Performance and efficiency of co-simulation for milling operations in robotic machining

Abstract

In robotic machining, tool/workpiece interaction generates forces that impact the performance of the robotic arm. This leads to defections due to its lower stiffness compared to dedicated machining machines, particularly at its joints. These deflections can degrade the quality of the machining operation, making trajectory and orientation optimization essential. Accurate optimization requires knowledge of the transient force field for the operation and a model of the robot. Currently, the simulations are performed using two in-house simulation frameworks : EasyDyn for the multibody dynamics of robot modelling and Dystamill for the modelling of the cutting forces. To enhance the simulation’s modularity, co-simulation between the multibody system (MBS) of the robot and the workpiece (with calculation of cutting forces for the latter) is proposed, allowing for better integration of the coupled dynamics. The decision on where to partition the model is critical. Given that the workpiece transmits force elements while the robot transmits position data, a displacement-force coupling approach is proposed. Co-simulation offers several advantages, including the ability to use dedicated software tailored to different parts of the model and facilitating the interfacing of various robots with different machining models. However, the modularity offered by co-simulation techniques is counterbalanced by less accurate results with respect to those obtained with monolithic models. This research aims to evaluate the influence of the accuracy of the results on a simple model.