Non-probabilistic Reliability Analysis And Optimization Of The Frame

The frame is an important load-bearing parts of the truck, carries all kinds of load. The structure performance is directly related to the comprehensive performance of the whole truck. After designing, this article uses the finite element analysis method for structural analysis, preliminarily understands the mechanics performance of frame structure. Traditional frame structure analysis is based on the deterministic, in fact the parameters have certain fluctuations, and this analysis will produce a certain error. Probability reliability theory can well describe the uncertainty, but it needs the probability distribution of parameters accurately, sometimes it is not easy to get on the project. Non-probabilistic convex theory describes the uncertainty with interval or ellipsoid; it only needs to know the parameter interval, and is structured simply, has strong applicability. Based on the reliability, this article structures the reliability optimization model of the frame. The main work is as follows:The frame geometry model needs to carry geometry cleanup, to apply boundary condition, to mesh with HyperWorks software. According to the truck running conditions, to apply load and constraint condition respectively, to analysis frame’ strength and stiffness in full working condition of bending and torsion condition, full of turning braking mode and full load condition, then to carry on the dynamic analysis of frame, the first ten frequencies of the frame are obtained.Taking the frame material elastic modulus and density as the uncertain parameters, so an interval model of the uncertain parameters can be constructed. Using Kriging model to build the stiffness and strength approximate function of frame in the design variables and uncertain parameters, selecting the optimization method of the interval model to calculate the non-probabilistic reliability index of stiffness and strength.Building frame ellipsoid model of uncertain parameters, and introducing in detail the method of ellipsoid method, building frame stiffness and strength approximation function with polynomial response surface method, using modified first order second moment method (iHL-RF) to solve the non-probabilistic reliability indices of stiffness and strength.Building the non-probabilistic reliability optimization mathematical model based on the non probabilistic reliability analysis, the objective function is minimizing the frame quality. The optimization model is solved with IP-GA. This algorithm can improve the solution efficiency. Then the optimized results should to be analyzed in order to verify the correctness of the optimization results.