Feed rate optimisation scheme in robotic machining operations for dynamic error compensation

Abstract

The manufacturing sector demands shift towards parts with more complex geometries with the need for flexibility in production, driving interest to robotic machining. This advancing technology brings advantages like affordability, versatility, and ease of implementation, making it well-suited for agile production environments. Nevertheless, robotic machining struggles with accuracy issues due to the inherent flexibility of robots, which results in deviations and vibrations. The positioning error along a robotic machining trajectory is composed of two contributions: the steady state error and the transient. Initially generated from CAM software, the trajectory is considered as a path with a speed profile. It is then discretised in elementary sections, modelled with Hermite splines and connected by nodes. To address, offline, the lack of accuracy, an updated trajectory is computed by iteratively replacing these nodes space based on the error estimated from the dynamics simulation, strongly reducing the steady state error. However, in transient sections, the error reduction is not sufficient.

This research focusses on the impact of the feedrate modification in transient areas, typically the entrance and exit of tool in the workpiece. Specific speed profiles are defined for these sections by applying linear segments with parabolic blends expressed in terms of the curvilinear abscissa. An optimisation scheme is proposed to update their feedrate considering the node repositioning necessary to compensate tool-tip deviation. The investigation of the feedrate update is based on the results from virtual machining simulator including the robot dynamical model, responsible for steady-state and transient errors respectively and a cutter-workpiece engagement module.