A non-contact calibration method of workpiece coordinate system for aero-engine blade-robotic grinding system

Abstract

Mounting error of the blade is an important factor that affects the grinding quality in aero-engine blade-robotic grinding system. In order to compensate for the mounting error, it is necessary to calibrate the pose of the workpiece coordinate system. However, most of the current workpiece calibration methods cannot meet the high accuracy requirements in the scenario of aero-engine blade grinding. Therefore, this paper proposes a new non-contact calibration method of workpiece coordinate system using scanner measurement, to solve the workpiece calibration problem in aero-engine blade-robotic grinding system. The method first calibrates the scanner based on the criterion sphere hand-eye calibration method and employs the Random Sample Consensus (RANSAC) algorithm to ensure the accuracy of the fitted sphere center coordinates. Then, the point cloud registration algorithms are used to align the scanned point cloud of the blade with the model point cloud. Finally, the space transformation matrix between the workpiece coordinate system and the base coordinate system is obtained through matrix calculation, achieving the accurate calibration of the workpiece coordinate system. The experimental validation of the proposed calibration method in this paper can maintain tiny and acceptable errors for grinding applications. Specifically, the hand-eye calibration error is less than 0.2 mm, and the point cloud registration error is less than 0.06 mm. Moreover, the calibrated blade exhibits excellent machining consistency after grinding, which can perfectly meet the requirements of actual machining accuracy.