A Study of the Factors Influencing the Construction Risk of Steel Truss Bridges Based on the Improved DEMATEL–ISM

Abstract

To enhance the safety management of steel-truss-bridge construction, an evaluation method based on the improved DEMATEL–ISM was proposed to analyze the risk factors involved in such construction. Decision Making Trial and Evaluation Laboratory (DEMATEL) is a method for systematic factor analysis that utilizes graph-theory and -matrix tools, allowing for the assessment of the existence and strength of relationships between elements by analyzing the logical and direct impact relationships among various elements in a system. The distinctive feature of Interpretative Structural Modeling (ISM) is the decomposing of complex systems into several subsystems (elements) and constructing the system into a multi-level hierarchical structural model through algebraic operations. Specifically, triangular fuzzy numbers are introduced initially to improve the direct influence matrix in the DEMATEL method, thereby reducing the subjectivity of expert evaluations. The degree of influence, influenced degree, centrality degree, and causality degree of each influencing factor are determined and ranked based on the above analysis. In response to the characteristics of top-push construction, 20 key factors were selected from four aspects: “human, material, environment, and management”. The top five identified influencing factors are displacement during pushing (X10), safety-management qualification (X18), local buckling (X14), overturning of steel beams (X13), and collision with bridge piers during guide beam installation (X7). Subsequently, corresponding solutions were proposed for different influencing factors. The results of the study offer targeted measures to enhance the safety management of steel truss bridge construction and provide a reference for accident prevention.

1. Introduction

While considerable progress has been made in my country’s bridge construction in the past 25 years, the level of safety management is still in need of improvement amid several reports of safety accidents in production. With the introduction of dual prevention mechanisms involving safety-risk grading and identifying hidden dangers in the “Safety Production Law”, more and more scholars have begun to focus on establishing a more comprehensive security-risk-evaluation management system [1]. Given the substantial investment, high risks, long construction periods, and technical difficulty associated with large-scale complex bridges, strict monitoring of construction is imperative to ensure the safety and quality standards of bridge projects. The use of top-push technology has become prevalent in bridge construction, particularly in large-span steel truss bridges [2]. The main advantage of this technology is its efficiency and cost-effectiveness while minimizing the environmental impact. However, it is crucial to acknowledge the potential safety hazards associated with this technology. Therefore, while studying and improving the push technology of steel truss bridges, it is paramount to prioritize safety considerations.

Xue et al. [3] developed a two-way top-pushing steel-box-beam construction-risk-source model to solve the difficulties and risk control problems in the construction process. The model integrates the triangular fuzzy number method, Bayesian network theory, and fuzzy comprehensive evaluation method to quantify the construction safety risk of two-way top-pushing steel-box-beam construction. Kang et al. [4] used the principle of adaptive control, employed Midas/Civil software to establish a finite element model for the construction process, and proposed a series of security control measures to guide the construction process of steel box beams. This method compares and analyzes the data of the theoretical calculations such as displacement and stress during monitoring and measurement to enhance security control in the construction process. Angelo Cardellicchio et al. [5] conducted research on the automatic identification of defects in existing Reinforced Concrete (RC) bridges. To attain the objectives of automatically identifying and interpreting defects in RC bridges, determining the priority order of risks, and planning additional intervention measures, this study utilizes existing Convolutional Neural Networks (CNN) algorithms in conjunction with Class Activation Map (CAM) methods from available explainable artificial intelligence (XAI) techniques. Zhu et al. [6] combined the response face method with an improved Delphi method to estimate safety risk in bridge construction. However, this method may be influenced by subjective factors. While there has been considerable research output on the safety of bridge construction by local and foreign scholars, the safety evaluation and risk control of steel truss bridge push construction are still relatively underdeveloped. This presents a challenge for the development and evaluation of the construction process for steel truss beams.

In this paper, an enhanced approach for the construction of large-span steel truss bridges has been presented, taking full account of the technical aspects as well as key and difficult points. The approach combines and improves the decision test and evaluation test method (DEMATEL [7]) with the interpretive structural model method (ISM [8]). Compared to traditional single DEMATEL or ISM methods, the improved DEMATEL–ISM assessment method enhances the specific analysis of risk-factor quantification. The combination of both methods allows for a more scientific decision-making process, complementing each other’s shortcomings. DEMATEL is utilized to identify the interaction relationships between factors and aids ISM in visualizing the causal organizational structure among various factors. ISM is employed to break down complex evaluation object systems into multiple simple subsystems. To reduce the subjectivity of expert evaluation in the DEMATEL method, the concept of the triangular fuzzy number (TFN) was introduced into the DEMATEL and defuzzification methods. This integration allows for a comprehensive assessment of both qualitative and quantitative risks associated with steel truss beam launching construction. Our approach effectively combines subjective and objective factors, providing valuable guidance for risk control and countermeasure research in this field.

