Handling Ambient Temperature Changes in Correlative Thermal Error Compensation

Abstract

Thermal errors are one of the leading causes for positioning inaccuracies in modern machine tools. These errors are caused by various internal and external heat sources and sinks, which shape the machine tool's temperature field and thus its deformation. Model based thermal error prediction and compensation is one way to reduce these inaccuracies. A new composite correlative model for the compensation of both internal and external thermal effects is presented. The composite model comprises a submodel for slow long-and medium-term ambient changes, one for short-term ambient changes and one for all internal thermal influences. A number of model assumptions are made to allow for this separation of thermal effects. The model was trained using a large number of FE simulations and validated online in a five-axis machine tool with measurements in a climate chamber. Despite the limitations, the compensation model achieved good predictions of the thermal error for both normal ambient conditions (21°C) and extreme ambient conditions (35°C).