Structural optimisation of a passenger seat according to crash injury criteria with artificial neural networks

Abstract

Ensuring occupant safety in passenger seats is one of the most challenging problems. This paper studied two-place passenger seats within the scope of ECE R14 and ECE R80 regulations. Finite element model of the seat was developed, and analyses were performed according to the ECE R14 static test, with results validated by the actual test. To accurately simulate the interaction between the dummy and plastic backrest cover of the passenger seat in a dynamic test of ECE R80, sample tests were conducted for the polymer material and material and failure models were created. To optimize the negative injury criteria results, the plastic backrest cover was divided into three sections and the thickness of each section was optimized with artificial neural networks. After the optimization, the seat met the regulatory requirements and weight reduction was achieved.