基于腐蚀损伤表征因子的疲劳寿命衰减影响研究

Research on Influence on Fatigue Life Attenuation Based Corrosion Damage Characterization Factors

应用MIT(Mean Impact Value)方法对LY12CZ铝合金试件疲劳寿命产生影响的腐蚀损伤表征因子进行筛选,得到了对疲劳寿命衰减影响较大的5个腐蚀损伤表征因子。定义了腐蚀疲劳寿命累积衰减函数与疲劳寿命衰减速率函数,建立了寿命累积衰减模型,验证了该模型的准确性,并以腐蚀损伤表征因子和腐蚀累积衰减函数为数据样本,用BP神经网络、自适应滤波的LMS算法分别预测了不同年限下的疲劳寿命。与实验测得的疲劳寿命数据对比后得出,BP神经网络、LMS方法计算产生的误差在工程上可以接受。

Corrosion damage characterization factors of LY12CZ aluminum alloy were filtrated using MIT（Mean Impact Value）method.Five corrosion damage characterization factors were obtained,which has significant influence on fatigue life attenuation.The functions of cumulative corrosion fatigue life and cumulative fatigue life attenuation rate were defined.The cumulative corrosion fatigue life model was established and its accuracy was verified.Corrosion damage characterization factors and the cumulative fatigue life attenuation function were taken as specimen.The fatigue lives of LY12CZ aluminum alloy of different service time were predicted with BP neural network and LMS algorithm.The results were than compared with experimentation results.It was proved that the error generated by BP neural network and LMS algorithm can be accepted in engineering.