| About this Abstract |
| Meeting |
2024 ASC Technical Conference, US-Japan Joint Symposium, D30 Meeting
|
| Symposium
|
2024 ASC Technical Conference, US-Japan Joint Symposium, D30 Meeting
|
| Presentation Title |
Localization of Impact Damage in Composite Material Subjected to Mechanical Loading Using Electrical Impedance Tomography |
| Author(s) |
Laura Homa, Tyler Tallman, Tyler Lesthaeghe, Norman Schehl, John Wertz |
| On-Site Speaker (Planned) |
Laura Homa |
| Abstract Scope |
Electrical impedance tomography (EIT) is a non-invasive imaging modality that is used to determine the conductivity within a domain based on voltage measurements taken on the boundary. When paired with composite materials that demonstrate stimulant-responsive conductivity, EIT can be used to map damage in terms of conductivity change. Here, we will demonstrate the ability of EIT to detect surrogate impact damage within a polymer-matrix composite (PMC) panel in the presence of mechanical loading. We consider EIT measurements from an eight-ply panel loaded in four point bend after indentation to induce delamination. EIT measurements of the non-indented panel before loading are used as a baseline. We then apply a novel inversion technique to perform difference imaging. It is well known that EIT is an ill-posed inverse problem which requires regularization to solve. Common practice is to use the Laplacian as a regularization operator to enforce smoothness in the solution. However, the Laplacian operator is insufficient on its own to detect impact damage in the presence of strain. Here, we solve the inverse problem in the Bayesian framework and propose a mixed prior that combines a smoothness prior with a conditionally Gaussian prior that favors sparse solutions. This prior is tailored to impact damage and is thus able to successfully image impact damage in the bent panel. |
| Proceedings Inclusion? |
Definite: Post-meeting proceedings |