Steel failure is always defined by phenomena such as yielding and buckling. A steel structure may fail under complex conditions such as cyclic or biaxial loading, but determining its critical stress is complex. A set of cyclic loading tests of H-steel columns with material Q345B and length 1500mm is selected as a case study, and thermodynamic-based failure analysis theories and methods are applied to reveal their failure loads. This work first transforms amplitude strain data under cyclic loading into state variables and divides them into six parts. The network-free renormalization method (based on the relativity of state variables/subparameters) can be used to characterize the stressing state evolution of H-steel columns. Further, the phase transition loads can be detected by applying clustering analysis criteria because of their attractor effect. On the other hand, Wilson's phase transition theory defines phase transition loads as the fixed points of renormalization, which can be verified by comparing the before and after renormalization (part and overall) results. Lastly, the economy index (EI) is defined as the ratio of the biaxial moment triangle area S2 to the steel column volume at the elastoplastic branch (EPB) point to verify the safety and economy of the EPB-based seismic performance design for H-steel columns. EPB-based designs increase the EI of Class IV and III H-steel columns by 258.8% and 365.8%, respectively. Thus, this work provides a new reference for thermodynamic-based failure and seismic performance analysis of H-steel columns.
Read full abstract