Lateral Asymmetric Cable Force Optimization of Cable-Stayed Bridges with Plate-Truss Composite Girders

To accommodate the alignment of approach roads on both banks, some long-span cable-stayed bridges with plate-truss composite girders adopt a laterally asymmetric lane arrangement. However, eccentric loading induces significant torsional effects in the main girder, which traditional cable force optimization methods struggle to account for effectively. To mitigate the adverse effects caused by lateral asymmetry, this paper proposes a stepwise optimization strategy for the finished bridge cable forces considering lateral asymmetry, based on the decoupling characteristics of vertical bending and lateral bending-torsion coupling in left-right symmetric cross-sections. Firstly, an equivalent single-beam model is established under the condition of consistent bending and torsional stiffness to facilitate cable force optimization analysis. The displacement response error between the simplified equivalent model and the refined model is within 6%. Subsequently, a stepwise cable force optimization procedure is proposed, consisting of two stages: average cable force optimization and cable force difference optimization. The first stage aims to minimize the bending strain energy to determine a reasonable distribution of average cable forces, while the second stage adjusts the cable force difference between the two sides to counteract the eccentric torque induced by the self-weight of the main girder. To reduce the variable dimensionality, B-spline curves are employed to replace the original dense cable force variables, with the control points of the spline curves serving as optimization variables. A gradient-based algorithm is used for solution. The results demonstrate that cable force difference optimization effectively balances the eccentric torque caused by the girder self-weight and significantly reduces the torsional angle of the main girder, thereby enabling the bridge to achieve the desired reasonable internal force state and configuration. Moreover, the stepwise optimization strategy significantly reduces computational cost through problem decoupling and dimensionality reduction, reducing the number of finite element analyses by approximately one-third while maintaining optimization accuracy and convergence stability.