Reliability Research For Boring Spindle Feed Mechanism Of Heavy-duty CNC Milling And Boring Machine

Heavy-duty CNC machine tools are one type of mainstream basic equipment of modern machinery manufacturing industry. Reliability of the core drive system of heavy-duty CNC machine tools is one of critical barriers to their entry into the world’s leading. Heavy-duty CNC milling and boring machine is one of the main numbers of the heavy-duty CNC machine tools. Reliability analysis of boring spindle feed mechanism, a core drive subsystem, is undoubtedly of great practical significance. In practical engineering, the performance of mechanical parts tends to degrade gradually until complete failure, thereby affecting the overall system performance. Therefore the system and mechanical parts generally present characteristics of multi-state. However, traditional fault tree analysis which is based on binary assumption cannot fully characterize the multi-state of heavy-duty CNC machine tools. Therefore, the multi-state fault tree analysis is adopted in this article to facilitate the reliability analysis of the boring spindle feed mechanism. Specific investigations include the following several aspects.Firstly, we analyze the basic function, structure and features of work environment of boring spindle feed mechanism. The states of the system and its components are defined. The multi-state fault tree is generated by utilizing multi-state reliability block diagram. This method can avoid the impractical assumption that multi-state system needs to be decomposed into several binary systems. Adopting this method, we then build the multi-state reliability block diagrams and the multi-state fault trees of boring spindle feed mechanism from each subsystem to the whole system layer by layer.Secondly, the multi-state fault tree of boring spindle feed mechanism constructed above is analyzed quantitatively. The occurrence probability of each state are calculated for top events of each subsystem and whole system layer by layer, where the methods of multi-valued decision diagram and Bayesian network are used. The pros and cons of multi-valued decision diagram and Bayesian network for multi-state fault tree analysis are compared as well. Bayesian network is then proposed to calculate the probability importance and critical importance of bottom events of the whole system and its subsystems. By analyzing the results of importance, weaknesses in the system are identified and suggestions are provided for improving the system.Finally, taking into account the impact of common cause failure and the uncertainty of system, we incorporate common cause failure and fuzzy probability into Bayesian network to analyze the multi-state fault tree. Adopting the technique of Bayesian network based common cause failure modeling, the common cause failure-Bayesian network modeling of boring spindle feed mechanism is constructed. However, this approach may lead to the occurrence of common parent nodes in Bayesian network, i.e., repeated events in fault tree. The fuzzy set theory is adopted further into the quantitative analysis of this model. In detail, triangular fuzzy number is used to describe the occurrence probability of each state of the root nodes. However it is difficult for discrete-time Bayesian network to calculate the occurrence probability of top event in the fault tree having repeated events based on fuzzy probability. To solve this problem, the methods of defuzzification and probability normalization are improved to handle the calculation of fuzzy probabilities. The bi-direction reasoning of common cause failure-Bayesian network modeling is then implemented for the boring spindle feed mechanism. Based on the analysis results, the impact of common cause failure on system reliability is highlighted and suggestion for system improvements is then provided.