Process reliability evaluation based on quality characteristics evolution

To evaluate the impact of machining process on product reliability quantitatively, a comprehensive evaluation method for process reliability was put forward based on Granger causality test and Cox Proportional Hazards Regression Model (PHRM). Granger causality test was conducted on the variable series of process characteristics and product characteristics to model the quality characteristics evolution. The quantitative relation between processing deviation and process quality deviation was determined to provide evidence for establishing accurate evaluation model. The process covariate effected on production reliability was confirmed based on evolution rule, and Cox PHRMs with cross term and without cross term were established. Meanwhile, the hazards ratio was given as an index of Cox model to judge the criticalities of different processes. The product reliability and life expectation were calculated under the Weibull Cox PHRM assumption. The case study on process reliability evaluation of radar phase-shift unit was demonstrated to prove the feasibility and effectiveness of the proposed method.