| Aiming at the three key quality characteristics, that is, precision, Accuracy lifetimeand reliability for China numerical control machine tool, especially the problem thathigh-grade numerical machine tool severely rely upon the overseas ones caused by thelatter two characteristics which lag behind the ones of overseas machine tool, the articleregarded the horizontal machining center as object and aimed at improving the three keyquality characteristics: machine tool precision, precision lifetime and reliability. Basedon the study of pedigrees of the machining object, machining conditions, loads,functions and failures during the working process of numerical machine tool, theTop-Down principle was employed to conduct thePedigree-Functions-Movement-Action (PFMA) structural decomposition on the basis ofthe study route from complex to simple. The machine tool structure was decomposed tothe most basic meta-actions and the meta-action assembly unite was defined. Besides,aiming at the meta-action assembly unite, the assembly precision, precision lifetime andreliability of numerical machine tool were studied, diagnosed and estimated. Theconcrete study contents include:①The PFMA structural decomposition method of numerical machine tool wasstudied. In order to meet the requirements raise the quality of numerical machine toolincluding the complex structure, the machining object, various loads and large amountof failure modes, the pedigree (Object Pedigree, Working Condition Pedigree, LoadPedigree, Function Pedigree and Failure Pedigree) structure which demonstrated theworking conditions of machine tool. Taking the functions and motions of machine toolas core, the PFMA structural decomposition method was put forward. Thedecomposition route and principles were determined. Regarding Design StructureMatrix (DSM) and Domain Mapping Matrix (DMM) as decomposition calculationmodel, the basic meta-actions were obtained by decomposition. Based on meta-actions,the meta-action assembly unite as the basic unite studying assembly quality was defined.On the foundation of DSM and DMM model, the typical product THM6380MachiningCenter was structurally decomposed. DSM and DMM decomposition matrixes afterstructural decomposing were gained.92meta-action assembly unites such as wormdrive were finally obtained by cluster partition and39meta-action assembly unites wereformed. The structural decomposition method, which leaves a sound foundation of following chapters, is the base of different parts of studied contents.②The assembly precision of meta-action assembly unites were studied. Startingfrom the analysis of assembly error source of the meta-action assembly unites ofmachine tool and the error transfer mechanism, the mathematical model of the threekinds of errors: the part position error, the shape error and the assembly position errorwere built based on the homogeneous coordinate transformation. The error transfer linkgraph and the error transfer link network were contributed to demonstrate the errortransfer regularity aiming at the error transfer of meta-action assembly unites andassembly functions. In addition, on the basis of D-H joint model involved the terminalposition of robots, the assembly error of meta-action assembly unites was statisticallyanalyzed and calculated. The assembly precision prediction model and estimationmethod required by the assembly functions in a single direction and in multipledirections were built. The verification, which provided support and a guarantee for thefurther study on the assembly precision of the entire machine, was conducted to thebuilt model by taking the assembly of one meta-action assembly unite, the worm drive.③The precision lifetime of the meta-action assembly unite was analyzed. TheFMEA analyses of some meta-action assembly unites which significantly impacted theprecision lifetime of machine tool, such as the feed system in the machine tool, ballscrew pairs and rolling linear guide pairs, etc. were conducted. The main factors leadedto precision degeneration of the system during the assembly process were found out andthe precision lifetime equation of the feed system was established. For a further researchon the variation regularity of the feed system locating precision failure which variedfollowing the accumulated running-in strokemm, the data collected from the test site,where the working conditions of the assembled servo feed system were simulated, wastaken for instance. The two-parameter Weibull distribution model was adopted toanalyze the precision reliability of the feed system. By measuring the feed systemlocating precision and discretization, the regularity that the locating precisiondistribution within the strokemm axis matched Weibull distribution was proved. Fromthis, the feed system locating precision lifetime prediction model was established basedon non-homogeneous Poisson Process of AMSAA power law model. The feed systemlocating precision prediction equation was deduced. Finally concrete measures werecarried out to improve the precision lifetime of the feed system during the assemblyprocess so as to lay a theoretical foundation for further improving the precision lifetimeof machine tool. ④The assembly reliability of meta-action assembly unite was studied. Toguarantee the meta-action assembly unite reliability of a machine tool, the meta-actionmodularization failure tree model was set up in combination withFunctions-Movement-Action mapping relation on the basis of FTA of meta-actions. Bymeans of ITE structural analysis method, the transformation from the modularizationfailure trees to binary decision diagrams and optimization were fulfilled. BooleanStructure Function of Meta-Action Particle Size was built. The tetra F=(S, P, T, Q)which described the multi-dimension dynamic mapping relation between meta-actionsand the performance properties of meta-action assembly unites was constructed.Moreover, reliability-driven assembly process flows were formulated. Taking THM6380exchange tray carrier for example, the article calculated the assembly reliability of theexchange tray carrier and demonstrated the feasibility and validity of the modelingmethod. For key assembly processes of various meta-action assembly unites in theexchange tray carrier, some specific reliability control measures were employed andreliability-driven assembly processes were formulated in order to lay the base of furtherimproving the assembly reliability of a whole machine.⑤The whole machine’s assembly quality was diagnosed and estimated. Focusingon the fuzzy indeterminacy between the assembly quality abnormity of numericalmachine tool and the causes, the equation described the relation between the assemblyquality abnormity and abnormity causes was built based on Fuzzy Relation Matrix andmembership in an abnormal mode. Improved neutral network (BP) algorithm wasadopted to transform the solution of the fuzzy relation equation to the neutral networkproblem. The neutral network algorithm solving the fuzzy relation equation’sapproximate solutions was designed and relevant parameters of the algorithm weredetermined. Specific diagnosis scheme for the assembly quality abnormity solved by BPbased on fuzzy relation equation was set up. The detailed diagnosis of the causesleading to the assembly quality abnormity of numerical machine tool was carried out.To further estimate the assembly quality of machine tool, Long-linear ProportionalIntensity Model (LPIM) was utilized to estimate the reliabilities of4machine toolwhich were assembled with reliability-driven assembly processes. Long-linearProportional Intensity Function of the failure time was established. The point estimationand interval estimation of model parameters were given by Maximum LikelihoodMethod and Fisher Information Matrix Method. The point estimation of machine toolreliability index and the interval estimation based on Delta Method were both given. At the same time, the time trend and reparation efficacy significance were inspected byLikelihood Ratio Test. The results showed that the failure intensities of the4numericalmachine tool were entirely declining and their reliability obviously improved. |
PDF Full Text Request