| As the essential equipment to modernize the manufacturing industry, NC machine tools arethe most important parts on the development of equipment manufacturing industry. Atpresent, since a big gap exists between the reliability levels of domestic NC machine toolsand the international advanced level, medium and high-grade NC machine tools and its keyfunction components depend mainly on import. The reliability level has become a key factorrestricting the development of domestic medium and high-grade NC machine tools.Therefore, it is extremely urgent to improve the reliability level of domestic NC machinetools.For these repairable products like NC machine tools, it is required that not only lessfaults, but also quick restoration after a fault. Therefore, both reliability and maintainabilityshould be considered in generalized reliability. The experience of domestic and overseasreliability work indicates that the design process determines the reliability level of theproducts. As the guiding and basic work in reliability design, reliability allocation andprediction should to be made full use of in each design stage. At present, some concretereliability design techniques have been applied to NC machine tools. However, due to theexisting data are incomplete and the existing methods are imperfect, most machine toolscompanies don’t start the work of reliability allocation and prediction, causing the reliabilityrequirements of various components to be indefinite. The dispersive work costs a great dealof manpower and material resources, yet the effect is not evident. For this case, this paperformulated the work flows in the design stage for NC machine tools reliability andmaintainability, which strung reliability modeling, failure analysis, formulating quantitativeindex, allocating and predicting the reliability and maintainability, and the allocation andprediction for reliability and maintainability are especially researched. These works arebased on the analysis of NC machine tools generalized reliability in the design stage, incombination with the requirement of the National Science and Technology MajorProject-“High Speed/Precision NC Machine Tools Reliability Design and Performance TestTechnology”. The main research contents consist of the following aspects:(1) According to the tasks in each design stages of NC machine tools, the reliability andmaintainability design work flow are formulated, and the considering factors of formulatingthe reliability and maintainability index are generalized. According the characteristics of reliability work, the whole machine are divided into several subsystems, and the structureand composition of each subsystem are clarified, thus laying foundation for carrying outthese works like failure analysis, reliability allocation and prediction.(2) The source of reliability data are analyzed, the data collecting method for NCmachine tools field test are formulated, and the field reliability test is carried out incombination with the method on similar products, obtaining a great deal of fault date andmaintenance data. Based on the collected fault data, this paper build the distribution modelof the time between failures, and test the models using graphic method and the analyticalmethod. The result shows that the time between failures of these products follows thetwo-paramter Weibull distribution with scale parameter563.21and shape parameter1.057. Through appling FMECA to these products, the criticality of each subsystem isobtained, where the feeding system has the greatest criticality, followed the automatic toolchanging system, electrical system and spindle system. The focus should be put on how todecrease the fault rate of these subsystems when we develop new products.(3) The reliability block diagram of the whole machine is built according to thesubsystem dividing situation. For the problem that there are many influencing factors inreliability allocating process and some of which are difficult to quantificationally analyze, areliability allocation method is proposed which combines the interval analysis, fuzzycomprehensive evaluation and analytic hierarchy process. According to the influencedegrees, each factor is assigned the corresponding weight, after comprehensively consideringthe factors of failure frequency, failure criticality, complexity, technology level and costeffectiveness. By comprehensive utilizing of field test information and expert experience,and replacing real number by fuzzy interval number to express uncertain information, thereliability allocating model of the NC machine tools is built. Then the allocation method isapplied to allocate reliability to the research object. The result shows that the method isfeasible and effective for reliability allocation of NC machine tools at present.(4) While using the conventional similarity comparison method to predict the reliabilitylevel of NC machine tools, it is difficult to judge the discrepancy degree of reliability levelbetween the research object and similar products. In view of this problem, a new reliabilityprediction method for NC machine tools is proposed which adopts the interval analytichierarchy process. Firstly, the relationships between the whole machine and its subsystemsare analysed, and the reliability prediction model for the whole machine is built. In theprocess of predicting the reliability of subsystems, the reliability data of similar products ismade full use. The reliability correction factor evaluation model is built, and the differences between the research object and similar products are analyzed comprehensively in theaspects such as design, manufacturing, using, management, product structure and so on. Thereliability correction factors are obtained using interval analytic hierarchy process, whichcould realize the complementation of assured information and fuzzy information. Incombination with four similar products, this method is applied to predict the reliability ofspindle subsystem, verifying the feasibility of this method.(5) Aiming at the problem that the weighted allocation method considering fault rateand design characteristics treats each design characteristic equally, and the designcharacteristics are hard to quantificationally analyze, the method is improved in this paper. After comprehensively considering the six design characteristics: fault detection method,reachability, replaceability, adjustability, complexity and criticality, the weighted factorcomprehensive evaluation model for maintainability is built, and the corresponding weightsfor each factor are assigned by interval analytic hierarchy process. The weighted factor ofeach subsystem is obtained by fuzzy comprehensive evaluation, overcoming the problems inconventional methods effectively. And the maintainability index of each subsystem isobtained by maintainability weighted factors in combination with allocated reliability valueof each subsystem. This method is applied to implement maintainability allocation for theresearch objects, verifying the feasibility of this method.(6) Aiming at the situation that the integrity degree of maintainability data of eachsubsystem differs greatly, this paper proposed a maintainability prediction method based ondata integrity degree for NC machine tools. This method differs with the conventionalmethods in that: the conventional analyze components and parts, yet this method analyzesfault modes. Comparatively speaking, the proposed method has enough data and a betteraccuracy. Aiming at these subsystems which integrity degree of maintainability data isdifferent, the functional level prediction method and the unit comparison prediction methodare adopted respectively to implement maintainability prediction. Finally, the maintainabilityprediction is implemented in combination with a case, verifying the feasibility of theproposed method. This method provides a new thinking for NC machine toolsmaintainability prediction, which could overcome the present problems that the maintenancedata is lack and the predicting accuracy is poor while using only one method. |
PDF Full Text Request