This content is not included in your SAE MOBILUS subscription, or you are not logged in.

Tool Wear Classification in Automated Drilling Operations of Aircraft Structure Components using Artificial Intelligence Methods

Journal Article
2022-01-0040
ISSN: 2641-9645, e-ISSN: 2641-9645
Published March 08, 2022 by SAE International in United States
Tool Wear Classification in Automated Drilling Operations of Aircraft Structure Components using Artificial Intelligence Methods
Sector:
Citation: Koch, J., Schoepflin, D., Venkatanarasimhan, A., and Schüppstuhl, T., "Tool Wear Classification in Automated Drilling Operations of Aircraft Structure Components using Artificial Intelligence Methods," SAE Int. J. Adv. & Curr. Prac. in Mobility 4(4):1072-1081, 2022, https://doi.org/10.4271/2022-01-0040.
Language: English

Abstract:

Structural components in fuselage barrels are joined with the help of riveting processes. Concerning the key feature of rivet drill hole size and drilling quality, a poorly executed drilling operation can lead to serious riveting defects such as rivet play or fracture due to non-uniform load distribution. Consequently, the drilling process of a rivet hole and its correct execution is of vast importance for the airworthiness of an aircraft. The condition of the drill used, i.e., the current tool wear, has a direct effect on the quality of the hole. Since conventional approaches, such as changing the tool after a predefined number of process cycles, do not reflect real tool wear, premature wear may occur, resulting in defects. Thus, the online-detection of tool wear for necessitated replacement may indicate a promising future direction in quality control. Since the aircraft industry has a particularly high requirement for defect-free production of structural components, this paper presents a study on the online-detection of tool wear in automated drilling processes using a combination of external sensor technology and Artificial Intelligence methods. For this reason, a laboratory setup to conduct automatic drilling operations in fuselage material is introduced. Two sensor types are utilized to capture the process data that is evaluated by machine learning algorithms. The performance of different machine learning algorithms is measured, and recommendations for action in sensor solutions, and the respective choice of algorithms for this task, are derived. Finally, the results of the study are discussed, and recourse for future work is elaborated upon.