Storage Reliability Prediction Of Aerospace Relay Based On Stress Relaxation Simulation

Most of the control systems for missiles and other weapons contain a large number of aerospace relays.The reliability of aerospace relays plays a decisive role in ensuring if they are readily available.Aerospace relays are mainly stored in the service state of weapons and equipment,and their storage period usually lasts for several years or even decades.Considering the high reliability and long life of aerospace relays,it is difficult to effectively predict the storage reliability in a short period of time even if the accelerated test method is adopted.Especially in the design phase of aerospace relays,the test time required for multiple rounds of design modifications and corresponding reliability prediction tests is difficult to estimate.In order to solve the problem that the space relay storage test is time-consuming,costly and its storage reliability can not effectively predict at the design stage,in this paper,by combining the stress relaxation degradation simulation of key components(reeds)with the relay virtual prototyping technology,the simulation of the batch relay storage degradation process is realized,and then the reliability is effectively predicted.Firstly,for the problem that the stress relaxation of the reed cannot be directly measured,the reed material(AgMgNi0.24-0.29 alloy)is selected as research object,and based on the sheet bending stress relaxation method in the American E328-13 standard,the bending stress relaxation test platform for equal stress samples is designed and manufactured.For the bending stress relaxation test platform for equal stress samples,and the stress relaxation test data for the subsequent simulations were obtained by stress relaxation tests on 16 samples under 4 temperatures and 4 initial stress conditions.Secondly,the stress relaxation behavior is described based on creep theory and viscoelastic model respectively.The stress relaxation test data is transformed into the material property parameters required for stress relaxation simulation,and verified by simulation analysis of stress relaxation test process.Feasibility of stress relaxation simulation.Aiming at the heterogeneous stress distribution of the cantilever beam reed and the regional difference of the stress relaxation process,the stress relaxation simulation of the whole reed is realized by the method of partitioning the reed according to the different stress distribution.Thirdly,in order to reflect the influence of individual relay differences on the degradation process and reliability of batch product storage,based on the static analysis of relay mechanical system and electromagnetic system,the dynamic simulation model of relay is established.At the same time,10 sets of manufacturing process data with significant influence on the output characteristics of the relay are selected,a plurality of virtual samples of the relay are generated,and qualified virtual samples are selected according to the requirements of the relay technical parameters.On this basis,the storage degradation simulation of the batch relay is realized by injecting the stress relaxation simulation result of the reed into each qualified virtual sample.Lastly,the storage reliability of the batch relay is predicted based on the storage degradation simulation data.According to the phenomenon that he simulation degradation data is strictly monotonically increasing,the storage degradation model of a single relay is given based on the inverse Gaussian process.In order to make the model simple and easy to process,the simulation data is time-scale transformed by the failure physical analysis,so that it approximates a linear degradation trend.The maximum likelihood estimation method is used to estimate the parameters of each relay's degradation model.The probability map is used to verify that the degradation process of each relay's key performance parameters is subject to the inverse Gaussian process.On this basis,the storage reliability of the batch relay is predicted by superimposing the failure distributions of multiple virtual relays.