2. Based on Improving DEMATEL–ISM

System science recognizes a system as a unified entity composed of interconnected elements. The DEMATEL method is a valuable tool for system analysis, and functions by dividing the system into its constituent elements and exploring the relationship between them [9]. Additionally, the DEMATEL, short for Decision Making Trial and Evaluation Laboratory, offers a decision-analysis approach that aids in determining causal and influential relationships within decision-making problems. The basic idea behind this method is to represent the problem factors as a network, where the relationships between factors are depicted using a directed graph [10]. This method employs expert interaction evaluation to identify causal and influencing relationships among problem factors, leading to the creation of a factor-influence matrix. Then, these calculated causal-measurement indicators are transformed into a relationship graph, portraying the interdependence and interactions between problem elements [11]. The DEMATEL method boasts several advantages: ① Analyzing both direct and indirect impacts, and establishing causal relationships. ② Guiding decision makers to understand and visualize the relationships between factors. ③ Ranking factors based on the calculated weights of key factors. The DEMATEL method aids in visualizing the causal relationship structure between factors, generating graphical outputs (or directed graphs) depicting causal relationships among variable factors, and identifying the most influential factors.

The Interpretive Structural Modeling (ISM) expresses the mechanisms by which elements interact by constructing a hierarchical relationship structure among them. This reveals the influence paths among elements, as well as the roles of elements within the entire system and the overall functional structure of the system. This method can effectively break down complex and intricate systems into sets of individual subsystems. The ISM boasts several advantages: ① Better identification of the structure within the system. ② Presenting complex structures in a simplified manner. ③ Providing a theoretical foundation for understanding “why” and “how” within the system. ④ Interpreting embedded objects. ⑤ Transforming less-comprehensible-logical-structure models into clear and visible models. The ISM model determines the mutual relationships between parameters or variables in the research problem, aiding in decomposing complex systems into simplified subsystems.

DEMATEL and ISM are both applicable for analysis and decision making in critical and complex situations. The integrated approach of DEMATEL and ISM strengthens and supports the decision-making process, complementing each other. However, due to the subjectivity of the qualitative evaluation provided by experts, the DEMATEL method incorporates the triangular fuzzy number [12] from fuzzy-set theory to enhance objectivity and reduce the subjectivity of qualitative analysis. And, there is limited research on such improvements. To achieve this, the method employs the ISM approach [13], which includes the establishment of a hierarchical structure of factors, the calculation of driving forces and dependencies among factors, and the construction of a factor control chart and a factor hierarchical structure model. These steps enable further analysis and comprehension of the interaction-relationship chart, facilitating the identification of key risk-influencing factors, main driving factors, and the determination of priority treatment elements. By leveraging the hierarchical structure and relationship between elements, the method allows for the determination of priority consideration and treatment elements, as well as the establishment of priority relationships between elements [14]. Additionally, incorporating triangular fuzzy numbers into the DEMATEL–ISM evaluation method within the realm of bridge-construction safety evaluation is a pioneering attempt in the field of bridge construction.

By combining the revised and improved DEMATEL and ISM methods for the joint analysis of risk factors in the evaluation object, considering the mutual influence and priority of factors [15], comprehensive and in-depth analysis and evaluation of the problem or system can be carried out, which helps to develop specific solutions and strategies to solve the problem. The specific modeling method is as follows:

• Invite seven evaluation experts to compare and rate the importance of each influencing factor while also grading the mutual influence relationship between these factors.
• Categorize the degree of mutual influence into five levels: no influence, little influence, little influence, big influence, and large influence.
• Assign triangular fuzzy values to each of the five-level indicators.

The definition of the triangle vague number [16] is as follows:

There are fuzzy numbers on the real domain R. Its affiliate function μ_{A (x)}: R → [0, 1], x∈ R, if there is

$${\mu }_{A(x)}=\left\{\begin{array}{l}0 , x\le a\\ \frac{x-a}{b-a} , a\le x\le b\\ \frac{c-x}{c-b} , b\le x\le c\\ 0 , x\ge c\end{array}\right.$$

(1)

Let A be a triangular vague number, represented as a = (a, b, c), where a ≤ b ≤ c. Here, a, b, and c denote the left-end, peak, and right-end points of the triangular fuzzy number, respectively.

The basic idea of defuzzification involves treating the vague number as a triangle and then calculating the position of its center of gravity to find a specific real number. As the result of the fuzzy number, the position of the center of gravity can be expressed as

$$x_0=\frac{{\displaystyle {\int }_a^{b}x\mu (x)dx}}{{\displaystyle {\int }_a^b\mu (x)dx}}$$

(2)

In this equation, the weighted average of the fuzzy number in the [A, B] interval is obtained, where the weight of the denominator corresponds to the weight of the vague number within the interval. This value can be regarded as the average value of the fuzzy number, i.e., the defuzzified result.

