| Machine tools are main equipments in the equipment manufacturing industry,and their quality directly affect the quality and cost of mechanical processing.With the continuous improvement of the requirements for the quality and precision of mechanical products,machine tools are developing in the direction of high precision and high reliability.The overall performance is an important sign that reflects the modernization level of a national machinery industry.Machine tool spindles,gears,bearings,ball screws,etc.,which are key components of machine tools,have an important impact on the improvement of machine tool performance and the improvement of processing accuracy.Among them,the quality of machine tool spindles directly affects the accuracy of processed products.Therefore,the reliability evaluation and optimization design of the machine tool spindle have very important theoretical research value and engineering practical value for ensuring the reliability of the machine tool.The main contents of this article are as follows:1)Firstly the basic theory of reliability is introduced,including the basic concepts of random variables,function functions,reliability,failure probability,and reliability indicators,and then common reliability analysis methods are introduced,including:moment method and response surface method.2)In order to ensure the reliability constraints of the reliability-based design optimization(RBDO)and improve the accuracy of the optimization results,aiming at the problem of low accuracy of reliability solution of structural performance function with higher non-linearity,based on the first order second moment reliability-based design optimization method,a fourth-order moment reliability-based design optimization method based on the Taylor series expansion is proposed.By transforming the reliability-based optimization problem into a single-loop deterministic optimization problem,the efficiency of the method is improved to a certain extent compared with the double-loop reliability-based optimization design method.Since the fourth-order moment reliability method has high calculation accuracy in the reliability solution,the reliability of the optimization result can be guaranteed.The proposed method are verified by a typical example and Monte Carlo simulation,which is compared with the reliability design optimization method based on the first order second moment.The results show that the reliability-based design optimization method has higher accuracy and optimization efficiency.3)To accurately evaluate the reliability of the machine tool spindle,the fourth-order moment method is used to analyze the reliability of various failure modes such as static strength failure,fatigue strength and torsional stiffness failure of the machine tool spindle.On this basis,considering the correlation between multiple failure modes of machine tool spindle,the reliability model related to multiple failure mode failure is established,and the correlation coefficient and reliability calculation method of multiple failure modes are given.The proposed method was verified by taking a certain machine tool spindle as an example using the Monte Carlo method.The results show that the fourth-order moment reliability method has higher precision and there is a certain correlation between multiple failure modes of the machine tool spindle.Therefore,considering the failure-related machine tool spindle reliability method has important guiding significance for the comprehensive reliability evaluation of the machine tool spindle.In addition,the method can be used for reliability assessments related to multiple failure modes of other mechanical parts4)The natural frequency of motorized spindle for machine tools has an important influence on its anti-resonance.In order to evaluate its natural frequency reliability,based on BP neural network and improved first-order and second-moment reliability method,a frequency reliability method of motorized spindle for machine tools based on BP neural network which is optimized by genetic algorithm is proposed.Firstly,the material parameters and dimension parameters of motorized spindle are regarded as design variables and the parameterized model of motorized spindle is established by ANSYS finite element software.ANSYS software is integrated by Isight software and sufficient samples are obtained by optimized Latin hypercube sampling method through Isight software.The functional relationship between the natural frequency of the motorized spindle and design variables is built by BP neural network method and the structural performance of motorized spindle is established.Finally,the improved first-order second-order method is used for reliability evaluation.The results show that the error of the fitting function after optimization by genetic algorithm is smaller,and the reliability value has a higher precision.On the other hand,in order to optimize the design parameters of motorized spindle and achieve the purpose of improving the natural frequency and weight reduction,Non-Dominated Sorting Genetic Algorithm(NSGA-II)is used to optimize the design of the machine tool spindle.Compared with before optimization,the natural frequency of the machine tool spindle is increased by15.8% and the quality is reduced by 17.53%. |
PDF Full Text Request