Torsional Effect And Parameter Optimization Analysis Of Main Girder Of Single-Cable Plane Concrete Cable-stayed Bridge

The single-cable plane concrete cable-stayed bridge has the advantages of large spanning capacity,good mechanical performance,wide driving vision and beautiful landscape,etc.It has good economy and applicability,and is a very competitive bridge type.With the rapid development of large-capacity and rapid transportation infrastructure in modern cities,the width of bridge decks continues to increase,and the torque borne by the structure under eccentric loads will become larger and larger.However,the cable-stayed cables of a singlecable-plane concrete cable-stayed bridge are anchored on the centerline of the main girder,and the cable-stayed cables can hardly provide effective torsional support for the main girder.Its torsional performance is one of the key performances to ensure the safety of bridge operation.Therefore,it is necessary to conduct in-depth research on the torsional performance of the main girder of single-cable plane concrete cable-stayed bridges.Relying on the Linhai Bridge,this paper analyzes the torsional performance of different main girder section forms in the bridge state,and combines deep learning and genetic algorithm to optimize the main girder section parameters.The details are as follows:(1)Based on Umansky’s second theory,this paper studies the torsional effect of box girder under eccentric loading.The stress calculation expressions of the thin-walled box girder when free torsion and constrained torsion occur,and the constrained torsion differential equation of the thin-walled box girder are deduced and analyzed.(2)Based on Umansky’s second theory,the calculation formula of torsional geometric properties of single-box three-chamber box section with inclined web is derived.The torsional effect of the concrete box girder is calculated by the analytical method and the finite element method of the plate and shell respectively,and the distribution law of the torsional stress of the main girder under the eccentric load is obtained.The trend of the analytical solution and the numerical solution is basically the same,and the results are more reliable.(3)Taking the vertical displacement of the section,the torsional distortion angle and the longitudinal normal stress as the key indicators,the influence of the section parameters of the main girder on the torsional performance of the full bridge in the bridge state is analyzed.The results show that the top plate and the sloping web have significant effects on the torsional performance of the section in the section size parameters;the beam height has the most obvious effect on the section torsional performance in the section shape parameters.(4)With the Linhai Bridge as the engineering background,a large number of data samples are constructed based on ANSYS for deep learning BP neural network learning and training,and the nonlinear mapping relationship between the input vector and the output vector is established.Finally,using the mass of the main beam as the constraint function,the genetic algorithm is used to optimize the objective function—the maximum torsional distortion angle of the mid-span section,and the section parameters for the rational arrangement of each chamber are obtained.The results show that the optimization method of torsional performance of single-cable plane concrete cable-stayed bridge in this paper is feasible,and this method can also be applied to the optimization of torsional design of similar bridges.