Using the above methods, the vague number of the triangle can be defuzzified to obtain the directly affecting matrix, M. Then, by perverting the following naturalization of the matrix M, the specification directly affects the matrix N:

$$N=\frac{1}{\max 1\le i\le n{\displaystyle \sum _{i=1}^n{M}_{ij}}}M$$

(3)

To further analyze the impact within the system, the comprehensive impact matrix T is calculated based on the matrix N:

$$T=\frac{1}{N(E-N)}$$

(4)

There are four criterions to assess the effect of elements on the systems: the degree of influence, influence, centrality, and cause serve [17], which can be calculated based on the value of t_ij through the comprehensive influence matrix T, as shown in Equation (4). Here, t_ij represents the direct and indirect impact of factor i on factor j, which collectively form the degree of comprehensive impact. Additionally, the degree of comprehensive impact of factor j is affected by the comprehensive impact of factor i.

The influence degree, denoted as D, refers to the sum of the values of the line matrix of the comprehensive impact matrix T, indicating the comprehensive impact value of the corresponding elements of each row on all other elements.

$$D=(D_1,D_2,D_3,\cdots ,D_n)$$

(5)

$$D_i={\displaystyle \sum _{j=1}^n{t}_{ij},(i=1,2,3,\cdots ,n)}$$

(6)

The degree of influence, represented by C, is the sum of the columns of each column of the comprehensive affecting matrix T, indicating that the corresponding elements of each column are affected by the comprehensive impact value of all other elements.

$$C=(C_1,C_2,C_3,\cdots ,C_n)$$

(7)

$$C_i={\displaystyle \sum _{j=1}^n{t}_{ij},(i=1,2,3,\cdots ,n)}$$

(8)

The centrality degree (M_i) indicates the position of a factor within the evaluation index system and the magnitude of its influence.

$$M_i=D_i+C_i$$

(9)

The reason degree (R_i) represents the reduction in influence and impact exerted by factor I.

$$R_i=D_i-C_i$$

(10)

Then, an ISM analysis was carried out for the above contents. Using the comprehensive influence matrix, an appropriate threshold was selected to establish the adjacent matrix K, which captures the influence of each factor on itself and other factors. The adjacent matrix K and identity matrix I are summed as follows to obtain the reachability matrix H.

$$H=\left(K+I\right)n+1=\left(K+I\right)n\ne \left(K+I\right)2=\left(K+I\right)$$

(11)

The process of drawing a hierarchical model based on the reachability matrix H, which establishes a hierarchical structure of factors and determines the priority and weight between factors, includes establishing a factor hierarchy, calculating the driving forces and dependencies among factors, and establishing a factor control chart and factor hierarchy model.

3. Identification of Risk Factors Influencing the Jacking Construction of Steel Truss Bridge

3.1. Engineering Examples

This article is based on the Xi’an Metro Line 10 Phase I Project, specifically Section 2 Work Area 2. The Weihe Extra Large Bridge within this Section spans a total length of 1412 m and consists of various spans arranged as follows: 124 + 132 + 132 + 168 + 300 + 168 + 132 + 132 + 124 m. The main bridge has a total of 10 piers (W01~W10), and the main beam installation adopts a walking-type pushing method. Additionally, 13 temporary pushing piers (L01~L13) are set between the main bridge piers. The pushing spans mainly range from 60 m to 72 m, with a maximum span reaching 84 m. Construction is conducted alternately on both sides of the river, with 55 sections of 664 m on the south bank (excluding 4 sections of guide beams) and 62 sections of 748 m on the north bank (excluding guide beams). These sections are joined together in the middle of the bridge.

The total overhead weight of the bridge is approximately 54,953 tons. The main truss of the steel beam of the Weihe Bridge has a center distance of 30.5 m, a truss height of 12 m and follows a triangular truss pattern with a standard pitch of 12 m. To accommodate the span arrangement, the side span has been adjusted to a length of 13.5 m. The upper and lower decks of the bridge consist of orthotropic steel plates in a plate truss composite structure. The bridge deck is connected to the top plate of the chord, enabling joint stress between the plates and trusses. The main structure adopts Q370qE, Q420qE, and Q500qE steel plates, with a maximum plate thickness of 64 mm for the main truss member and 70 mm for the node plate. The standard member is 12 m long, about 3.5 m high (including the height of the node plate structure), and is 2.19 m wide. The maximum weight of a member is 75.9 t, while the maximum weight at the junction of the stiffening chord and the top chord is 140.9 t.

Figure 1, Figure 2 and Figure 3 are used to illustrate the specific details of the Weihe River Bridge. Figure 1 shows the schematic diagram of a cross-sectional structure, Figure 2 presents the layout of Weihe River Bridge span, and Figure 3 displays the overall view of the bridge.

