Research On Load Testing Technology And Load Spectrum Compilation Of CNC Machining Centers

With the constant improvement of the level of mechanical industry of our country,machining centers is more and more widely used in automotive, aerospace, precisionmold sand other areas of the industry. However, compared to machining centersimported from industrial developed countries such as Germany and America,domestic made machining centers have lower processing accuracy stability, MTBF(Mean Time Between Failure) and reliability because of the lack of scientific andreliable load spectrum in the design stage, leading to the short of fatigue strengthdesign of the key features of the domestic made machining centers. In addition, loadspectrum is the important basis of the reliability test specification for the currentmachining centers and all kinds of key features in the early days of the factoryconduct troubleshooting test, empty running test or a simple load test because of alack of load spectrum information. Therefore, research on load testing technology andload spectrum compilation of CNC machining centers is of great significance to thepromotion of reliability of the machining centers and key features of them.According to the national major projects of science and technology research of"High-grade CNC Machine Tools and Basic Manufacturing Equipment",takingdomestic processing center as the research object, this paper has designed field andlaboratory testing technology and a large amount of load data has been collectedbased on the technology mentioned above. Then, this paper has carried out researchon formation of velocity spectrum, cutting force spectrum and cutting torque spectrum,fuzzy comprehensive decision analysis method was used for optimizing thedistribution model best fitted the collected load data. As such, the main content of thispaper is as follows:1.To get the accurate machining centers load data, this paper has formulated fieldand laboratory load spectrum testing technology. Field testing technology was usedfor data acquisition on site. For load data unable to collected on site, this paper hasdesigned laboratory test techniques.2.After the acquisition of data preprocessing, the velocity spectrum, maincutting force spectrum, backward force spectrum and the feed force and torquespectrum of the focused machining centers have been adopted. According topreliminary analysis of the statistics of load spectrum data, commonly used statistical distribution models such as Gaussian Distribution, Fourier series, Beta Distribution,Lognormal Distribution, Gamma Distribution and Weibull Distribution have beenpreliminarily chosen to fit the collected load data. Multiple linear regression methodand simulated annealing optimization method have been engaged to estimate theunknown parameters in each chosen distribution model. Finally, Cole AlamogordoRove testing method was used to carry on the test of goodness of fit, so thedistribution model of load spectrum has been acquired. This paper adopted fuzzycomprehensive decision analysis method to choose the distribution models whichhave pass the primary test at the same time so that the statistical distribution modelbest fitted the collected data was gained.3.On basis of the gained distribution models, torque load testing scheme appliedto the main spindle has been formulated.This load spectrum compiling method of machining center in this study hasimportant reference value. The load spectrums obtained in this paper will provideimportant basis for fatigue strength design and reliability test of domestic mademachining centers.