April 19 @ 4:00 pm - 5:00 pm
Plastic Deformation Reconstruction Based on Acoustic Emission
PhD candidate, Department of Mechanical Engineering
Johns Hopkins University
Interpreting acoustic emission (AE) is a non-destructive evaluation (NDE) technique based on the elastic waves caused by irreversible deformations in materials. The core challenge is decoding AE signals and quantitatively determining the source, or deformation mechanisms. We adopt variational and machine-learning approaches to decode the complex interconnections between plastic deformation mechanisms in metals and the AE they emit. Surrogate measurements of AE associated with different dislocation mechanism were first obtained from full 3D elastodynamic-field solution incorporated into 3D Discrete Dislocation Dynamic simulations. Data assimilation procedures use the AE signals at a few sensors only, and reconstruct the plastic deformations. We show that the dislocation activities history in 3D can be accurately reconstructed using a few AE sensors. A reconstruction accuracy of 90% was reached and can further be improved with increased number of sensors and/or high acquisition rate. Acoustic emission measurements were also augmented by physics informed neural networks to improve the reconstruction accuracy with lower number of sensors. Generative adversarial networks are also being considered for reconstructions from noisy signals. This approach allows for the decoding of hard-to-interpret surface AE measurements and reconstruct plastic slip intrinsically in the material during deformation.
Junjie Yang is a third year PhD student working with Prof. Jaafar El-Awady and Prof. Tamer Zaki in the Department of Mechanical Engineering at the Johns Hopkins University. His research focuses on dislocation plasticity of metals and alloys using computational methods and machine learning techniques.