3.2. Construction Difficulties

This bridge incorporates a diverse array of steel structural members in substantial quantities. The total weight of the steel truss beams is 62,000 tons, necessitating the assembly of numerous members across a considerable distance, especially for the extensive main span measuring 300 m. The entire bridge is designed without a deck expansion joint, adding complexity to the assembly and installation of steel truss beams. Moreover, the construction precision of the large temporary structures must be maintained. The primary beam-pushing construction involves multiple mechanical equipment, diverse types of work, various construction levels, and a prolonged construction period for simultaneous operations. The construction content involves lifting activities and high-altitude operations, heightening potential hazards and safety risks. Given these circumstances, it is necessary to conduct a comprehensive investigation, identification, and risk assessment of the factors associated with relevant operation items involved in the construction process. Based on the assessment results, corresponding risk control measures can be formulated to ensure the safe and smooth construction of the project.

3.3. Evaluation Index System for Risk Factors of Steel Truss Beam Jacking Construction

The decomposition of push construction activities includes the division of sub-projects and process (unit) operations. According to the push construction process, the construction operations are decomposed into different levels, with the main processes, construction methods, operation procedures, mechanical equipment, and other characteristics being clearly defined. The breakdown table of construction activities is shown in

Table 1. Construction work breakdown table.

Sub Project	Number	Subdivisional Projects and Operational Processes	Main Mechanical Equipment
Temporary steel structure.	1	Installation and dismantling of gantry crane and assembly platform.	Crawler crane, truck crane, flatbed truck.
	2	Installation and dismantling of temporary pier for top pushing.	Crawler crane, truck crane, flatbed truck, truck pump.
	3	Lifting and vertical movement.	Crawler crane, truck crane, flatbed truck.
Steel beam.	4	Lifting and vertical movement.	200 t gantry crane.
Steel beam pushing.	5	Incremental launching construction.	Pushing system.

.

The risk-influencing factors in steel truss bridge launching construction should not be limited to theoretical and construction process levels but should be combined with actual working conditions to ensure the accuracy of research results and the rationality of the indicator system. To identify the risk-influencing factors more comprehensively and clearly, the risk factors are decomposed from several key work nodes such as the steel beam assembly platform, temporary pier, guide beam, pier side bracket, and launching system. Following the four elements of “human-machine environmental management” safety [18], seven experts and scholars were invited to evaluate the factors affecting the safety risks of push construction, including two university professors, two project senior engineers, and three senior engineering management personnel. All invited personnel have extensive experience in bridge engineering and safety, both in practical work and scientific research. Consistent with the characteristics of the push construction in this project, 20 main factors were selected from four aspects: “human, material, environmental, and management”. The risk-influencing factor for the push construction of the Weihe Steel Truss Bridge was constructed, as shown in

Table 2. Risk factors and indicators of steel truss bridge jacking construction.

Dimension	Code	Influence Factor	Introduction to Factors
Human factors	X1	Cognitive factors of construction personnel.	Workers lack correct safety awareness.
	X2	Illegal operation.	Operators operating machinery in violation of regulations and errors.
	X3	Fatigue work.	Long-term high-intensity work.
Material factor	X4	Rigidity and strength of assembly platform.	Does the design strength of the assembly platform meet the standards?
	X5	Settlement and displacement of assembled platforms.	Is there any displacement on the assembly platform during the pushing process?
	X6	Matching degree between guide beam design and steel beam.	Does the guide beam design meet the standards and match with the steel beam?
	X7	Collision with the pier when the guide beam is on the pier.	Is there any collision between the guide beam and the temporary pier during the pushing process when it moves up the pier?
	X8	Rationality of pier side bracket design.	Does the design of the bracket next to the pier match the permanent pier?
	X9	Does the temporary pier design meet the standards.	Is the temporary pier self-load-bearing and compressive-condition-qualified?
	X10	Pushing occurs with offset.	The steel beam deviates during the jacking process.
	X11	Pushing device failure.	Pushing device malfunction.
	X12	Limit device malfunction.	The limit device has malfunctioned.
	X13	Steel beam overturning.	The eccentric force on the main beam during the jacking process causes the steel beam to overturn.
	X14	Local buckling.	Steel beam instability.
	X15	Inadequate safety-management system.	Incomplete management system.
	X16	Incomplete on-site safety-management measures.	Inadequate on-site safety-management and prevention measures.
	X17	Inadequate safety-education disclosure.	Incomplete and detailed technical disclosure to construction personnel.
	X18	Safety-management qualification investigation.	Poor safety qualification.
Environmental factors	X19	Impact of surrounding buildings.	The impact of surrounding buildings on construction.
	X20	Natural environmental impact.	Wind power, hydraulic influence.

.

4. Building a Safety-Evaluation Model for Steel Truss Bridge Jacking Construction

4.1. Determine the Initial Direct Impact Matrix and Comprehensive Impact Matrix

Seven experts were invited to rate the degree of mutual influence between various risk factors in the construction of this project.

Table 3. Evaluation table of triangular fuzzy numbers.

