The Accelerated Degradation Test Method Based On Multiple Performance Parameters Of The Study

We can evaluate the reliability by operating accelerated degradation test for the productswith high reliability and long life. But the products often has more than one failure mechanism,itmeans the products’ multiple performance parameters degradating at the same time; and aperformance parameters of the degradation may accelerate or inhibit another performanceparameters of degradation,it means there is correlation between the multiple performanceparameters’ degradation. In this context,it is not comply with engineering practice to evaluate thereliability of the products limited to a single performance parameters of degradation.So, in thisdissertation, researched the reliability statictical model，the statistics analysis method of test data,the optimal design of constant-stress accelerated degradation test plan and simulation evaluationmethod based on multiple performance parameters.During fulfill this work,the main researchincludes:Firstly, briefly stated the background and meaning of the dissertation,the methods and trendsfor reliability test.The reliability statistics model and the statistics analysis method of test dataand the optimal design of test plan for accelerated life test and accelerated degradation test werereviewed.The problems of the reliability research considering the multiple performanceparameters’ degradation of the product’s and the current research situation and problems of servoelectro-hydraulic were generalized.Secondly, The product’s reliability statistics model was established based on multipleperformance parameters and them correlation. With electro-hydraulic servo mechanism forexample,by identifing failure mechanism of electro-hydraulic servo mechanism under storagecondition, the major failure modes of seal failure was confirmed,and established reliabilitystatistics model of (Nitrile Butadiene Rubber)NBR O-Ring used on electro-hydraulic servomechanism.Thirdly, the statistics analysis method of test data based on multiple performanceparameters was researched. Model parameters of the reliability statistics model were estimatedby the method of least squares estimation. With NBR O-Ring for example, by operatingdiagnostic test,A rough estimation of the parameters for the model was got,which provides thebasis for the optimal design of test plan and simulation evaluation.Fourthly, the rule and method of optimal design based on constant-stresses accelerateddegradation test was proposed; minimum variance of product’s reliability estimated is definedas goal with products under working stress level and to the service life of the specified;and eachtest stress levels and distribution proportion of test samples are adopted as design variables,the optimal design mathematic model of accelerated degradation test based on multiple performanceparameters is built by using maximum likelihood estimation theory.The result about the optimaldesign of accelerated degradation test plan of NBR O-Ring shows that the optimized test plancan raise estimation precision of reliability estimated,also cut down the numbers of test samplesunder the same degree of confidence.Fifthly,the method of simulation evaluation of constant-stresses accelerated degradation testplan was proposed.By using the method of Mento Carlo,with criterion of mean variances andstandard distance values of estimated product’s reliability estimated,was used to evaluaeconstant-stresses accelerated degradation test plan for O type sealing ring. The results show thatthe optimized test plan is more reasonable and feasible.Finally,a conclusion was made and the future research about accelerated degradation testbased on multiple performance parameters was briefly presentated.