| Résumé | Due to their greater workspace and lower cost, industrial robots can provide an alternative solution to CNC machine tools for milling of aerospace parts. However, current machining robots fail to meet the required finishing tolerances (0.1 mm) of the aerospace industry. To improve the accuracy of industrial robots during milling, this paper proposes a new feedback method using an off-the-shelf passive CMM-arm. Unlike laser trackers and optical CMMs, CMM-arms are not susceptible to occlusion or line of sight issues, which may arise due to machining debris or robot movements. In this work, the performance and robustness of a CMM-arm used to guide a robotized machining process in optically challenging conditions are studied. First, the kinematic model of the CMM-arm is developed and the shared workspace between the CMM-arm and a KUKA KR300 Ultra industrial robot is analyzed. This ensures adequate reach of the robot’s tool center point (TCP) while the CMM-arm is attached to the machining spindle. The CMM-estimated tool center point is then calculated in real time on an external PC based on the combined kinematic chain of the robot and the CMM-arm. The positioning error of the TCP is determined in real time at a rate of 80 Hz, and the corresponding correction signals are sent to the robot via the Robot Sensor Interface (RSI). Experimental results show over 80% reduction in maximum error during 5-axis milling of aluminum. |
|---|