Evaluation Terminology	Triangular Fuzzy Number
No effect.	(0.00, 0.00, 0.25)
The impact is minimal.	(0.00, 0.25, 0.50)
Less impact.	(0.25, 0.50, 0.75)
Significant impact.	(0.50, 0.75, 1.00)
Serious impact.	(0.75, 1.00, 1.00)

presents the evaluation terms and their corresponding triangular fuzzy values. By analyzing the data information in the table, we can construct an impact matrix A, which represents the interrelationships between each influencing factor.

The comprehensive influence matrix T obtained using defuzzification and normalization calculation of triangular fuzzy numbers in Formula (2) is shown in

Table 4. Comprehensive Impact Matrix T.

	X1	X2	X3	X4	X5	X6	X7	X8	X9	X10	X11	X12	X13	X14	X15	X16	X17	X18	X19	X20
X1	0.061	0.146	0.128	0.051	0.063	0.075	0.082	0.054	0.047	0.122	0.076	0.054	0.107	0.087	0.145	0.153	0.149	0.135	0.025	0.037
X2	0.071	0.093	0.090	0.059	0.095	0.085	0.120	0.062	0.052	0.129	0.083	0.087	0.115	0.111	0.149	0.159	0.155	0.143	0.027	0.041
X3	0.091	0.134	0.067	0.051	0.089	0.076	0.097	0.056	0.046	0.137	0.076	0.054	0.107	0.088	0.142	0.151	0.147	0.092	0.025	0.038
X4	0.081	0.118	0.091	0.068	0.139	0.149	0.142	0.124	0.068	0.180	0.101	0.104	0.167	0.162	0.115	0.182	0.149	0.176	0.033	0.079
X5	0.085	0.149	0.106	0.128	0.085	0.153	0.146	0.127	0.069	0.186	0.104	0.107	0.171	0.167	0.122	0.189	0.171	0.172	0.034	0.081
X6	0.085	0.133	0.095	0.085	0.098	0.102	0.152	0.108	0.095	0.181	0.119	0.104	0.166	0.162	0.171	0.174	0.152	0.179	0.032	0.052
X7	0.137	0.215	0.152	0.150	0.170	0.209	0.130	0.161	0.131	0.221	0.173	0.155	0.209	0.202	0.212	0.233	0.222	0.216	0.043	0.096
X8	0.069	0.111	0.076	0.092	0.119	0.122	0.129	0.067	0.102	0.152	0.089	0.078	0.152	0.149	0.098	0.122	0.115	0.150	0.028	0.045
X9	0.061	0.099	0.068	0.068	0.078	0.100	0.104	0.117	0.048	0.138	0.080	0.070	0.140	0.136	0.087	0.097	0.092	0.146	0.025	0.050
X10	0.117	0.207	0.161	0.102	0.152	0.190	0.169	0.143	0.126	0.160	0.166	0.138	0.197	0.191	0.204	0.221	0.212	0.213	0.052	0.092
X11	0.089	0.166	0.115	0.076	0.106	0.122	0.118	0.083	0.071	0.172	0.077	0.097	0.131	0.126	0.139	0.176	0.160	0.159	0.043	0.075
X12	0.090	0.167	0.116	0.076	0.091	0.114	0.118	0.082	0.071	0.162	0.126	0.064	0.146	0.125	0.155	0.166	0.160	0.159	0.042	0.059
X13	0.145	0.203	0.144	0.115	0.134	0.196	0.166	0.140	0.108	0.214	0.163	0.121	0.136	0.188	0.201	0.218	0.209	0.201	0.035	0.089
X14	0.160	0.211	0.165	0.135	0.155	0.184	0.172	0.145	0.111	0.222	0.168	0.140	0.201	0.136	0.208	0.228	0.218	0.208	0.052	0.078
X15	0.140	0.179	0.152	0.099	0.115	0.138	0.128	0.090	0.077	0.185	0.117	0.087	0.142	0.136	0.119	0.191	0.184	0.182	0.034	0.063
X16	0.134	0.170	0.145	0.076	0.091	0.113	0.103	0.082	0.071	0.173	0.110	0.080	0.131	0.125	0.168	0.121	0.174	0.170	0.042	0.058
X17	0.134	0.170	0.145	0.076	0.091	0.113	0.103	0.082	0.071	0.173	0.110	0.080	0.131	0.125	0.168	0.179	0.115	0.170	0.042	0.058
X18	0.148	0.191	0.160	0.134	0.141	0.158	0.156	0.117	0.115	0.176	0.127	0.112	0.157	0.151	0.188	0.206	0.197	0.140	0.047	0.053
X19	0.031	0.028	0.023	0.018	0.021	0.051	0.025	0.020	0.018	0.032	0.022	0.019	0.028	0.027	0.029	0.031	0.029	0.030	0.008	0.060
X20	0.069	0.103	0.076	0.066	0.077	0.622	0.109	0.081	0.071	0.158	0.089	0.078	0.131	0.127	0.123	0.128	0.114	0.124	0.074	0.041

.

