Numerical and experimental modeling of thermal errors in a five-axis CNC machining center

Abstract

This work aims at the establishment of methodology to model and analyze the thermal errors of a five-axis CNC machining center, from an estimated temperature field, to finally model an artificial neural network (ANN) algorithm to accurately predict with robustness the thermal error. The thermoelastic behavior of the machining center was modeled through two different approaches: experimental (or data-driven) model and numerical (or physical) model. The thermal behavior of the machine was first modeled using finite element method (FEM) techniques based on theory of friction heat and convection heat and validated with the various experimentally raised temperature fields using temperature sensors and thermal imaging. The main machine subsystems were initially validated, such as ball screw system, linear guides, and spindle, which allowed for validating of the thermal behavior of the entire machine for five different duty cycles obtaining a maximum error of less than 8% when comparing the numerical results with the experimental results. The components of the thermal errors in X, Y, and Z directions were obtained through FEM by measuring the displacement of the spindle tip in relation to the reference bushing located on the worktable. The same procedure was experimentally performed using a touch probe system clamped in the spindle, and the results were compared obtaining a maximum deviation of 17 μm. The validation of the finite element model allowed for the use of the results obtained by the simulation to train and validate an ANN for predicting the thermal errors of the machining center. The relative errors between the thermal errors predicted by the ANN and the FEM simulation results were less than 1% indicating that the methodology developed in this work that combines the use of physical models with data-driven models is an accurate and robust tool to predict the thermal errors of the machine for various working conditions, even with the machine moving at different speeds or alternating the movement of the axles. © 2018, Springer-Verlag London Ltd., part of Springer Nature.