摘要
以空气中腐蚀疲劳裂纹扩展速率为基准,引入环境加速因子Cenv,且把Cenv与Cair合并为C,用于表达07MnNiCrMoVDR钢在硫化氢溶液中的腐蚀疲劳裂纹扩展速率。利用Matlab中的神经网络工具箱,以介质浓度、载荷频率、应力比作为输入节点,C作为输出节点,训练该网络预测不同条件下的裂纹扩展速率。根据训练完毕神经网络预测不同条件下的C,利用Original找出这三个影响因素在裂纹扩展速率系数C中的关系表达式,从而由Matlab中的统计工具箱回归出07MnNiCrMoVDR钢在硫化氢溶液中的腐蚀疲劳裂纹扩展速率数学模型。
The formula of crack growth rate in H2S solution could be expressed with C combined by environment accelerative coefficient Cenv and Cair, based on the formula of crack growth rate in air. Neural network toolbox was used to train the network, in which medium concentration and load frequency and stress ratio were regarded as input nodes and crack growth rate coefficient C was regarded as output node. Crack growth rate under different environments can be predicted with this network. On the basis of the datum predicted by trained network, the relationship between crack growth rate and these three influencing factors can be gained by using original software. So the maths model of crack growth rate of 07MnNiCrMoVDR under H2S solution can be regressed with statistical toolbox in Matlab.
出处
《压力容器》
2003年第8期1-3,52,共4页
Pressure Vessel Technology
