Interactive Buckling in Prestressed Stayed Members and Load-Carrying Capacity Intelligent Evaluation Method
Prestressed stayed members have beautiful shape and high material efficiency,but they are susceptible to multi-modal interaction failure so-called"interactive buckling"owing to their slender configuration,which can affect their load-carrying capacity and buckling behaviour significantly.This paper investigates the interactive buckling in the symmetric and asymmetric prestressed stayed members as columns and beam-columns,such as its effects on buckling behaviour and load-carrying capacity,and develops the load-carrying capacity evaluation model via artificial neural network(ANN)based on regression algorithm.The results show that,for the prestressed stayed member,modal interaction can lead to unstable post-buckling behaviour and significant load-carrying capacity reduction.This reduction is dominated by structural geometric parameters and the buckling loads of interactive modes,and the reduction degree is affected by stay asymmetry and prestressing levels.The proposed ANN model for load-carrying capacity evaluation can take interactive buckling into accounts,and its accuracy,reliability and applicability are acceptable.This study can provide research foundation for developing the intelligent design software of the prestressed stayed member,which can satisfy the requirement for structural design in engineering practices.
prestressed stayed memberfinite element analysismachine learninginteractive bucklingload-carrying capacityartificial neural network