4.2. Determine the Degree of Influence, Degree of Cause, and Centrality of Each Factor

The comprehensive impact matrix was used to calculate the degree of influence, degree of cause, and centrality of each influencing factor according to Equations (6) and (8)–(10), as shown in

Table 5. Table of element indicator values.

Code	Impact Degree D Value	Affected Degree C Value	Centricity D + C Value	Cause Degree D − C Value (R)
X1	1.800	1.997	3.797	−0.198
X2	1.928	2.993	4.922	−1.065
X3	1.764	2.274	4.038	−0.510
X4	2.429	1.724	4.152	0.705
X5	2.551	2.110	4.661	0.441
X6	2.442	3.073	5.516	−0.631
X7	3.438	2.466	5.905	0.972
X8	2.064	1.940	4.004	0.123
X9	1.803	1.571	3.374	0.232
X10	3.211	3.271	6.482	−0.060
X11	2.302	2.177	4.479	0.126
X12	2.288	1.830	4.119	0.458
X13	3.128	2.866	5.994	0.261
X14	3.299	2.721	6.020	0.578
X15	2.558	2.943	5.501	−0.385
X16	2.338	3.326	5.663	−0.988
X17	2.338	3.125	5.463	−0.788
X18	2.872	3.164	6.036	−0.291
X19	0.548	0.745	1.292	−0.197
X20	2.462	1.246	3.708	1.216

. Figure 4 is Impact degree—affected degree graph. Figure 5 illustrates the centrality causality diagram. Figure 6 presents the comprehensive impact relationship diagram. Figure 7 provides the weight chart.

Figure 1 shows the schematic diagram of a cross-sectional structure, Figure 2 presents the layout of Weihe River Bridge span, and Figure 3 displays the overall view of the bridge.

4.3. Constructing a Multi-Level Hierarchical Model Based on ISM

Based on the results obtained from the Fuzzy DEMATEL analysis, the ISM method was employed for further analysis to reveal the hierarchical structure between the various influencing factors in the jacking construction. The adjacent order matrix and the identity matrix are summed using Formula (11), and the threshold value = 0.05 is determined by combining expert inquiry and repeated testing. The resulting reachability matrix is shown in

Table 6. Reachability matrix.

	X1	X2	X3	X4	X5	X6	X7	X8	X9	X10	X11	X12	X13	X14	X15	X16	X17	X18	X19	X20	Drive
X1	1	1	1	0	1	0	1	0	0	1	1	1	1	1	1	1	1	1	1	0	15
X2	1	1	1	0	1	0	1	0	0	1	1	1	1	1	1	1	1	1	1	0	15
X3	1	1	1	0	1	0	1	0	0	1	1	1	1	1	1	1	1	1	1	0	15
X4	0	0	0	1	1	0	1	0	0	1	1	1	1	1	0	0	0	0	1	0	9
X5	0	0	0	0	1	0	1	0	0	1	1	1	1	1	0	0	0	0	1	0	8
X6	0	0	0	0	1	1	1	0	0	1	1	1	1	1	0	0	0	0	1	0	9
X7	0	0	0	0	1	0	1	0	0	1	1	1	1	1	0	0	0	0	1	0	8
X8	0	0	0	0	1	0	1	1	0	1	1	1	1	1	0	0	0	0	1	0	9
X9	0	0	0	0	1	0	1	0	1	1	1	1	1	1	0	0	0	0	1	0	9
X10	0	0	0	0	1	0	1	0	0	1	1	1	1	1	0	0	0	0	1	0	8
X11	0	0	0	0	1	0	1	0	0	1	1	1	1	1	0	0	0	0	1	0	8
X12	0	0	0	0	1	0	1	0	0	1	1	1	1	1	0	0	0	0	1	0	8
X13	0	0	0	0	1	0	1	0	0	1	1	1	1	1	0	0	0	0	1	0	8
X14	0	0	0	0	1	0	1	0	0	1	1	1	1	1	0	0	0	0	1	0	8
X15	1	1	1	0	1	0	1	0	0	1	1	1	1	1	1	1	1	1	1	0	15
X16	1	1	1	0	1	0	1	0	0	1	1	1	1	1	1	1	1	1	1	0	15
X17	1	1	1	0	1	0	1	0	0	1	1	1	1	1	1	1	1	1	1	0	15
X18	1	1	1	0	1	0	1	0	0	1	1	1	1	1	1	1	1	1	1	0	15
X19	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0	1
X20	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	1	1
dependence	7	7	7	1	18	1	18	1	1	18	18	18	18	18	7	7	7	7	19	1	199

, and further analysis leads to the generation of a hierarchical decomposition table (

Table 7. Hierarchy breakdown table.

