EUVIMEDObjevte EUVIMED
Crossref

Physics-Informed Filters for Reconstructing Missing Data in Ultrasonic Guided Wavefields

Michael MacIsaac, Amanda Beck, Woohyun Eum, Charlie Tran, Joel B. Harley, Ghatu Subhash

e-Journal of Nondestructive Testing · 2026-07 · 31 (7)

DOI: 10.58286/33505 ↗

At a glance

Journal article in e-Journal of Nondestructive Testing, 2026-07. Written by Michael MacIsaac and 5 co-authors.

AbstraktBibliografické údajeCiteRelated studies

Abstract (original language)

Ultrasonic guided wavefields are valuable nondestructive evaluation datasets for revealing material or structural heterogeneities. Traditionally, wavefields are sampled on a dense spatial grid that covers the entire structure. Approaches exist to reduce the grid density but usually maintain a near-grid-like structure. This paper explores a strategy based on optimal physics-informed filters for removing this constraint and reconstructing wavefields with large spatial data gaps. These filters incorporate the wave equation or Helmholtz equation into the reconstruction. The filters ensure the wavefields satisfy this physics and learn the associated physics parameters from the data under test. We demonstrate performance with and without the defects in the wavefield and discuss the strengths and limitations.

Cite

APA

MacIsaac, M., Beck, A., Eum, W., Tran, C., Harley, J. B., & Subhash, G. (2026). Physics-Informed Filters for Reconstructing Missing Data in Ultrasonic Guided Wavefields. e-Journal of Nondestructive Testing, 31(7). https://doi.org/10.58286/33505

Vancouver

MacIsaac M, Beck A, Eum W, Tran C, Harley JB, Subhash G. Physics-Informed Filters for Reconstructing Missing Data in Ultrasonic Guided Wavefields. e-Journal of Nondestructive Testing. 2026;31(7). doi:10.58286/33505

BibTeX
@article{MacIsaac2026,
  author = {MacIsaac, Michael and Beck, Amanda and Eum, Woohyun and Tran, Charlie and Harley, Joel B. and Subhash, Ghatu},
  title = {Physics-Informed Filters for Reconstructing Missing Data in Ultrasonic Guided Wavefields},
  journal = {e-Journal of Nondestructive Testing},
  year = {2026},
  volume = {31},
  number = {7},
  doi = {10.58286/33505}
}
Start for free →