Research and Example Verification on Stability Coefficient of Vertical Bar of Fastener Steel Pipe Support

Abstract

The calculation method of steel pipe support in current specifications is not uniform, and it is unsafe to directly apply the stability coefficient of the vertical bar in the current code, which leads to frequent safety accidents of fastener steel pipe supports. Through the field statistics of the initial defects of the steel pipe, it is found that the initial bending values of the 3 m and 6 m steel pipes with the largest size are larger in L/400. The stability coefficient table is obtained by taking the initial bending of L/400 as the initial defect, which is 43.15% smaller than the stability coefficient given by the specification. Through the comparative analysis of the actual engineering monitoring results, the calculation results of the original specification and the calculation results after the correction of the stability coefficient, it is concluded that the stress calculated by the stability coefficient proposed in this paper is similar to the actual monitoring results of the project and is greater than the stress of the vertical rod calculated by the stability coefficient of the original specification. It is suggested that the corrected stability coefficient should be applied to the actual project.

1. Introduction

Fastener steel pipe support has the advantages of convenient transportation, convenient disassembly and low cost in construction. Because of these advantages, the application frequency of fastener-type steel pipe support is becoming higher and higher, so there are more and more safety accidents [1]. Whether it is a fall in the process of dismantling the frame by the staff or a collapse caused by the overall or local instability of the frame, the reason is due to the lack of understanding of the overall stability of steel pipe support. At present, there is no uniform design and calculation standard for the calculation of the stability of steel pipe support. Therefore, the unified standard calculation method is of great significance to the safe construction of steel pipe supports.

The overall stability coefficient value used in the stability calculation of the fastener-type steel pipe support comes from the stability coefficient table in the “Technical Specification for Cold-Formed Thin-Walled Steel Structures” (GB 50018-2002) [2]. Due to the multiple turnover and transportation of the support frame, the initial bending of a large number of support frames is higher than the value of the initial bending of the axial compression member in the “thin gauge” (initial bending value L/750), and initial bending is an important factor affecting the stability of the vertical rod. Therefore, it is unsafe to directly apply the “thin gauge” value table to the design of the vertical rod of the fastener-type steel pipe support. In 2011, Hu Changming et al. [3] gave the stability coefficient of the vertical bar under the initial bending at L/250 on the basis of considering the cross-section characteristics and initial geometric defects of the steel pipe. The determination of the initial bending is obtained by comparing the test results with the standard calculation results. The initial bending of the steel pipe is not measured, so the determination of the initial bending is open to question. Chu Qi et al. [4] found that the weakness of the horizontal bar is the bearing capacity of the weld at the connection between the socket and the horizontal bar. Preventing the buckling failure of the weld at the horizontal bar and socket is crucial to ensure the bearing capacity of the horizontal bar. Under the action of shear force, weld buckling failure occurs at the connection between the wheel and the vertical bar. As the number of connecting horizontal rods on the wheel decreases, the shear capacity of the wheel decreases significantly.

2. Methodology

2.1. Field Experimental Statistical Analysis of Initial Bending of Steel Pipe

According to the understanding of the initial bending of steel pipes used in practical engineering, 100 steel pipes of 3 m and 6 m were randomly selected from a construction site in Xi’an for field measurement. Through processing the data, it was found that the number of steel pipes in the range of L/350~L/300 was 31, the number of steel pipes in the range of L/400~L/350 was 59, the number of steel pipes in the range of L/450~L/400 was 57, and the number of steel pipes in the range of L/500~L/450 was 31. The maximum initial bending of steel pipes of 3 m and 6 m was L/290 and L/280, respectively. The minimum initial bending was L/560 and L/428, respectively. Figure 1 shows that the initial bending of most steel pipes is in the range of L/500~L/300, and the steel pipes with initial bending of L/400 account for a large proportion.

2.2. Finite Element Simulation Analysis of the Steel Pipe Support Model

In the formwork support system, SAP2000 finite element software can automatically generate the spatial structure system of the support frame when simulating the spatial structure system of the support frame. It is convenient and quick to establish the model of the frame structure and the bar system, and it is more favorable for the treatment of semi-rigid joints. Therefore, in this paper, the steel pipe support model is established by SAP2000 finite element software. On the basis of buckling analysis, the stable bearing capacity of the frame under different initial defects is obtained by nonlinear analysis. By comparing the stable bearing capacity obtained by the test in the literature [3,4,5,6], the validity of the finite element model is verified. Compared with the critical bearing capacity calculated by the stability coefficient in the “thin gauge” (GB 50018-2002), the initial bending of the steel pipe closest to the actual construction conditions is determined. Then, the stability coefficient which is more in line with the actual engineering is calculated, which provides a reference for the calculation of formwork support under actual construction conditions.

