Study on sensitivity analysis of tooth surface roughness parameters and contact stress

The reasonable selection of three-dimensional roughness parameters can evaluate the service performance machined surface, and its excessively high dimension greatly increases difficulty selection. In order to reveal correlation law between 3D tooth surface contact parameters, analysis theory is used explore correlation, nonlinear mapping model constructed by using BP neural network, genetic algorithm optimization Morris Sobol′ two global sensitivity. qualitative method conducts quantitative each parameter, selects that have greatest impact on determines primary secondary relationship, solves problem parameter redundancy, provides theoretical basis for anti-fatigue design manufacturing surface. results show overall influence maximum shear stress gear pressure are arranged in following order: arithmetic average height ( S a ) peak volume V mp ), which has strong correlation.