory of Hydraulic Engineering Simulation and Safety, Tianjin University, Tianjin 300072, China
b
Shanghai Tunnel Engineering Co., Ltd, Shanghai 200000, China
c
China Academy of Building Research, Beijing 100013, China

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

Article history: This paper presents a systematic experimental and analytical study on the stability behavior of steel
Received 19 December 2014 tube–coupler scaffolds (STCSs), wherein upper horizontal tubes transfer construction loads. Four
Revised 6 October 2015 specimens with common geometric sizes were constructed and tested statically in the lab to obtain
Accepted 9 October 2015
the strength and failure modes. Advanced nonlinear finite element analysis, which considers the initial
Available online 11 November 2015
member imperfection and semi-rigid behavior of right-angle couplers, was conducted on the specimen
models by using ANSYS to obtain the stress distribution, strength, and failure modes of the STCS.
Keywords:
Analysis results were compared with the experimental results. The following can be concluded from
Steel tube–coupler scaffold
Right-angle coupler
the experimental and analytical results. (1) The main failure mode of the STCSs is the local instability
Slip behavior of a vertical steel tube if the upper layer joints have sufficient anti-slipping capacity. Otherwise, the main
Experimental research failure mode is the slipping and fracture of right-angle couplers. (2) The strength of STCSs is sensitive to
Nonlinear finite element analysis upper joint type. The load capacity of ST2 with the strengthened joint is 1.30 times that of ST1 with a
common joint, and the load capacity of ST4 with a common joint is 1.16 times that of ST3 with a weak
joint. (3) The numerical simulation method presented in this paper is generally consistent with the
full-scale test results and has a maximum difference of 11.08%.
Ó 2015 Elsevier Ltd. All rights reserved.

1. Introduction theoretically. Pienko [9] determined the capacity of the key joint
in inserting-type scaffolds on the basis of numerical analysis and
Many steel scaffold systems, such as door-type scaffolds by considering nonlinear materials and the interaction between
(Fig. 1(a)), inserting-type scaffolds (Fig. 1(b)), cuplok-type scaffolds particular joint elements. Peng et al. [10] investigated the load
(Fig. 1(c)), and steel tube–coupler scaffolds (STCSs) (Fig. 1(d)), are capacities and failure modes of inserting-type scaffolds on the
commonly used as temporary structures to support workers, basis of experimental tests supplemented by analyses. Zhang
materials, and structural members during construction. The STCS [11–13] conducted research on cuplok-type scaffolds by using
system is the most popular system in China, United Kingdom, and the probability-based design method. Beale [14–16], Ao [17], Yue
several other countries because of its easy fabrication, installation, [18], and Liu [19,20] conducted systematic studies on stability
and dissemination. In the past, these systems usually fail in capacity and design method of STCSs.
construction sites because of inadequate design, poor installation, Two common force transfer paths are usually used for STCSs
and overloading, thus causing not only project delays but also during construction, as shown in Fig. 3: one is the load transfer
serious injuries and casualties to construction workers [1,2]. A to the vertical members directly through the U-heads (Fig. 3(a));
picture of a steel scaffold collapse is shown in Fig. 2. the other is the load transfer to the vertical members by the top
Many studies have been conducted on scaffolds in the last horizontal members (Fig. 3(b)). Many experimental and analytical
15 years because of the injuries and casualties induced by scaffold studies have been conducted for the first type [14–20]. However,
failure in construction sites. Chan et al. [1], Peng et al. [2–6], Yu [7], few studies have been performed for the second type, which is
and Weesner [8] produced many studies on the stability capacity used in the construction of large-span spatial structures such as
and design method of door-type scaffolds experimentally and gymnasiums, train stations, garages, and bridges.
Many disastrous collapses of STCSs have occurred in several
⇑ Corresponding author. Tel.: +86 02227409933.
countries. Thus, the safety of this type of system during construc-
E-mail address: zhchen@tju.edu.cn (Z. Chen).
tion has come into question. Moreover, structural properties using

http://dx.doi.org/10.1016/j.engstruct.2015.10.015
0141-0296/Ó 2015 Elsevier Ltd. All rights reserved.
H. Liu et al. / Engineering Structures 106 (2016) 80–95 81