2.2.1. Simulation Assumptions for the Steel Pipe Support Model Structure Treatment

The initial geometric defects of the structure, the P-Δ effect, the P-δ effect, semi-rigid joints and other treatments are considered as follows:

• The initial geometric imperfections of the structure are imposed by adding the initial geometric stiffness.
• P-Δ: set by defining the load conditions, using nonlinear, P-Delta follow-up analysis (including modal analysis) in order to maintain the stiffness of the structure, recorded as P-Delta, and linear analysis.
• P-δ: The deflection of the bar is defined under P-Delta and determined by the subdivision of the bar.
• Joint semi-rigid: achieved by releasing the end stiffness at both ends of the horizontal bar [7,8]. We determined that the rotational stiffness of the fastener has a great relationship with the tightening degree of the fastener bolt. The greater the tightening torque, the greater the rotational stiffness of the fastener. It can be obtained from the formula $M=-126.4952{\theta }^2+20.6169\theta +0.06$ and 20 kN·m/rad can be taken as the rotational stiffness of the fastener when the tightening torque is 40 N·m. The tightening torque is 40 N·m, and the stiffness of the end of the horizontal bar is 20 kN·m/rad to simulate the semi-rigid problem of the frame.
• Other: Under all working conditions of this model, the vertical bar and the bottom constraint are regarded as hinged, the connection between the scissors and the vertical bar is regarded as hinged, and the connection between the scissors is lap joint.

2.2.2. Build Finite Element Model

According to the literature survey and field investigation, combined with the statistics of the section size of the steel pipe in [9], a steel pipe support model with a size of 5.8 m × 3.6 m × 5.994 m was established. The steel pipe specification is Φ 48 × 3.2 Q235 steel pipe, and its section characteristics are shown in

Table 1. Section characteristics of steel pipe of support frame.

Category	Cold-Formed Thin-Walled Steel Pipe
Specification (mm)	Φ 48 × 3.2
Weight (N/m)	35.5
Cross sectional area (×102) mm2	4.50
Moment of inertia (×104) mm4	11.36
Modulus of section (×103) mm3	4.73
Radius of gyration (mm)	15.89

.

The semi-rigidity of the steel tube connection joint is set to 20 kN·m/rad in three rotation directions in the form of a point spring. The layout of the scissors brace is vertical 5 steps and horizontal 3 steps. The vertical concentrated load of 20 kN is specified on each node on the top of the frame to replace the actual stress of the frame.

The fastener-type steel pipe support models with initial bending of L/300, L/350, L/400, L/450 and L/500 (L is the height of the support frame) are established, respectively, and the support constraints are set up. The basic working condition analysis model is as shown in Figure 2, and the first-order buckling instability mode of the basic working condition is as shown in Figure 3.

2.2.3. Verification of Finite Element Model

To validate the rationality of the finite element model, the load-displacement curves of the four corners of the frame are obtained by using the finite element software and compared with the load-displacement curves of the corresponding positions obtained from the test in [3] (as shown in Figure 4, Figure 5, Figure 6 and Figure 7).

By comparing the load-displacement curves of the test and the finite element, it can be calculated that the ratio of the finite element to the test results is between 1.04 and 1.12; that is, the finite element calculation results are generally larger than the test results. The main reason for the error is that the initial defect addition in the finite element simulation is too small.

2.2.4. Simulation Result Analysis

The load-displacement curves of JD-1 at the top of the corner pole under different initial bending are obtained by simulation (as shown in Figure 8). The stable bearing capacity is obtained by the load-displacement curve, as shown in

Table 2. Ultimate stability bearing capacity of brace under different initial bending (kN).

Initial Crookedness	L/300	L/350	L/400	L/450	L/500
Buckling strength	73.62	79.83	88.57	92.36	96.14
Test values	84.95	84.95	84.95	84.95	84.95
Standard calculated value	113.6	113.6	113.6	113.6	113.6

.

