Automation of three-dimensional inspection using the iterative closest point algorithm: application to a gas turbine blade

Abstract

This paper examines an innovative approach to the automated inspection of turbine blades in power generation systems. By integrating point cloud generated from computer-aided design (CAD) with iterative methods, this methodology aims to improve the efficiency and accuracy of the inspection process. The combination of theoretical blade geometry data, captured via CAD point clouds, with advanced algorithms derived from iterative methods, allows for a rapid assessment of the blade’s condition. This approach has the potential to reduce inspection times, increase assessment accuracy, and detect anomalies at an early stage. The paper explores in detail the key aspects of this method, including the creation and alignment of CAD point cloud with real blades, the application of iterative methods to detect degradation and anomalies, and the preliminary results obtained from specific case studies. However, challenges remain, such as the quality of input data and the need to develop iterative models specific to each application. Future prospects include the refinement of data capture techniques, the exploration of new iterative methods, and the integration of machine learning for even more advanced automation. Overall, this approach represents a significant advance in the field of industrial maintenance, enabling proactive and efficient management of turbine blade inspection and maintenance. It offers advantages in terms of cost, speed, and reliability for maintaining the sustainability and performance of power generation systems.