In order to improve the performance of wet joints in prefabricated ultra-high-performance concrete (UHPC) bridges, a comparative study was conducted to investigate the flexural behaviour of UHPC joint beams. Experimental tests were implemented on a complete beam and on five further specimens with, respectively, vertical joint, rhombus joint, vertical joint with strips, tied rhombus joint with strips, and welded rhombus joint with strips. The specimens’ failure modes, ultimate flexural capacities, strain distributions and load–deflection curves were investigated. The test results show that the flexural capacity of the joint beams is lower than that of the complete beam. Among the five specimens with wet joints, the welded rhombus joint with top and bottom strips had the greatest stiffness, the highest cracking resistance capacity, and the best flexural performance. From the best to the worst performing, the remainder of the wet joint specimens took the following order: the unwelded rhombus joint with strips, the vertical joint with strips, the rhombus joint and the vertical joint. In accordance with the test results and the related theoretical analysis, this paper proposes formulas to calculate the flexural capacity of the beam with UHPC joints.

A combined analytic hierarchy process (AHP) method, Delphi method, and cloud model, i.e., a cloud clustering–based group decision method, is proposed. The risk probability and risk loss of the main risk factors are scored by experts. A backward cloud generator is used to calculate the digital characteristics of the sample from the experts’ scoring. A forward cloud generator is used to generate the cloud droplets according to the scoring samples from the experts. An iteration method is approved to quantify the coherence. Introduction of the expert authority, group decision risk probability, and risk loss can be obtained based on the cloud model. Similarly, the group decision weight of risk resources can also be obtained by the cloud-clustering group decision method. According to the risk resource weight, risk probability, and risk loss, risk level can be determined based on the risk judgment matrix. The study results show that the risk level of the example cable-stayed bridge is Level II, which is very similar to the computation achieved by a method derived from the current safety risk assessment guide in China. However, cloud clustering can ensure higher coherence for the samples. Consequently, the risk evaluation result will be more stable. The real bridge example and comparison analysis show that the cloud-clustering group decision method is reasonable and can objectively estimate the bridge risk.

AHP-GST method for bridge risk assessment is proposed. The expert scoring method is used to get the risk assessment data sample, and the analytic hierarchy process (AHP) is combined with the grey system theory. The risk possibility is calculated by analytic hierarchy process, and the weight matrix of each risk objective layer is obtained.The grey system theory is used to judge the severity of the risk, and the whitening value weight matrix of each index is obtained. Five kinds of risk weights are calculated and the engineering risk grade is determined according to the principle of maximum membership. Taking Jiangshun Bridge as an example, the AHP-GST method is used to evaluate the risk of the bridge's substructure in construction stage. The result of risk assessment is consistent with that obtained by the current guidance method, which shows that the method is reasonable and can be applied to the risk assessment of bridge construction.

A practical combined algorithm for the corbel of common reinforced concrete bridge is put forward. Based on material nonlinearity, combining the professional finite element software Midas Civil and MIDAS FEA with Smeared crack model, the Thorenfeldt and Constant curves are used as concrete compression and tension constitutive model to numerical analyze the corbel's stress and crack. Example results show that the normal stress of the corbel and crack are greater than design value of C35 concrete tensile strength, and part areas are in the plastic state. The groove corner region of the corbel has abrupt stress. The crack of the corbel is firstly generated at the groove and the top plate, and gradually expands with the increase of the load coefficient. It develops sufficiently in the groove region, but there is not any full-length crack. The corbel has a group-strip-like crack farther from groove corner region. The calculation results are consistent with the bridge disease condition, which shows that Finite Element-Smeared Crack Combined Algorithm proposed in this paper is effective for the analysis of corbel.