It can be found from Figure 8 that when the initial bending value is the largest, the stable bearing capacity of the steel pipe support is the largest. When the initial bending value is the smallest, the stable bearing capacity decreases by 23.4% when the initial bending increases. This also shows that the stable bearing capacity of the frame is greatly affected by the initial bending of the steel pipe. The larger the initial bending of the steel pipe, the smaller the stable bearing capacity. When the initial bending value is in the range of L/350~L/400, the load-displacement curve is closest to the load-displacement curve obtained under the field test, and the stable bearing capacity is closest to the test value. It can be clearly seen from Table 2 that the results of finite element simulation are very different from the calculation results of the specification and are close to the experimental values, which further indicates that the initial bending of the steel pipe support used in the specification to calculate the stability coefficient is debatable.

3. Results

3.1. Determination of Initial Defect Calculation of Component

Comprehensive field measurement and finite element simulation analysis: In this paper, L/400 is taken as the initial bending of steel pipe support. It can be calculated from Table 1 that the cross-section parameter of the steel pipe is 1.51. Therefore, 1.51 is taken as the cross-section parameter of the universal steel pipe for calculation.

The comprehensive initial defects of the bar are deduced from the relative initial bending of the bar obtained by the section core distance:

$$\rho =\frac{W}{A}$$

(1)

$${\epsilon }_0=\frac{e_0+v_0}{\rho }$$

(2)

According to the stability theory of steel structure:

$${\epsilon }_0=\frac{e_0+v_0}{\rho }=0.05+\frac{{\lambda }_i}{n\rho }$$

(3)

According to the principle of ‘thin gauge’ (GB 50018-2002), the mathematical model of the comprehensive initial defect of the rod is shown in Equation (4):

$$\begin{gathered} \left\{a{\lambda }_n\right\}\left(0\le {\lambda }_n\le 0.5\right) \\ \left\{0.05+b{\lambda }_n\right\}\left(0.5\le {\lambda }_n\le 1.0\right) \end{gathered}$$

(4)

a, b and c are the corresponding coefficients in the mathematical model of comprehensive defects, λ_n is the regularized slenderness ratio of components. When the initial bending of the bar is L/400, through calculation a = 0.39, b = 0.285, c = 0.285, so the comprehensive defect formula of the bar is calculated when the initial bending is L/400:

$$\begin{gathered} \left\{0.39{\lambda }_n\right\}\left(0\le {\lambda }_n^{}\le 0.5\right) \\ \left\{0.05+0.285{\lambda }_n\right\}\left(0.5\le {\lambda }_n\le 1.0\right) \end{gathered}$$

(5)

In the Equations (1)–(5):

W—section modulus in bending (mm⁴);

A—sectional area (mm²);

e⁰—The initial deformation value at the midpoint of the component (mm);

v⁰—initial crookedness.

3.2. Determination and Comparison of Stability Coefficient

Table 3. Stability factor of fastener-type steel pipe support.

