Reliability Of The Fatigue Life Of The Contact Area Of ​​the Genetic Algorithm-based Study

The conventional method is not exact and flexile for the contact problem if the contact model's strain breadth is simulated in consider of the randomicity of material parameter, geometry parameter and load condition. So in this text the strain breadth of high nonlinear contact position is simulated by Artificial Neural Network. The network structure and connect power value is optimized by Genetic Algorithms. And then, the strain-life model is established. The randomization of the Masson-Coffin formula is attained by two related random parameters. Considering the material test data, the strain-life formula under consideration mean stress revise is matched. The fatigue life correlated with reliability is simulated by Monte-Carlo method.The following is the method of strain breadth simulation. (1) The network structure and connect power value is coded by binary coding and float coding of Genetic Algorithms;(2)The target function value is computed according to the individual performance type. The fitness function is obtained by optimizing the target function;(3)The higher fitness value individual is operated by proportional Select Operator, two point Crossover Operator or three point Crossover Operator and simple Mutation Operator. After completing the optimization of the network structure and connect power value, a choiceness performance Artificial Neural Network is formed;(4)The strain breadth of contact dangerous position under influence of the random parameter is simulated by the Artificial Neural Network;(5)The above-mentioned method is verified by Hertz classical contact problem.The contact modal of turbine disk and blade rabbet is made aiming at the high reliability. The 3D heat-structure coupling modal is computed by ANSYS and the dangerous position is confirmed. The 15 scattering point of strain breadth is computed by APDL language of center-composition in probability design mold. The teacher signal is intervened referring to the points. The dangerous position strain breadth of the contact modal is simulated by the Artificial Neural Network which is optimized by Genetic Algorithms using C++ language. Then the fatigue life of the contact modal is analyzed by the method of this text. The result is better accord to the engineering fact.