(a) Door-type Scaffolds (b) Inserting-type scaffold

(c) Cuplok-type Scaffold (d) Steel Tube–Coupler Scaffold


Fig. 1. Four types of scaffolds.

the second force transfer path are relatively complicated because horizontal member (Fig. 4(a)), and the rotational coupler is used
of the semi-rigid nature of the coupler and the lack of rational to connect the X-bracing member to the vertical member (Fig. 4
design guidelines in the national standards for STCSs in China (b)). The mechanical property of a right-angle coupler has a signif-
[21], the European Union [22,23], the United States [24], and Japan icant effect on the load bearing of STCS. Therefore, the rotational
[25]. Consequently, understanding the structural behavior of stiffness and slipping stiffness of the right-angle coupler were
STCSs through experimental studies and proposing rational design studied through experiments.
recommendations is important. The tightening torque of the right-angle coupler has a signifi-
cant effect on its mechanical property, and the ‘‘technical code
for safety of steel tubular scaffold with couplers in construction
2. Rotational stiffness and slipping stiffness of right-angle
(JGJ 130-2011)” requires such torques to be approximately
coupler
40 N m. Therefore, the rotational stiffness and slipping stiffness
of the right-angle coupler were studied only with a 40 N m tighten-
The STCS mainly includes three types of members: vertical
ing torque.
members, horizontal members, and X-bracing members. The
Abdel-Jaber have carried out a systematical research on the
right-angle coupler is used to connect the vertical member to the
rotational strength and stiffness of connectors of STCS, and the
slipping stiffness was not included [26]. Considering the material
differences, the rotational stiffness and slipping stiffness of the
right-angle coupler were studied in this paper.

2.1. Test setup and loading procedure

Two specimens with a tightening torque of 40 N m were tested


to obtain the rotational stiffness of right-angle couplers. The test
setup is shown in Fig. 5. To eliminate the influence of elastic
deformation of vertical member and horizontal member, the dial
indicators were set 200 mm away from the right-coupler, as
shown in Fig. 5. The moment was loaded from 0 to 900 N m with
an increment of 100 N m. Weights were used as load. In Fig. 5,
‘‘1” represents the horizontal members, ‘‘2” represents the vertical
members, and ‘‘3” represents the right-angle coupler. The unit is
millimeter.
Two joint types were usually used in construction as shown in
Fig. 2. Collapse of steel scaffold during construction in Anhui Province, November Fig. 6. The first type uses two right-angle couplers (Fig. 6(a)), and
3, 2011. the second type uses three right-angle couplers (Fig. 6(b)). Three
82 H. Liu et al. / Engineering Structures 106 (2016) 80–95

(a) force transferred by U-head (b) force transferred by top horizontal tube
Fig. 3. Two typical types of steel tube and coupler scaffold.

(a) Right-angle coupler (b) Rotational coupler


Fig. 4. Two main joint in steel tube–coupler scaffold.

(a) sketch map (b) practical view


Fig. 5. Test setup for rotational stiffness tests.

specimens with tightening torque of 40 N m were tested to obtain 2.2. Test result analysis
the slipping stiffness of the first joint type. Two specimens with
tightening torque of 40 N m were tested to obtain the slipping stiff- Fracture failure does not occur in the test process. The moment
ness of the second joint type. The test setup for slipping stiffness and rotation angle curves obtained from the rotational stiffness
tests includes one jack, one pressure sensor, steel-load distribution tests are shown in Fig. 7. The quadratic polynomial was used
beam, and loading frame (Fig. 6(c)). to fit the test data, and the expressions were as follows. Two
H. Liu et al. / Engineering Structures 106 (2016) 80–95 83

(a) Two-coupler joint (b) Three-coupler joint (c) Test setup


Fig. 6. Joint type and test setup for slipping stiffness tests.

Fig. 7. Moment and rotation angle curves of the right-angle coupler. Fig. 9. Moment and rotation angle curves of right-angle coupler.

moment–rotation curves were obtained and named with N40-1


and N40-2, and the two test data were fitted with a quadratic
function named N40-1F and N40-2F, as shown in Fig. 7. The mean
rotation for each moment was calculated and shown in Fig. 7 with
curve N40-AVER and N40-AVER. The rotational stiffness used
in subsequent numerical simulation was the initial rotational
stiffness of curve N40-AVER.
On the basis of the fitting expressions of Eqs. (1)–(3), the initial
rotational stiffness was 15.70 and 38.64 kN m/rad for the N40-1
and N40-2. The average value was 22.58 kN m/rad.