$\mathsf{\lambda }$	0	1	2	3	4	5	6	7	8	9
0	1	0.9937	0.9874	0.9813	0.9752	0.9691	0.9632	0.9572	0.9513	0.9444
10	0.9397	0.9339	0.9282	0.9225	0.9168	0.9111	0.9055	0.8998	0.8942	0.8886
20	0.8830	0.8773	0.8682	0.8661	0.8605	0.8548	0.8492	0.8435	0.8379	0.8322
30	0.8265	0.8209	0.8151	0.8094	0.8036	0.7979	0.7921	0.7863	0.7804	0.7749
40	0.7695	0.7639	0.7584	0.7527	0.7471	0.7415	0.7358	0.7303	0.7229	0.7196
50	0.7142	0.7088	0.7033	0.6978	0.6923	0.6868	0.6812	0.6756	0.6700	0.6644
60	0.6587	0.6531	0.6473	0.6416	0.6359	0.6302	0.6244	0.6186	0.6104	0.6071
70	0.6013	0.5955	0.5898	0.5839	0.5781	0.5724	0.5666	0.5600	0.5532	0.5466
80	0.5399	0.5333	0.5268	0.5203	0.5139	0.5075	0.5012	0.4950	0.4888	0.4827
90	0.4767	0.4706	0.4647	0.4588	0.4524	0.4459	0.4396	0.4334	0.4271	0.4211
100	0.4151	0.4092	0.4034	0.3977	0.3920	0.3865	0.3810	0.3756	0.3703	0.3651
110	0.3600	0.3549	0.3500	0.3451	0.3403	0.3356	0.3310	0.3264	0.3220	0.3176
120	0.3132	0.3090	0.3048	0.3007	0.2967	0.2927	0.2888	0.2850	0.2812	0.2776
130	0.2739	0.2704	0.2668	0.2634	0.2600	0.2567	0.2534	0.2502	0.2471	0.2439
140	0.2409	0.2379	0.2349	0.2320	0.2292	0.2264	0.2236	0.2209	0.2183	0.2156
150	0.2131	0.2106	0.2080	0.2056	0.2032	0.2008	0.1985	0.1962	0.1939	0.1917
160	0.1900	0.1874	0.1853	0.1832	0.1812	0.1791	0.1772	0.1752	0.1733	0.1714
170	0.1695	0.1677	0.1659	0.1641	0.1624	0.1607	0.1590	0.1573	0.1557	0.1540
180	0.1525	0.1509	0.1493	0.1478	0.1463	0.1448	0.1434	0.1419	0.1405	0.1391
190	0.1377	0.1364	0.1350	0.1337	0.1324	0.1311	0.1299	0.1286	0.1274	0.1262
200	0.1250	0.1238	0.1226	0.1215	0.1204	0.1193	0.1182	0.1171	0.1160	0.1149
210	0.1132	0.1129	0.1118	0.1109	0.1099	0.1089	0.1079	0.1070	0.1060	0.1051
220	0.1042	0.1033	0.1024	0.1015	0.1007	0.0998	0.0989	0.0981	0.0973	0.0965
230	0.0957	0.0949	0.0936	0.0928	0.0920	0.0913	0.0906	0.0898	0.0891	0.0884
240	0.0877	0.0870	0.0863	0.0856	0.0849	0.0847	0.0840	0.0834	0.0827	0.0820
250	0.0814	-	-	-	-	-	-	-	-	-

is the recommended stability coefficient calculated in this chapter. Figure 9 is the comparison between the recommended stability coefficient curve calculated in this chapter and the curve given in the ‘thin gauge’ (GB50018-2002) and the class c curve of the European steel structure design code EN1993 [10] (the European steel structure design code classifies cold-formed steel tubes as class c).

It can be seen from Figure 9 that the proposed stability coefficient curve calculated in this chapter is relatively different from the stability coefficient curve in the “thin gauge” (GB50018-2002) of steel pipe support in China and is closer to the stability coefficient curve given in the European standard EN1993. The closest range of slenderness ratio section is 50~250, which is also a commonly used section of steel pipe support in engineering. In this section, the stability coefficient curve of “thin gauge” (GB50018-2002) is significantly higher than that of Europe and the proposed curve given in this paper. Comparing the stability coefficient, it is found that within a certain range, with the increase of slenderness ratio, the stability coefficient varies greatly, and the influence of defects is also greater. The maximum value can reach 94%. This paper lists the relative difference between the recommended stability coefficient and the ‘thin gauge’ (GB50018-2002), as shown in

Table 4. The relative difference between the proposed stability factor in this paper and the code for design of thin steel structures (GB 50018-2002).

$\mathsf{\lambda }$	3~27	28~72	73~93	94~153	153~250
elative difference	1%~10%	10%~30%	30%~40%	40%~94%	40%~50%

.

3.3. Verification of Engineering Example

3.3.1. Project Profile

The project is a teacher training base and a double-creation building in a university in Xi’an. An outdoor construction project, its main purpose is as an office building. It contains garage, teaching and other functions. The total planning construction area of the project is 17,300 square meters. The fastener-type steel pipe support is located in the report hall of the double-creation building. The height of the double-creation building is one floor underground and six floors above ground. The report hall is a frame structure; the project adopts the fastener-type steel pipe full hall support as the temporary support system, and the steel pipe size is Φ 48 × 2.8 mm.

3.3.2. Design Scheme of Fastener-Type Steel Pipe Support

• The height of fastener-type steel pipe support is 6.4 m and the span is 18 m. The design value of steel pipe compressive strength is 205.0 N/mm².
• Support frame structure size: vertical rod step h = 1.50 m, vertical rod longitudinal distance (span direction) L = 1.2 m, horizontal distance 1.2 m, horizontal rod spacing 0.6 m, top extension height a = 0.2 m and a = 0.5 m.
• The self-weight of the formwork is 0.20 kN/m², the self-weight of the steel pipe of the support frame is 1.35 × 0.751 = 1.014 kN, the self-weight of the steel pipe of the support frame is 1.35 × 0.201 = 0.271 kN, the reduction coefficient of the fastener is 1.00 and the standard value of the toppling concrete load is 2002.

