The Effect of Different Regularization Approaches on Damage Imaging via Electrical Impedance Tomography

• Published in: Journal of nondestructive evaluation Vol. 44; no. 2
• Main Authors: Tallman, Tyler N., Smyl, Danny, Homa, Laura, Wertz, John
• Format: Journal Article
• Language: English
• Published: New York Springer US 01.06.2025; Springer Nature B.V
• Subjects: Characterization and Evaluation of Materials; Classical Mechanics; Control; Damage detection; Damage localization; Dynamical Systems; Electrical impedance; Engineering; Image quality; Inverse problems; Ionizing radiation; Nondestructive testing; Regularization; Regularization methods; Solid Mechanics; Tomography; Vibration; Electrical impedance tomography; Embedded sensing; Nondestructive evaluation; Inverse problems; Regularization
• ISSN: 0195-9298; 1573-4862
• DOI: 10.1007/s10921-025-01178-w

Electrical impedance tomography (EIT) has attracted attention for its potential application to nondestructive evaluation of materials due to a lack of ionizing radiation, low cost, and the potential for integration with the material system. However, EIT is an ill-posed inverse problem and requires regularization to achieve a physically meaningful solution. Many materials-based practitioners of EIT make use of relatively simple regularization methods. This is important because the choice of the regularizer significantly impacts final image quality. Choosing poorly can misrepresent the damage state of the material to the inspector, ultimately undermining the potential of this modality. This manuscript thus serves as a primer for researchers using EIT for damage detection and localization by applying several common regularization techniques and one new technique to six experimental data sets. Notably, experimental data is taken from components of representative complexity and/or subject to non-trivial loading or realistic damage in an effort to transition the research away from the overly-simplified shapes and conditions that permeate the current state-of-the-art. It is shown that there is no one-size-fits-all regularization method; materials-based EIT practitioners must guide selection of the appropriate regularization method with knowledge of what they intend to image.