Level	Essential Factor
1st floor (top floor)	X19, X20
Layer 2	X5, X7, X10, X11, X12, X13, X14
3rd layer (bottom layer)	X1, X2, X3, X4, X6, X8, X9, X15, X16, X17, X18

) and system-directed graph (Figure 8):

4.4. Countermeasure Analysis and Method Warning

According to the results of the fuzzy DEMATEL analysis, it can be concluded that the importance of each factor in the system is reflected using centrality. Specifically, the higher the centrality, the higher the importance of the factor in the system. As indicated in Table 7, the top six influencing factors in the descending order of values are X10 (displacement during pushing), X18 (safety-management qualification), X14 (local buckling), X13 (steel beam overturning), X7 (collision with the pier during the installation of the guide beam), and X16 (incomplete on-site safety-management measures). These key factors closely influence each other and are prone to dangerous accidents during the construction of steel truss bridge jacking, thereby highlighting the need to prioritize prevention measures against them. The two factors of X18 (safety-management qualification) and X16 (incomplete on-site safety-management measures) can be addressed from a management perspective, including strengthening the supervision of safety qualifications and on-site safety management of construction units, as well as developing comprehensive safety-management regulations. Meanwhile, X10 (offset during pushing), X14 (local buckling), X13 (steel beam overturning), and X7 (collision with the pier when the guide beam is on the pier) are technical factors that could be managed by implementing technical disclosure and strengthening deep drawing.

Reason degree is an important indicator for categorizing factor attributes. Factors with a reason degree greater than 0 are considered cause factors, while those with a reason degree less than 0 are regarded as result factors. According to Table 7, there are a total of 10 cause factors and 10 result factors. The top five factors based on reason degree are X20 (natural environment impact), X7 (collision with the pier when the guide beam is on the pier), X4 (stiffness and strength of the assembly platform), X14 (local buckling), and X12 (failure of the limit device). Factors with higher positive cause values are more likely to have an impact on other factors while being less susceptible to their effects. Therefore, it is necessary to pay attention to these factors during the construction process, focus on prevention and control, and proactively prevent the occurrence of safety accidents.

A negative degree of causation indicates a higher vulnerability of factors to external influence. Among the factors analyzed, the top five ranked are X2 (illegal operation), X16 (incomplete on-site safety-management measures), X17 (inadequate safety education and disclosure), X6 (matching degree of guide beam design with steel beam), and X3 (fatigue operation). Therefore, in construction, it is not only necessary to control such risk factors but also to prevent the antecedents that lead to these factors.

The analysis of the ISM hierarchical model revealed that the risk and harmful factors leading to accidents during the launching construction of steel truss bridges can be divided into three levels. At the highest level, the direct contributing factors include X19 (environmental impact) and X20 (natural environmental impact), while at the middle level, the transition factors are X5 (settlement and displacement of the assembly platform), X7 (collision with the pier when the guide beam is on the pier), X10 (deviation during the launching), X11 (failure of the launching equipment), X12 (limit device failure), X13 (steel beam overturning), X14 (local buckling). The underlying essential factors are X1 (cognitive factors of construction personnel), X2 (illegal operation), X3 (fatigue operation), X4 (stiffness and strength of assembly platform), X6 (matching degree of guide beam design with steel beam), X8 (rationality of pier side bracket design), X9 (compliance of temporary pier design with standards), X15 (incomplete safety-management system), X16 (incomplete on-site safety-management measures), X17 (incomplete safety education disclosure), and X18 (safety-management qualification). From the analysis, it is evident that the essential factors contributing to accidents can be classified into two categories: inadequate management education and incomplete facility structure design. To address the former, it is crucial to enhance the safety awareness of construction personnel, focusing on the following measures.

Provide safety training: Offer thorough training programs to workers, focusing on increasing their potential hazards and risk awareness in the work environment, and ensuring they understand how to correctly use safety equipment and tools.

Establish a safety culture: Enterprises can foster a safety culture by implementing mechanisms for safety rewards and penalties. Issue and communicate clear safety standards and regulations, emphasizing the importance of safety to employees.

Conduct regular inspections and evaluations: Regularly inspect the workspace to assess the safety status, evaluate employees’ compliance with safety regulations and procedures, and promptly identify and solve potential safety hazards.

Hold engaging safety activities: Organize safety-promotion activities such as Safety Month, Safety Week, and Safety Day, using diverse forms of media and activities to enhance employees’ safety awareness.

Establish effective supervision and implementation: Enterprises can implement a robust responsibility system to supervise and manage workers’ safety works. Ensure the strict implementation of safety regulations throughout the organization, and establish a clear safety-responsibility system, wherein each employee understands their safety responsibilities and obligations.

To address the issue of incomplete facility structure, the following measures can be considered:

Enforce adherence to standard design specifications: During the construction process, strict enforcement of standard design specifications is essential to ensure that the design of structural facilities meets industry standards.

Establish a quality-control system: Establish a robust quality-control system that ensures thorough quality control and acceptance during the design process. Consider engaging third-party testing agencies to conduct testing and evaluations, thereby ensuring that structural facilities meet design requirements and quality standards.

Continuously monitor and revise design: During the construction process, it is necessary to maintain continuous monitoring and revision of the design plan. If problems arise, timely adjustments and improvements should be made to ensure that the design of structural facilities meets actual needs and requirements.

