Performance monitoring of a machining centre

This paper is concerned with the generation and validation of a thermal deformation model that is to be used as one aspect of monitoring the performance of a machining centre. Due to the need for high quality and high accuracy products, the effects of thermal deformation on a machine tool need to be minimised. There has been much research on models for predicting and correcting thermal deformation but not on their use of evaluating the condition of the machine. If a model contains information about any "normal" characteristics of a machine then it can be used to detect "abnormal" characteristics. This approach requires both validation of the model and of its inputs. Validation of the model is required to ensure that it does truly represent the "normal" characteristics of the machine and validation of the model's inputs Is required to ensure that they accurately reflect the state of the machine. The models considered in this paper were empirically derived using data from temperature and displacement sensors positioned on a vertical machining centre at Southampton Institute. The thermal deformation was produced by using the machining centre under normal operating conditions but without cutting. Comparison between two different modelling techniques enabled the most "suitable" model to be chosen. These models were validated by assessing their prediction success on new, data. Validation of each model's inputs was achieved through the use of an expert system shell and rules about the correct behaviour of each of the temperature sensors derived from information gathered during the tests.