Development Of High Precision Machine Tool Spindle Bearing Fault Prediction System

The bearing quality directly affects the performance of the workpiece in the field of mechanical processing. However, bearing application environment is often complex. Therefore, the research on the fault prediction technology of rolling bearings has been an important topic in machinery fault diagnosis fields.The high precision machine tool spindle bearing fault directly leads to the decrease of the machining accuracy. Under the existing state, it can only stop to purchase and replace bearings after bearing fault. This leads to decreased equipment utilization and disorganized production. This paper describes the current development status of bearing fault prediction system and research results of scholars at home and abroad. There are many shortcomings on the research of bearing fault prediction. On the basis of getting plenty of machined workpiece and bearing replacement date data, the paper proposes to develop a new bearing fault prediction system, which predicts bearing fault occurred from the perspective of the workpiece data analysis. This proposes a new research direction for bearing fault prediction.The paper analyzed the collected data of machined workpiece based on the Statistical Process Control(SPC). First, it elaborated the development process of statistical process control and introduced the main methods of SPC used in data processing. Then through application examples of Q-DAS software illustrated that SPC was widely used in manufacturing enterprises. Secondly, it made tolerance analysis on the collected data of the workpiece. Then it made data analysis with the application of the SPC theory methods, such as parameter calculation, distribution model, analysis chart and process capability analysis, etc. after the analysis object is determined. And it made comparative analysis and fit analysis on the multiple parameters before and after the bearing replacement node. Eventually it determined the mean, standard deviation, process capability index(Cpk) three parameters as the judgment of bearing replacement. At last, it made a research on the distribution model of collected data. And it made normality test on the data distribution. Eliminate individual outliers from the group through outlier test. The sample was divided into benchmark sample and warning sample by using the methods of sample extraction and separation. Calculate the warning threshold of three parameters using extraction and separation of the sample. And make experimental analysis verification and correction on the warning thresholds.The paper designed a bearing fault prediction system with the connection between Lab VIEW software development platform and SQL Server database. It not only plays the strengths of LabVIEW visualization on the virtual panel user interface but also takes advantage of SQL Server powerful data management capabilities. Realize the functions of large-scale data computation, storage, display, warning and query. The functions of the program which is designed using LabVIEW are broadly divided into three parts: Data computation and storage module, Data display and alarm module and Data query module. Establish the whole event structure through the branch of the three module and each branch realize their functions independently of each other. Detect workpiece data by offline statistics and enter them into system program. Set three warning parameters to “or” gate relations. One of them exceeds the threshold will trigger the indicator. Automatically prompt the replacement of bearings before the machine bearings fault. Achieve the purpose of bearing fault prediction.