Profile tracking for multi-axis ultrasonic inspection of model-unknown free-form surfaces based on energy concentration

Abstract

This paper presents a novel profile tracking method to endow a multi-axis ultrasonic inspection system the ability of sampling model-unknown free-form surface without relying on external measuring devices. The proposed profile tracking is realized by adaptively implementing two steps: rough navigation by predicted trajectory and autonomous refinement through a process of concentrating the received ultrasonic energy. Different from existing unknown-surface sampling methods, both of surface points and surface normal are concerned in this paper to cater the unique requirements of multi-axis ultrasonic inspection. With the extended ability of profile tracking, the commonly used CAD model based path planning method can be generalized for multi-axis ultrasonic inspection of unknown free-form surfaces. Experiment results show that the profile tracking errors representing by maximal deviation of sampling points and associated normal from practical surface do not exceed 0.65 mm and 0.8°, which is suitable for multi-axis ultrasonic inspection of large model-unknown free-form surface.