3.3.3. Stability Checking of Fastener Steel Pipe Support

It is known that the rotation radius of the vertical rod section is i = 1.60 cm, the net cross-section area is A = 3.974 cm², the net cross-section resistance moment is W = 4.25 cm³, and the compressive strength of the steel pipe is designed. The maximum support reaction force of the horizontal bar is N₁ = 6.072 kN (including the combination coefficient); the top vertical rod section N = 6.072 + 0.271 = 6.343 kN and the non-top vertical rod section N = 6.072 + 1.014 = 7.086 kN. The stability coefficient in the specification [11] and the stability coefficient given in this paper are used to calculate the fastener-type steel pipe support of the project. Taking the wind load as an example, the calculation results are as follows:

When considering the wind load, according to the standard [12] calculation of the maximum axial pressure of the pole N:

Top upright:

N = 6.072 + 1.350 × 0.201 + 0.9 × 0.980 × 0.016/0.600 = 6.366 kN

(6)

Non-top upright:

N = 6.072 + 1.350 × 0.751 + 0.9 × 0.980 × 0.016/0.600 = 7.109 kN

(7)

The bending moment of the vertical bar section generated by the wind load design value is calculated according to the formula in the specification [13,14]:

$${\mathrm{M}}_{\mathrm{W}}=0.9\times 1.4{\omega }_{\mathrm{k}}{\mathrm{l}}_{\mathrm{a}}\frac{{\mathrm{h}}^2}{10}$$

(8)

$${\mathrm{w}}_{\mathrm{k}}={\mathrm{u}}_{\mathrm{z}}\times {\mathrm{u}}_{\mathrm{s}}\times w_0=0.250\times 1.000\times 0.185=0.046\mathrm{kN}/{\mathrm{m}}^2$$

(9)

In the formula:

W_k—characteristic value of wind load (kN/m²);

H—Pole step distance, h = 1.5 m;

l_a—The spacing of the windward side of the pole, l_a = 1.2 m.

Therefore, the bending moment generated by wind load is

$${\mathrm{M}}_{\mathrm{W}}=0.9\times 1.4{\mathrm{w}}_{\mathrm{k}}{\mathrm{l}}_{\mathrm{a}}{\mathrm{h}}^2/10=0.9\times 1.4\times 0.046\times 1.200\times 1.500\times 1.500/10=0.016\mathrm{kN}/\mathrm{m}$$

(10)

For the top vertical bar section: when the vertical rod extension length a = 0.2 m, check the specification Schedule C-2 to get μ₁ = 1.649, the calculated length l₀ = 3.619 m, and the slenderness ratio is allowed (when k takes 1):

λ₀ = 225.888/1.155 = 195.574 < 210. The slenderness ratio calculation meets the requirements. Φ = 0.144 (standard value) and φ’ = 0.0998 (recommended value).

Calculated according to the specification φ value:

$$\sigma =\frac{\mathrm{N}}{\mathrm{A}}+\frac{{\mathrm{M}}_{\mathrm{W}}}{\mathrm{W}}=\frac{6366}{0.144\times 397.4}+\frac{16000}{4248}=115.040\mathrm{N}/{\mathrm{mm}}^2$$

(11)

Calculated as recommended value φ’:

$$\sigma =\frac{\mathrm{N}}{\mathrm{A}}+\frac{{\mathrm{M}}_{\mathrm{W}}}{\mathrm{W}}=\frac{6366}{0.998\times 397.4}+\frac{16000}{4248}=164.279\mathrm{N}/{\mathrm{mm}}^2$$

(12)

When σ = 115.040 N/mm², σ’ = 164.279 N/mm², the stability check of the pole is σ, σ’ = [f], which meets the requirements.

