Research On Reliability Evaluation And Optimization Method For Micro System On Intelligent Ammunition

The development of miniaturization in modern intelligent ammunition has been firmly promoted by the research and application of miniaturization technology,which further promotes the capability of the new ammunition in precise strikes and efficient damage.Because of the miniaturization of Safety and Arming device(S?A device)in fuse,there are more space for the detection circuit with multisensors and Main Charge,which then enhances the precision and power of the ammunition.However,present designs of the micro system usually cannot achieve the high reliability demanded and effective methods should be developed for the reliability assessment and improvement during the design stage.The paper mainly focuses on the issues arising from the miniaturization of intelligent ammunition and its reliability promotion based on miniaturization technology.It basically contains such problems as quantitative assessment of micro system with adhesion failure,reliability analysis of structure with coupling fields,reliability analysis with correlated failure modes,and so on.The main contents of the paper are as follows:(1)Typical failure modes of micro structures such as fracture and adhesion are analyzed,and the failure mechanisms are figured out through simulations and experiments.The external loads,geometric dimensions and material properties are identified as main factors in structural reliability analysis.The basic principles of sampling technique and iterative method based on MPP for structural reliability analysis methods are proposed and the corresponding probabilistic sensitivity methods are presented.The integration method of deterministic simulation programs for efficient micro structure reliability simulation is proposed.Details about simulation parameters updating,simulation program driving and simulation result extracting are discussed.Those methods can applied for efficient reliability simulation in digitized design and realization of micro structures.An efficient least square approximation method is presented,where a straight step calculation technique in gradient descent algorithm is achieved.A new optimization method for correlation function parameters is given to improve the efficiency of Kriging interpolation method.(2)A method for micro structure adhesion reliability analysis based on energy method is introduced,where the failure probabilities of the cantilever in micro accelerometer with various external impact loadings are calculated and the loading,thickness and length of the cantilever are identified as the key affecting factors.Analysis method for micro switch reliability evaluation through reduced-order model method is given.Finite difference method is used for the modeling of reduced-order model considering multi-field coupling and Krylov subspace algorithm is adopted for its calculation.Reliability model for a micro electrical switch with fluidic-electrical-mechanical coupling,where the failure probabilities of the switch under different driving voltages are calculated and the voltage and length parameter are figured out as main factors.(3)The method for measuring the correlation of two failure modes is presented based on their performance functions.Numerical examples are used to demonstrate how the system reliability is affected by the correlations.General reliability models for series system,parallel system and hybrid system are established based on probabilistic fault tree method.The transformation method between a hybrid system and a serial-parallel system is proposed,where the method about how to calculate the correlation coefficients of different sub-parallel systems in the serial-parallel system is discussed.A more efficient reliability calculation method based on PCM for multiple failure modes is proposed.The numerical integration approach is used for the calculation of bivariate normal distribution problem,which improves the precision of conditional normal fractiles in the proposed method.It then updates the corresponding coefficient matrices based on the precise fractiles which then efficiently overcomes the inaccuracy of original PCM method especially being used for the reliability evaluation of series system.Probabilistic sensitivity methods are presented for reliability problem with multiple failure modes.(4)The reliability of a micro gyroscope with multiple failure modes is analyzed based on the modal simulation and impacting analysis.Performance functions for gyroscope instability,damping characteristic dissatisfaction,impacting failure and initial electric capacity dissatisfaction problem are presented.The correlations among multiple failure modes and the system reliability are analyzed,where the key failure modes and variables are listed.(5)Reliabilities in objective or constraint can be transformed to reliability indexes or probabilistic performance measurements in the modeling procedure.Improvement of reliability optimization loop strategy is studied,where an improved search algorithm for most probable point of inverse reliability is presented.The search direction and step size are controlled in the iteration procedure of the proposed method.Numerical results are used to demonstrate the better convergence comparing with the AMV method.A reliability optimization model is established,where the maximization of the sensitivity of micro gyroscope is as the optimization goal.And reliabilities corresponding to the failure modes are required not to be less than 0.99.The deformation of the removable comb drive is constrained to be less than 10?m.Results of the reliability optimization show that the sensitivity of gyroscope is improved 48%when all the constraints are satisfied.The feasibility and efficiency of the proposed methods are also demonstrated by the example.(6)The reliability simulation method is applied to the reliability assessment of a micro S?A device in fuse.Micro structure designs are presented according to the function requirements of the device.Feasibility of the designs is validated through deterministic FE simulations.It utilizes performance function to represent the failures of the device mathematically,and then constructs the corresponding reliability models according to structural reliability theory.The reliabilities of the key structures in the device are evaluated,where the randomness of geometric parameters,material properties and loading are taken into account.The significant factors are also identified for further structural improvements.A prototype of the micro S?A device is manufactured.which is further delivered to centrifugal experiments.Results from the experiment demonstrate the feasibility of the design.All the presented methods and procedures in the paper can be also used for the quantitative reliability evaluation of other micro systems,which may contribute to the applications of micro system in various engineering problems.