Assessment of maximum and minimum hanger forces in a triple-cable suspension bridge under live load: Analytical algorithm with manta ray foraging optimization

Abstract

Triple-cable suspension bridges (TCSBs) are widely used in sea-crossing, island-linking projects with extra-long spans and ultra-wide decks. Their safe operation requires sufficient resistance of all bridge components (towers, cables, girders, and hangers) to static/dead and cyclic/live loads. Therefore, reliable assessment of maximum and minimum axial forces acting on each hanger is topical. This study proposes an analytical algorithm for assessing the internal forces and deformations of a TCSB. Specific steps and formulae for solving the coefficients are described by taking the estimation of girder response under random eccentric line loading as an example. Parameters of the suspension bridge under dead load are treated as known quantities. Actual engineering scenarios, including girder displacements along the bridge, are analyzed and discussed in detail. Basic unknown quantities are determined via governing equations under conditions of elevation differences’ closure and conservation of main cables’ unstressed lengths, yielding direct solutions for deformations of the main cables, girder, towers, and hangers. The proposed analytical model’s feasibility is verified via the finite element method. Optimal solutions to the above equations are derived by the manta ray foraging optimization (MRFO) algorithm to find the most unfavorable load distributions for all vehicle lanes of the bridge and the maximum and minimum axial forces of random hangers. Finally, the proposed method is successfully applied to a TCSB with a 1600 m main span under two operating conditions. The most unfavorable load distributions and the maximum and minimum axial forces of hangers at the mid-span and at the quarter-span point are estimated. The calculation results are compared against those obtained by the influence surface method, showing their good correlation and superiority of the proposed method in computation workload.