M1 ¼ 90 þ 15; 697x  67; 067x2 ; ð1Þ

M2 ¼ 98 þ 38; 637x  476; 480x2 ; ð2Þ

MA ¼ 91:80 þ 22; 575x  150; 243x2 : ð3Þ


The load–coupler displacement curves for the double-coupler
joint and triple-coupler joint are shown in Figs. 8 and 9, respectively.
The maximum vertical loads with coupler displacements below
Fig. 8. Load–coupler displacement curves for the double-coupler joint. 2 mm were 40.29, 39.8, and 37.00 kN for the double-coupler joint.
84 H. Liu et al. / Engineering Structures 106 (2016) 80–95

(a) Local flattening (b) Coupler Fracture A (c) Coupler Fracture B


Fig. 10. Failure modes for slipping stiffness tests.

The maximum vertical loads with coupler displacements below (1) All steel members and connectors were purchased randomly
2 mm were 53.69 and 44.01 kN for the triple-coupler joint. There- from the market.
fore, the average slipping resistance capacity of the double- and (2) All test specimens were built by workers from actual con-
triple-coupler joints were 39.03 and 48.85 kN, respectively. The struction sites.
maximum friction for the double-coupler joint is lower than the (3) The torque applied to the coupler bolts was controlled at
loading capacity of triple-coupler joint. The failure modes for about 40 N m.
the slipping tests are the local flattening of the horizontal steel tube
and coupler fracture, as shown in Fig. 10.
3.2. Material properties

3. Experimental program for the load bearing capacity of STCS The basic stress–strain characteristics of the steel tube were
obtained through tensile coupon tests. Coupons were machined
3.1. Specimen design from the complete section of the steel tube and subsequently
tested in accordance with code GB/T228-2002. The key results
Four full-scale STCSs without U-heads, in which loads are trans- from the coupon tests are summarized in Table 1 and Fig. 13.
ferred from the upper horizontal tubes to the posts, were tested
statically to obtain their strength and failure modes. The ST1 and
ST2 have the same geometric sizes. ST3 and ST4 also have the same
geometric sizes. The geometric details are summarized in Fig. 11. 3.3. Test setup and loading procedure
These specimens have different story heights, different post hori-
zontal spacings, and different connection joints on the upper layer, Four loading frames were used in the tests. Two 50 t hydraulic
as shown in Figs. 11 and 12. The test results can identify the main jacks were fastened to the top of each loading frame, and vertical
factor for the bearing capacity of STCS. The vertical X-bracing was loads were applied to the two top steel distribution beams firstly.
arranged according to the technical code. The loads were then transferred to a group of bottom steel distri-
For the STCS studied in this paper, the connection of the beam bution beams in the perpendicular direction and finally to the scaf-
and post in the upper layer has a large effect on the strength and fold system, as shown in Fig. 14. During the static tests, uniform
failure mode. Three connections were designed as shown in vertical loads were applied to the top horizontal tube of the STCS
Fig. 12. For the Connection A, mutually perpendicular tubes were specimens at a loading rate of 60 N/s until the specimens failed.
connected to the post by right-angle couplers at the same position, For safety, the specimens were not allowed to collapse. Given
as shown in Fig. 12(a). This connection was used in ST1 and ST4. that the pressure in the hydraulic jacks decrease once the STCS
For the Connection B, a right-angle coupler was added below the specimens reached the peak load and started to unload due to
low horizontal tube to strengthen the anti-slip stiffness at each instability, the specimens would be considered ‘‘failed” at this
upper layer connection on the basis of the first type. An additional point and the test would be stopped.
loaded horizontal tube was also connected to the non-loaded hor- In practice, some peripheral members of STCS are not loaded. To
izontal tube, which was used in ST2, as shown in Fig. 12(b). For the make the test results reflect the practical performance, the area
Connection C, a right-angle coupler was added below the lower marked by triangles in Fig. 11 indicate the non-loading area and
horizontal tubes in the same way as that of the second. The loaded the area marked by black dots in Fig. 11 indicate the loading areas.
horizontal tube was connected to another non-loaded horizontal
tube, and used in ST3, as shown in Fig. 12(c).
The steel tubes used in the test specimens had a diameter of 3.4. Instrumentation
48 mm and a thickness of 2.8 mm (old tubes) or 3.2 mm (new
tubes). New steel tubes and couplers were used in the loading Six displacement transducers were mounted on the middle post
area marked by black dots in Fig. 11, and old steel tubes and of the north loading façade and east loading façade at the upper,
couplers were used in the non-loading area marked by triangles middle, and bottom points, as shown in Figs. 11, 15 and 16.
in Fig. 11. To simulate the realistic behavior of STCSs on Twenty-seven strain measurements, including 18 vibrating wire
construction sites, the test specimens were fabricated in the gauges and 9 resistance gauges, were mounted at the upper, mid-
following manner: dle, and bottom points for some post, as shown in Figs. 11 and 16.
H. Liu et al. / Engineering Structures 106 (2016) 80–95 85

