Study On Reliability And Maintainability Analysis Of CNC Machine Tools Based On Grey System Theory

Evaluation and prediction of reliability and maintainability are the basis of CNCmachine tool reliability research, and which are also the premise of reliability design,reliability growth and maintenance. CNC machine tools is a complex system of mechanical,electrical and hydraulic hybrid. The complexity of its structure led to more intricate failuremechanism, uncertain failure microscopic appearance, and more clear macroscopiccharacterization. So the part of the fault information of CNC machine tools and itssubsystem is known. Such kind of systems are called Grey Systems. In this paper, the CNCmachine tools and its subsystems are regarded as a grey system, and its reliability andmaintainability should be studied.Due to the limitations of the existing reliability index assessment methods, this paperproposes to use function increase and decrease to assess reliability and maintainability index.In view of a single indicator can not fully measure the overall level of reliability of CNCmachine tools, this paper selects multiple evaluation indicators, and the grey whiteningweight function theory of reliability and maintainability of the implementation ofcomprehensive clustering evaluation. Considering the importance of prediction of reliabilityand maintainability to the reliability improvement, the routine repair and maintenance ofCNC machine tools, this paper presents the reliability prediction model established by usinggrey theory and the maintainability prediction established by using the exponentialsmoothing. Meanwhile, in order to make up for the deficiency of ignoring the failuresequence of empirical modeling methods, the two-dimensional reliability model of CNCmachine tools model is raised to get the changing tendency of the fault. Finally, a set ofmethodologies for reliability and maintenance assessment and prediction of CNC machinetools are proposed. The specific work of this paper is illustrated as follows:1. Reliability evaluation of CNC machine tools(1)Combined with previous experience modeling method, this paper established areliability model based on domestic and foreign development status of CNC machine toolsand research status of CNC machine tools reliability. In order to solve the problem that theevaluating results do not tally with the practical engineering under the exponentialdistribution, this paper proposes the adopted interval estimation method of MTBF obeysWeibull distribution, which lays the foundation for the comprehensive assessment of thereliability.(2) Due to the single index is not able to measure the reliability of the whole system,the method of multiple reliability index and combined with whitening weight function of comprehensive clustering evaluation is raised. This paper uses previous interval estimationresults to partition clustering which can avoid the shortcomings of expert scoring and makethe evaluation results more consistent with practical.2. Maintainability evaluation of CNC machine tools(1) With the help of the existing maintenance data and function of maintainability, weobtain the point estimation and interval estimation of maintainability index, get thepreliminary evaluation of maintainability of CNC machine tools.(2)In this paper we adopt whitening weight function to realize comprehensiveclustering evaluation of maintainability and use interval estimation results onmaintainability index to partition clustering interval. The method makes comprehensiveclustering evaluation results more truly which can reflect the maintainability of CNCmachine tools.(3) Considering the influence of the subsystem maintainability on the whole machinemaintainability, the method of comprehensive evaluation of subsystem maintainability byusing the information entropy theory and the whitening weight function is proposed.3. Subsystem risk assessment of CNC machine tools(1) In view of the micro relations of CNC machine tool subsystems without explicitcharacterization, grey relational theory in grey theory is proposed to make correlationanalysis of reliability and maintainability. According to the results of the experiments, thismethod can effectively achieve the “whitening” of “grey” between sub systems.(2) Based on the influence of different subsystems on the reliability of CNC machinetools, this paper applies Bayes theory to establish the posteriori probability model ofreliability and maintainability, and uses the model to determine the key subsystem. The casestudies show that: the establishment of the posterior probability model can quickly andaccurately find out the key subsystems of CNC machine tools. The critical subsystems ofCNC lathe are the spindle system, feeding system, numerical control systems and turret.(3) This paper presents the use of primary and secondary diagram and FMECAanalysis for the risk assessment of critical subsystems’ reliability, finds out the main failurereason, and proposes appropriate maintenance measures of the failures according to theanalysis results. The analysis shows that: the main reason of the failure of criticalsubsystems is component damage, the primary failure mode are component damage andgeometric precision tolerance.4. Prediction of reliability and maintainability(1) This paper establishes the reliability models of machine tools and subsystems which can predict the changing tendency of the fault in the assessment periods. The resultsshow that: the fault intervals of CNC lathes accord with the two-parameter Weibulldistribution, and the maintenance time accords with the logarithmic normal distribution.(2) Grey theory is applied to establish grey fault prediction model. Based on thecharacteristics of maintenance time, the exponential smoothing method is adopted toestablish the prediction model of maintainability. The results indicate that: the predictionmodel can accurately predict subsequent fault time and maintenance time.(3) In the light of existing empirical modeling methods ignore the fault sequence, thispaper puts forward using subsequent forecast fault time to establish the two-dimensionalreliability model of CNC machine tools to get the changing tendency of the fault; the modeladopts the cumulative fault time and fault interval as variables.