Machine tool condition monitoring using statistical quality control charts

The application of statistical quality control charts to the monitoring of machine tool performance parameters is presented. A novel method of measurement normalization is included to compensate for performance parameter inter-dependence. The method is based on a ''normalization chart'' which defines the normal relationships between the performance parameters. The difference between a ''current'' measurement and the corresponding value from a normalization chart is then used to calculate a residual. Using the principles of statistical quality control, the residual is then plotted on a control chart to monitor the condition of the machine tool. The method is illustrated with some of the machine tool parameters logged by a data acquisition system, designed and implemented as part of a European Union research project. (C) 1997 Elsevier Science Ltd. All rights reserved.