(a) Plan view of ST1 and ST2 (b) Plan view of ST3 and ST4

(c) Elevation view of ST1 and ST2

(d) Elevation view of ST3 and ST4


Fig. 11. Layout for the specimens.
86 H. Liu et al. / Engineering Structures 106 (2016) 80–95

(a) Connection A (b) Connection B

(c) Connection C
Fig. 12. Connection type of the upper layer members.

Table 1 For ST2, no observable phenomenon occurred when the load is


Material mechanical properties from tensile coupons tests.
below 33.33 kN. Some upper right-couplers began to slip when
Specimen Young’s Yield Ultimate the load was applied to 36.17 kN. Two right-couplers fractured
modulus Es stress fy strength fu during loading from 39.51 kN to 41.98 kN, and an additional
(MPa) (MPa) (MPa)
right-coupler fractured during load duration. Subsequently, the
New tube S1 2.34  105 357.18 401.83 load was reduced to 39.26. Continue loading when three addi-
S2 2.19  105 355.54 400.94 tional right-angle couplers were installed below the three fracture
S3 2.23  105 367.72 403.61
Average 2.25  105 360.15 402.13
couplers. The hydraulic jacks began unloading when the load
applied was 43.21 kN again, and the test was terminated. The fail-
Old tube S4 2.53  105 418.57 476.24
S5 2.45  105 447.01 511.40
ure modes of ST2 were fracture and slipping of the right-angle
S6 2.42  105 461.20 509.94 coupler, bending, and flatting of upper horizontal tube, as shown
S7 2.43  105 434.43 511.64 in Fig. 19.
S8 2.80  105 409.86 490.97 For the ST3, no observable phenomenon occurred when the load
S9 2.79  105 415.25 480.56
is below 16.67 kN. Some upper right-couplers began to slip, and
Average 2.57  105 431.05 496.79
the upper layer posts began to have observable bending when
the load 20.83 kN was applied. A right-angle coupler fractured
when the load was from 20.83 kN to 25.0 kN. The upper layer posts
4. Experimental results analysis exhibited remarkable bending, and some right-angle couplers frac-
tured when the load reached 25.0 kN. Subsequently, the test was
4.1. Experimental phenomenon analysis and failure modes terminated. The failure modes of ST3 were fracture and slipping
of right-angle coupler, bending, and flatting of the upper horizontal
For the ST1, no observable phenomenon occurred when the load tube, as shown in Fig. 20.
was below 29.63 kN. A right-coupler fractured during loading from For the ST4, no observable phenomenon occurred when the load
29.63 kN to 33.33 kN, as shown in Fig. 17. The load was then was below 16.67 kN. The right-angle fracture and some upper
reduced to 18.89 kN. Continue loading when an additional right-couplers began to slip when the load 20.83 kN was applied.
right-angle coupler was installed below the fracture coupler. The hydraulic jacks began to unload slowly, and the upper layer
Another two right-couplers fractured when 23.58 kN load was posts had observable bending when the load 29.17 kN was applied.
applied. Two additional right-angle couplers were installed below A right-angle coupler fractured during loading from 29.17 kN to
the fracture couplers. Given the significant deformation of upper 33.33 kN. Thereafter, the load was reduced to 17.22 kN. To study
loaded horizontal steel tube and large slip length of 40 mm, the the load capacity change, continue loading when a right-angle
test was terminated when a load of 33.33 kN was applied again. coupler was added. The load was stable at 30.00 kN, and the test
The failure modes of ST1 were the fracture and slipping of the was terminated. The failure modes of ST4 were the fracture and
right-angle coupler, bending, and flatting of the upper horizontal slipping of the right-angle coupler, bending, and flatting of the
tube, as shown in Fig. 18. upper horizontal tube, as shown in Fig. 21.
H. Liu et al. / Engineering Structures 106 (2016) 80–95 87

