Optimal Design of Failure Mode Control for Semi-rigid Steel Frame Based on Elitist Retained Genetic Algorithm

Abstract

The post-earthquake reports show that failure mode control design is of great significance to improve the seismic performance of structures. To realize the optimal design of failure mode in semi-rigid steel frame, a failure mode optimisation design method based on an elite retained genetic algorithm was developed in this study. Firstly, a database of 96 extended end-plate connections was established using the ABAQUS finite element software to obtain the initial rotational stiffness and yield moment under failure criteria. Secondly, with section size as the optimisation variable, failure mode as the constraint condition, and total steel consumption as the objective function, elite gene transfer was realized through gene sequence replication, hybridisation and mutation. With a 10-story, 3-span, semi-rigid steel frame as an example is proposed to demonstrate the application of the proposed semi-rigid design method, and pushover and time history analyses were conducted on the optimized results to verify the accuracy of the proposed method. The analysis results showed that the stable and reliable optimal design results were obtained through 1317 generation genetic iterative calculations, and the energy consumption ratio of the semi-rigid connection (including the beam ends) was 95.82% and that of the column was 4.18%; in addition, 76% of the beam ends had plastic hinges to dissipate energy in the pushover analysis. This finding indicates that the optimized frame can achieve the global failure mode during earthquakes, and the design method can effectively improve the seismic performance of the frame.