Analysis of trajectory deviation during high speed robotic trimming of carbon-fiber reinforced polymers

Although carbon-fiber reinforced polymers (CFRPs) are used extensively in the aerospace industry, trimming of CFRPs in high speed robotic end milling has however not yet received its due attention within the research community. For such an application, the robot should be very stiff for the machining operation to be generated. If the robot is not sufficiently stiff, deviations in shape and position of the workpieces will occur. In this paper, a methodology has been developed and implemented for quantifying the relative contributions of cutting parameters (feed, speed) and robot configuration to workpiece accuracy during robotic trimming of CFRP laminates. Trajectory deviation and part accuracy are evaluated with respect to robot configurations and cutting conditions using data from two sources. The first set of data is obtained through the laser interferometer measurement system, while the second set of data is obtained from measurements carried out on the trimmed specimens using the coordinate measuring machine (CMM). Results from the laser interferometer system show that the observed errors are consistent with gear transmission errors. This is revealed by periodic shapes along the measured path. It was observed also that joint 1 has an important influence on the robot accuracy. The less the joint 1 displaced, the better the performance of the robot. Machining results show that a lack of stiffness in the robot causes large deviations in shape and position and, further, that these deviations are affected by the cutting force and the machining parameters. Results show also that the robot configuration which is optimally suited to perform the task is reached by using a relatively folded configuration and a minimal displacement of the joint 1. Furthermore, the optimum cutting conditions leading to a minimum trajectory deviation are achieved at a low feed rate (0.2 mm/rev) and a moderate cutting speed (400 m/min). (C) 2014 Elsevier Ltd. All rights reserved.