(a) New tube (b) Old tube


Fig. 13. Steel stress–strain curve for the tube.

Fig. 14. Loading mechanism and test setup.

Fig. 15. Displacement transducers.


Fig. 22 shows that the upper part of the vertical post exhibits
different degrees of bending and that the lower part has no
observable bending. 4.2. Stability strength analysis
The bottom end of the scaffold system was unrestrained in the
horizontal directions. However, no horizontal displacements were The equivalent post load pcr test , which represented the capacity
observed during the tests and in practice. The boundary condition of each specimen, was defined as the ratio of overall load and the
here could also be considered pinned. number of loaded vertical post. The pcr test of all specimens are

(a) Vibrating wire gauge (b) Resistance gauge


Fig. 16. Strain measurement.
88 H. Liu et al. / Engineering Structures 106 (2016) 80–95

is larger than that of ST1 because one group of additional loaded


horizontal steel tubes were installed, as proven by the nodal
displacement–load curves shown in Figs. 23 and 24.
The strength of STCS using Connection C on the upper layer is
lower than that by using Connection A, as proven by test ST3
(pcr test = 25.00 kN) vs. ST4 (pcr test = 29.17 kN). The reasons may be
as follows: (1) the posts in ST3 are approximately axial compres-
sion, but the posts in ST4 exhibited remarkable eccentric compres-
sion; (2) the lateral stiffness of ST3 is larger than that of ST4
because a frame was not formed between the upper loaded hori-
zontal steel tube and vertical steel tube in ST4, as proven by the
nodal displacement–load curves, shown in Figs. 25 and 26.
The strength of STCS increased with decreasing story height, as
proven by test ST1 (pcr test = 33.33 kN) vs. ST4 (pcr test = 29.17 kN).
The member stress–load curves for the upper loaded horizontal
Fig. 17. Additional right-angle couplers installed below the fracture coupler. tubes in Specimen ST1 are shown in Fig. 27. There is a remarkable
difference for the stress. The upper loaded horizontal tube is a
continuous beam theoretically. The member stress distribution
summarized in Table 2, and the load–displacement curves, which for each span should have no difference. The practical significant
represented the stiffness of STCS, are summarized in Figs. 23–26. difference may be induced by excessive slipping displacement or
The following conclusions were obtained according to Table 2 a fracture on individual upper anti-slippery right-angle couplers.
and Figs. 23–26. Therefore, the span of upper loaded horizontal tube should be
The strength of STCS using Connection B on the upper layer is doubled when calculating its bending strength for safety.
remarkably larger than that using Connection A, as proven by test The member stress–load curves for the vertical post in Speci-
ST1 (pcr test = 33.3 kN) vs. test ST2 (pcr test = 43.21 kN). The reasons men ST1 are shown in Fig. 28. The axial stress was 74 Mpa if the
may be as follows: (1) the slipping stiffness of Connection B is post was assumed an axial compression member. An obvious dif-
larger than that of Connection A; (2) the lateral stiffness of ST2 ference between the tested stress and ideal axial stress was also

(a) Horizontal tube bending (b) Coupler fracture (c) Coupler slipping

Fig. 18. Failure modes for ST1.

Post bending Coupler fracture


Fig. 19. Failure modes for ST2.
H. Liu et al. / Engineering Structures 106 (2016) 80–95 89

Post bending Horizontal tube bending

Coupler fracture Coupler slipping


Fig. 20. Failure modes for ST3.