Stability comparison:

$$\frac{164.279-115.040}{115.040}\times 100\%=42.8\%$$

(13)

For the non-top vertical bar section: check the specification Schedule C-4 to get the effective length coefficient μ₂ = 1.951, the effective length l₀ = 3.380 m, and then the slenderness ratio:

$$\lambda =\frac{l_0}{i}=\frac{3380}{16.0}=210.993$$

(14)

Permissible slenderness ratio (when k takes 1):

$${\lambda }_0=\frac{210.993}{1.155}=182.678<210$$

(15)

The slenderness ratio calculation meets the requirements. φ = 0.164 (standard value) and φ’ = 0.1132 (recommended value). Calculated according to the value of specification φ:

$$\sigma =\frac{\mathrm{N}}{\mathrm{A}}+\frac{{\mathrm{M}}_{\mathrm{W}}}{\mathrm{W}}=\frac{7109}{0.164\times 397.4}+\frac{16000}{4248}=112.778\mathrm{N}/{\mathrm{mm}}^2$$

(16)

Calculated as recommended value φ’, the stress of the pole: σ = 112.778 N/mm² and σ’ = 161.795 N/mm². The stability check of the pole is σ, σ’ = [f], which meets the requirements. The stability comparison is calculated according to the recommended φ’ values:

$${\sigma }^{\prime }=\frac{\mathrm{N}}{\mathrm{A}}+\frac{{\mathrm{M}}_{\mathrm{W}}}{\mathrm{W}}=\frac{7109}{0.1132\times 397.4}+\frac{16000}{4248}=161.795\mathrm{N}/{\mathrm{mm}}^2$$

(17)

Calculated as recommended value φ’, the stress of the pole: σ = 112.778 N/mm² and σ’ = 161.795 N/mm². The stability check of the pole is σ, σ’ = [f], which meets the requirements. Stability comparison:

$$\frac{161.795-112.778}{112.778}\times 100\%=43.5\%$$

(18)

3.4. Analysis of Effect

According to the calculation results of the stability of the vertical bar, it is concluded that for the steel pipe support frame of the fastener steel pipe support system, under the condition that the slenderness ratio calculation meets the requirements, and for different vertical bar segments, the stress of the vertical bar calculated by the stability coefficient φ given in the specification is less than the stress value calculated by the stability coefficient φ’ given in this paper. The average stability value of the vertical bar calculated according to the stability coefficient of the specification is 43.15% smaller. Obviously, this is due to the fact that the stability coefficient φ of the specification is larger than the actual value. The main reason is that the determination of the stability coefficient of the specification fails to fully consider the circular section characteristics of the steel pipe and the initial bending degree of the bar. The geometric initial defects of the steel pipe support of the temporary support system are equivalent to the geometric initial defects of the permanent structure. The initial bending value of the steel pipe used in the calculation is small, which is quite different from the actual engineering situation.

4. Conclusions

Based on the field measurement of the initial defects of the steel pipe and the finite element simulation frame, the analysis and correction method is used to calculate the stability coefficient of the vertical bar under the actual situation. Through analysis and calculation, the following conclusions are drawn:

• Through the field measurement of the initial defects of steel pipes in a construction site, it is found that the initial bending values of most steel pipes are in the range of L/500~L/300, and the initial bending values of the most used 3 m and 6 m steel pipes are in L/400.
• Therefore, the initial bending value of the bar L/400 that is closest to the actual construction condition is determined, and the stability coefficient table of the vertical bar when the initial bending value is L/400 is given, which provides a reference for the calculation of fastener-type steel pipe support under actual construction conditions.
• The stability coefficient table is obtained by taking the initial bending of L/400 as the initial defect, which is 43.15% smaller than the stability coefficient given by the specification. It is suggested that the corrected stability coefficient should be applied to the actual project.

5. Discussion

In terms of continuing to improve the study of the stability coefficient of vertical bar of fastener steel pipe supports, the author believes that further work needs to be done in the following aspects:

In terms of research background and references, based on the relevant data that the research group has mastered, there is incompleteness. In the follow-up, a large number of literatures are needed to study and analyze the main reasons affecting the instability of fastener-type steel pipe supports.

The field measurement of initial bending of steel pipe is only for 100 Q235 steel pipes of 3 m and 6 m. The measurement data only represent the initial defects of some steel pipes. In the future, a large number of steel pipes of different sizes will be measured to ensure the reliability of the data.

The experimental study of high-formwork systems with couplers is not deep enough. There are few field tests, and they affect the data accuracy. There are many factors of uncertainty. The test data in this paper only rely on a high formwork project in Xi’an, and the data are not perfect.