Energy-absorbing characteristic prediction of thin-walled structure based on explicit finite element
XIE Su-chao;TIAN Hong-qi;ZHOU Hui
（Key Laboratory of Traffic Safety on Track (Central South University), Ministry of Education; School of Traffic & Transportation Engineering, Central South University, Changsha 410075, Hunan, China）
Abstract： In order to discuss the influence rule of experimental parameters on energy-absorbing characteristic of thin-walled structure, predict and analyze energy-absorbing characteristic of thin-walled structure of the same type, the prediction model of thin-walled structure of the same type was set up based on BP(Back Propagation) artificial neural network and explicit finite element. Take the square thin-walled structure for example, to obtain various energy-absorbing characteristic parameters under different experimental conditions, plentiful numerical experiments were performed on thin-walled structures of the same type by changing the influence factors of energy-absorbing characteristic such as structural mass, impact velocity, cross section size, wall thickness and so on, then the energy-absorbing characteristic prediction model of thin-walled structure of the same type was set up, and the network model was trained till getting the desired network error by 2035 steps. The results show that the output samples of the BP artificial neural network match the aim well, the difference of specific energy absorption(SEA) is -2.53％, the difference of effective impact force is 4.67％ and the difference of effective impact journey is -3.90％, which indicates that the model has relatively high accuracy.
2010, Vol. 29 