(a) Post tube bending (b) Coupler fracture (c) Coupler slipping
Fig. 21. Failure modes for ST4.

present, which proves that a significant moment exists in the post initial imperfection and instability of members, the steel tube
and that the failure is locally unstable. between two adjacent right-angle couplers was simulated by using
10 BEAM188 elements, as shown in Fig. 29. A convergence study
was conducted to determine the number of beam element between
5. Numerical analysis joints. The bearing capacities of Specimen ST1, when the element
number is 2, 4, 6, 8 and 10, are 51.23, 48.77, 47.53, 47.53, and
5.1. Numerical simulation method 46.91 kN respectively. Therefore, the elements number with a
value of 10 can ensure the results accuracy. A spring-damper
Commercial finite element analysis software ANSYS was used to element (COMBIN14) was used to model the right-angle couplers.
study the behavior and predict the strength of STCS tested speci- Classification modeling and node coupling techniques were used in
mens. Three-dimensional beam element (BEAM188) was used to theoretical analysis, and the main modeling steps of ST1 specimen
model the horizontal beams and vertical posts. To consider the were as follows:
90 H. Liu et al. / Engineering Structures 106 (2016) 80–95

Fig. 22. Deformation of vertical post.


Fig. 25. Displacement–load curve of Specimen ST3.

Fig. 26. Displacement–load curve of Specimen ST4.


Fig. 23. Displacement–load curve of Specimen ST1.

Fig. 24. Displacement–load curve of Specimen ST2. Fig. 27. Member stress–load curve of the loaded upper horizontal tube of ST1.
H. Liu et al. / Engineering Structures 106 (2016) 80–95 91

Two nodes that belong to the Y-direction horizontal steel tube


or vertical steel tube were used in the same position to model
a spring-damper element at the ROTX degree of freedom. A
spring constant of 22.58 kN/rad was assigned to the spring-
damper element (COMBIN14), which was the mean initial
rotational stiffness obtained from previous coupler tests.
Step 7: A uniform load was applied to the top horizontal steel
tube in the X-direction. For the boundary conditions, the bot-
tom of the scaffolds was considered pinned, and the top of
the posts was horizontally constrained to some extent. The con-
strained stiffness with value 5–10 kN/m was determined by
inversion analysis based on the test results. Inversion analysis
is a method to determine a parameter value in numerical model
using test data. The main steps are as follows: (1) carry out test
to obtain some structure characteristic value, such as nodal dis-
placement, member stress, and load; (2) assume a value for the
determined parameter and establish the numerical model; (3)
Fig. 28. Member stress–load curve of the post of ST1. compare the nodal displacement and member stress obtained
from numerical analysis and experiment, and if the difference
is enough small, the assumed value is the parameter value;
(4) repeat the Steps (1)–(3) until the parameter value is
Table 2
Load capacity of the four specimens.
determined.

Specimen no. pcr_T (kN) pcr_E (kN) pcr_NE (kN) pcr_JM (kN) Error (%)
The modeling method of ST4 is as same as ST1 shown in Fig. 29
ST1 33.33 65.51 58.02 31.48 5.55 (a). Some differences were present for ST2 and ST3. For ST2, addi-
ST2 43.21 68.69 54.32 39.51 8.56 tional X-direction horizontal steel tubes around the loading area
ST3 25.00 42.07 27.77 27.77 11.08
ST4 29.17 58.98 38.88 30.00 2.85
were modeled by using the nodes nearest the intersection of the
horizontal tube and vertical tube and to the Y-direction horizontal
steel tubes, as shown in Fig. 29(b). For ST3, only the first group
Step 1: The vertical steel tube nodes and elements were mod- spring-damper element at the ROTY degree of freedom was
eled, and the freedom to rotate in the Z-direction was limited. modeled.
Step 2: The X-direction horizontal steel tube nodes and Eigenvalue buckling analysis was conducted firstly to obtain the
elements were modeled, and the freedom to rotate in the elastic critical buckling loads and the failure modes for nonlinear
X-direction was limited. analyses in later steps. Note that neither the material nonlinearity
Step 3: The Y-direction horizontal steel tube nodes and nor the geometric imperfections were considered in the eigenvalue
elements were modeled, and the freedom to rotate in the buckling analysis because it is purely linear elastic.
Y-direction was limited. Nonlinear analysis was then conducted, which consider the
Step 4: The X-bracing steel tube nodes and elements were geometrical imperfection, geometric nonlinearity and material
modeled, and the freedom to rotate in the self-axis direction nonlinearity. A uniform load was applied incrementally to the
was limited. top horizontal steel tube in the west-east direction until the STCS
Step 5: Three nodes with same coordinates were modeled at the became unstable or the solution could not converge. Material non-
intersection of horizontal tubes and vertical tubes in Steps 1–3, linearity was considered by using a perfect elastic–plastic material
and the translational degrees of freedom of the three nodes at model, and the parameter’ value were determined by coupon tests.
the same position were coupled. According to previous literature, the first buckling mode identified
Step 6: Two nodes that belong to the X-direction horizontal from the eigenvalue buckling analysis with maximum amplitude of
tube or vertical steel tube were used in the same position to 0.05 m was applied to the model as the initial geometric imperfec-
model a spring-damper element at the ROTY degree of freedom. tion [21].