Establish a sound communication mechanism: Establish a clear and effective communication channel between designers, engineers, and construction personnel. This will facilitate openness and transparency and, therefore, prevent discrepancies during the design and construction phases.

Establish a culture of safety awareness: Cultivate a strong culture of safety awareness among designers, engineers, and construction personnel, emphasizing the importance of adhering to safety regulations and standards.

Establish a responsibility system: Implement a robust responsibility system that clarifies the responsibilities and obligations of each person involved in the construction process. Similarly, establish an effective supervision and punishment mechanism to hold accountable and punish those who fail to meet their responsibilities.

Through the above measures, the problem of imperfect structural facility design during construction can be effectively addressed. This will ensure the quality and safety of the project.

4.5. Effectiveness Analysis of Improved DEMATEL–ISM Method

According to the improved DEMATEL–ISM analysis and evaluation method, considering four aspects, 20 risk factors that may lead to safety accidents in the top-pushing construction of steel truss bridge were identified. Subsequently, new safety-management policies were formulated, categorized from the perspectives of ‘human, material, management, and monitoring’.

To address factors X1 (cognitive factors of construction personnel), X2 (violations of regulations), and X3 (fatigue operations), it is necessary to strengthen safety technical disclosures, deepen safety awareness, and enhance skill proficiency for workers. For example, organizing a monthly safety production activity could significantly improve workers’ safety consciousness.

For factors X5 (assembly platform settlement and displacement), X10 (deviation during top-pushing), X11 (top-pushing equipment failure), X13 (overturning of steel beams), and X14 (local buckling), focusing on the aspects of equipment and monitoring, the quantity of safety recorders, and monitoring alarm devices had been increased on the top-pushing platform, alongside temporary pier caps and guide beams. Implementing 24 h uninterrupted monitoring for the entire bridge significantly reduced the occurrence of unsafe incidents due to equipment shortages, human errors, and delayed detection.

In terms of management, there was an emphasis on professional knowledge training for management personnel and in-depth disclosure of bridge drawings, allowing management personnel to have a more comprehensive understanding of the entire bridge. Three months after implementing the new safety-management policies, a satisfaction survey regarding the new safety-management policies was distributed to all personnel in the project department. The satisfaction rate for the 20 management personnel and 40 workers who participated in the survey was 96.6%. Further comparison with past construction logs revealed a significant reduction in the number of unsafe incidents, confirming the effectiveness of the improved DEMATEL–ISM analysis and evaluation method.

5. Conclusions

Bridges play a crucial role in people’s lives and the economic development of regions. As China’s bridge-construction scale continues to expand, addressing safety issues related to bridges becomes imperative. This study focuses on the top-push construction of a steel truss bridge, aiming to identify and analyze potential risk factors that could lead to accidents. The improved DEMATEL–ISM method is employed, identifying 20 risk factors from four key aspects and ranking them based on centrality and causal relationships. The higher the centrality, the more significant the factor in the system. Arranged in descending order, the top five highly prioritized risk factors are as follows: X10 (displacement during pushing), X18 (safety-management qualification), X14 (local buckling), X13 (overturning of steel beams), and X7 (collision with bridge piers during guide beam installation).

For factors X10 (displacement during pushing), X14 (local buckling), X13 (overturning of steel beams), and X7 (collision with bridge piers during guide beam installation), from a monitoring perspective, corrective measures can be implemented by increasing bridge monitoring points, enhancing monitoring during the top-push process, and real time observation and data collection to correct deviations in the pushing process. For X18 (safety-management qualification), it is necessary to strengthen safety-management qualification screening from a management perspective. Additionally, efforts should be made to enhance the safety awareness of construction personnel, focusing on the following points: Provide Safety Training; Establish a Safety Culture; Inspect and Evaluate; Conduct Safety Activities; Ensure Implementation.

The hierarchical analysis of the ISM method provides a visual representation of the interrelationships between each factor. Eleven fundamental factors have been identified as causes of accidents during the launching construction phase, with environmental factors confirmed as the key influencing factors, including X19 (impact of surrounding buildings) and X20 (impact of the natural environment), particularly X20. The bridge discussed in this paper is situated near a river in the city of Xi’an, characterized by a temperate continental monsoon climate. During the summer and autumn seasons, the bridge is often affected by tropical cyclones, leading to hazardous weather conditions such as thunderstorms, strong winds, and hail. Consequently, there is a high demand for the bridge’s overall rust resistance and the pier’s flood resistance. Therefore, during the bridge-construction phase, it is advisable to reinforce the flood-resistant structure of the piers and apply paint to the steel structural components of the bridge to resist corrosion. Additionally, regular monitoring of river water levels should be conducted to address natural disasters such as floods. The mentioned preventive measures and improvement methods offer a reference for effective safety protocols in the future launching of the construction of a truss steel bridge over water.