Robotic path compensation training method for optimizing face milling operations based on non-contact CMM techniques

I. Iglesias,A. Sanchez,Francisco J.G. Silva

DOI: https://doi.org/10.1016/j.rcim.2023.102623

IF: 10.103

2024-02-01

Robotics and Computer-Integrated Manufacturing

Abstract:Currently, the use of industrial robots in the machining of large components in metallic materials of significant hardness is proliferating. The low rigidity of industrial robots is still the main conditioning for their use in machining applications, where the forces developed in the process cause significant deviations on the cutting tool path. Although there are already methodologies that facilitate the pose study of the robot mechanical behaviour, predicting deviation values of the cutting tool path and facilitating the selection of process variables, robotic cell users still request new methods able to allow them to optimize the use of these production systems. On the other hand, non-contact measurement technologies have burst into many fields of knowledge, their use is becoming consolidated, and they allow the digitization of complex surfaces. This research presents the development of a new method of robotic machining trajectory compensation that allows optimizing the manufacture of flat surfaces using an industrial anthropomorphic robot. The new training method determines the actual deviations of the cutting tool after the machining process, and checks if these are within the admissible range of flatness error. This method is a novel iterative technique that incorporates the algorithm that uses the measured deviations and a reduction factor fr to calculate the offset that modifies the coordinate value of the programmed path points outside the admissible range and generates a new machining path to be tested. The method has been tested on a pre-industrial scale for aluminium machining, and the algorithm has carried out two iterations to generate a compensated robotic milling path within a flatness tolerance range of 300 µm, improving the error deviation by 37% comparing to the initial path.

robotics,computer science, interdisciplinary applications,engineering, manufacturing