(a) FE model of ST1 and ST4 (b) FE model of ST2


Fig. 29. FE model for the four specimens.
92 H. Liu et al. / Engineering Structures 106 (2016) 80–95

(a) ST1 (b) ST2

(c) ST3 (d) ST4


Fig. 30. First buckling mode of the four specimens.

Fig. 31. Failure mode for ST1 obtained from FE analysis.

Fig. 32. Failure mode for ST2 obtained from FE analysis.


H. Liu et al. / Engineering Structures 106 (2016) 80–95 93

Fig. 33. Failure mode for ST3 obtained from FE analysis.

Fig. 34. Failure mode for ST4 obtained from FE analysis.

Fig. 35. Friction force of the upper right-angle couplers obtained from FE analysis Fig. 36. Friction force of the upper right-angle couplers for ST2 obtained from FE
for ST1. analysis.
94 H. Liu et al. / Engineering Structures 106 (2016) 80–95

strength pcr_NL was generally consistent with the full-scale test


results with a maximum difference of 11.08%. The discrepancies
may be the result of the rotational stiffness of couplers, variance
in the mechanical properties of steel tubes, and initial imperfec-
tions of the steel tubes, among others. The FEM simulation results
for the STCS test specimens were considered generally precise, and
the STCS models could be used for future parametric studies.

6. Summary and conclusions

A comprehensive study was conducted on the stability and


strength of STCSs through four full-scale tests and FEM analyses.
The following conclusions were made:

(1) Full-scale tests and FEM analyses show that the main failure
mode of STCSs is the local instability of the vertical steel
tube if the upper layer joints have sufficient anti-slipping
capacity. Otherwise, the main failure mode of STCSs is the
slipping and fracture of right-angle couplers.
(2) The full-scale tests and FEM analyses show that the strength
Fig. 37. Friction force of the upper right-angle couplers for ST3 obtained from FE of STCSs is sensitive to upper joint type. The load capacity of
analysis. ST2 with the strengthened joint is 1.30 times that of ST1
with a common joint, and the load capacity of ST4 with a
common joint is 1.16 times that of ST3 with a weak joint.
(3) The numerical simulation method presented in this study is
generally consistent with the full-scale test results, with a
maximum difference of 11.08%.

Acknowledgements

This work was funded by National Natural Science Foundation


of China (No. 51321065) and Tianjin Urban and Rural Construction
Commission (No. 2015-4).

