Study Of Dynamic Performance Testing, Analysis Model And Process Planning Method In Cylindrical Grinding

Grinding has been widely used as precision machining technology in modern mechanical area, at the same time, it is the last process in precision machining. Different related elements in the grinding process system is complex, and a large number of complex nonlinear coupling relationship exists in different related elements. The stability of grinding process has been greatly effected by different elements, which will terminally the quality and efficiency of grinding. In some respects, the existed optimization method of grinding processing will improve the machining quality and efficiency. However, the existed optimization method is limited. On the one hand, it is difficult to cover different workpiece materials and processing requirements, on the other hand, the actual machining environment of individual grinding tool can not be actually reflected. According to the problem listed in the above, this paper developed some confirmed research meothod. The main work and results of this thesis are listed in below:(1) According to the study of the dynamic behavior caused by dynamic grinding force in grinding process, the dynamic force and the dynamic response effected by grinding contact length, maximum undeformed thickness has been researched. The time-domain dynamic analytic model(TDFMD)of grinding process has been proposed which focuses on the dynamic instability factors. This model is fixed with multi sensor detection data, and creats the result of grinding force and power dynamic response.(2) Consider the dynamic performance in grinding system, a process optimization method(PODP) has been proposed. Using multi-sensor detection and fusion method, the dynamic performance parameters in the grinding system can be obtaioned, which can fix the dynamic grinding force analytical model. The dynamic performace has been used as the input system to optimize the grinding parameters, and the grinding optimization results are closely to the actually grinding process.(3)According to the difficult problem of online detection in grinding process, an online multi sensor detection system of grinding process has been developed. The improved PSD method has been proposed to detect the amplitude information of spectrum, respectively, to find out the influence of the dynamic characteristics of the grinding wheel and the workpiece with grinding machining quality. Based on LMD decomposition and Lempel-Ziv complexity calculation method, the stability of grinding process was evaluated, and the influence of different process parameters on grinding stability was found. Fusion analysis has been used to analysis the dynamic information in grinding process, the stability diagram between the wheel speed vs. grinding depth and the stability diagram between the workpiece speed and grinding depth are described to define the stable lobe.(4) Based on the grinding quality and efficiency, the optimization process model of cylindrical grinding has beem established. The model proposed to merge dynamic characteristics feature of grinding process into the input variables. According to the description of cylindrical grinding in the optimization model, KPCA and improved ELM algorithm has been used to predict the grinding quality. The actual value of the roughness and error detection is less than 5%, respectively, the grinding process optimization prediction model considering the dynamic characteristic has been verified to have a greater accuracy than the traditional grinding process prediction optimization. According to the grinding process optimization model combined with improved Pareto multi-objective optimization method, the optimization results are ideal with grinding quality, efficiency, cost, time and other optimization objectives of process optimization.(5) With hard machining material ceramic used in aerospace and steel automobile spindle as engineering verification object, the grinding dynamic analytical model and multi-sensor detection method has been used to observe the regulation under different process parameters, respectly, the pareto multi-object optimization method has been used to define the grinding process. As the input parameter of actually grinding process, the effectiveness of the optimization method has been improved.