References

[1] Chan SL, Zhou ZH, Chen WF, Peng JL, Pan AD. Stability analysis of semi-rigid
steel scaffolding. Eng Struct 1995;17:568–74.
[2] Peng JL, Pan AD, Rosowsky DV, Chen WF, Yen T, Chan SL. High clearance
scaffold systems during construction–1: structural modeling and modes of
failure. Eng Struct 1996;18:247–57.
[3] Peng JL, Pan AD, Rosowsky DV, Chen WF, Yen T, Chan SL. High clearance
scaffold systems during construction–2: structural analysis and development
of design guidelines. Eng Struct 1996;18:258–67.
[4] Peng JL, Pan AD, Chen WF, Yen T, Chan SL. Structural modeling and analysis of
Fig. 38. Friction force of the upper right-angle couplers for ST4 obtained from FE modular falsework systems. ASCE J Struct Eng 1997;123:1245–51.
analysis. [5] Peng JL, Pan AD, Chan SL. Simplified models for analysis and design of modular
falsework. J Construct Steel Res 1998;48:189–209.
[6] Peng JL, Pan AD, Chen WF. Approximate analysis method for modular tubular
falsework. ASCE J Struct Eng 2001;127:256–63.
5.2. Numerical analysis results [7] Yu WK, Chung KF, Chan SL. Structural instability of multi-storey door-type
modular steel scaffolds. Eng Struct 2004;26:867–81.
The critical buckling load pcr_E obtained from the eigenvalue [8] Weesner LB, Jones HL. Experimental and analytical capacity of frame
scaffolding. Eng Struct 2001;23:592–9.
buckling analysis is listed in Table 2. The first buckling mode iden- [9] Pienko M, Blazik-Borowa E. Numerical analysis of load-bearing capacity of
tified from the eigenvalue buckling analysis is shown in Fig. 30. The modular scaffolding nodes. Eng Struct 2013;48:1–9.
connection type on the upper layer has little effect on the buckling [10] Peng JL, Wu CW, Chan SL, et al. Experimental and numerical studies of
practical system scaffolds. J Construct Steel Res 2013;91:64–75.
mode. The main failure mode is local post instability. [11] Zhang H, Rasmussen KJR, Ellingwood Bruce R. Reliability assessment of steel
The failure modes and friction force distribution of the upper scaffold shoring structures for concrete formwork. Eng Struct 2012;36:81–9.
right-angle couplers obtained from nonlinear analysis are shown [12] Zhang H, Rasmussen KJR. System-based design for steel scaffold structures
using advanced analysis. J Construct Steel Res 2013;89:1–8.
in Figs. 31–35. The distribution of coupler friction is very non-
[13] Zhang H, Chandrangsu T, Rasmussen KJR. Probabilistic study of the strength of
uniform as shown in Figs. 35–38. The major failure modes were steel scaffold systems. Struct Saf 2010;32:393–401.
overloading of upper horizontal steel tube and coupler failure for [14] Godley MHR, Beale RG. Sway stiffness of scaffold structures. Struct Eng
1997;75(1):4–12.
ST1, right-angle fracture and slipping for ST2, and overloading for
[15] Beale RG, Godley MHR. Numerical modeling of tube and fitting access scaffold
upper horizontal steel tube and local post instability for ST3 and systems. Adv Steel Construct 2006;2:199–223.
ST4. The failure modes and key node displacement shown in [16] Beale RG. Scaffold research – a review. J Construct Steel Res 2014;98:188–200.
Figs. 23–26 are consistent with the results of the full-scale tests, [17] Ao HF, Li GQ. Investigation of overall load-bearing stability capacity of tube-
and-coupler scaffolds. Chin Q Mech 2004;25:213–8 (In Chinese).
thus proving that the FE analysis is rational. The strength pcr_NL [18] Yue F, Yuan Y, Li GQ, Ye KM, Chen ZM, Wang ZP. Wind load on integral-lift
obtained from the nonlinear analysis is listed in Table 2. The scaffolds for tall building. ASCE J Struct Eng 2005;131:816–24.
H. Liu et al. / Engineering Structures 106 (2016) 80–95 95

[19] Liu HB, Zhao QH, Wang XD, et al. Experimental and analytical studies on the [23] British Standards Institution. BS EN 12812: Falsework—performance
stability of structural steel tube and coupler scaffolds without X-bracing. Eng requirements and general design; 2004.
Struct 2010;32:1003–15. [24] American national standards institution. ANSI/ASSE A10.8-2001: safety
[20] Liu HB, Chen ZH, Wang XD, et al. Theoretical analysis and experimental requirements for scaffolding; 2001.
research on stability behavior of structural steel tube and coupler falsework [25] Japanese industrial standards committee. JIS A 8951-1995: tubular steel
with x-bracing. Adv Steel Construct 2010;6(4):946–62. scaffolds; 1995.
[21] Chinese Standards Institution. GJG 130-2001: technical code for safety of steel [26] Abdel-Jaber MS, Beale RG, Godley MHR, Abdel-Jaber M. Rotational strength
tubular scaffold with couplers in construction; 2011. and stiffness of tube-and-fitting scaffold connectors. Proc ICE Struct Build
[22] British Standards Institution. BS EN 12811: Temporary works equipment; 2009;162:391–403.
2003.

You might also like