EUVIMEDEUVIMED entdecken
Crossref

Prediction of Grain Orientation from Thermal Maps Using a Data-Driven and AI Approach

Nishi Bhemani, Krishnan Balasubramaniam

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

DOI: 10.58286/33512 ↗

Kurz gefasst

Fachartikel in der Zeitschrift e-Journal of Nondestructive Testing, 2026-07. Verfasst von Nishi Bhemani und 1 weiteren Autor:innen.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Understanding microstructure is fundamental to predicting the mechanical or thermal, properties of materials. The present study focuses on microstructure characterization using thermal imaging by integrating data-driven and AI approach. Synthetic microstructures are generated using Voronoi tessellation, representing polycrystalline materials with 100 orthotropic grains. Each grain is oriented in random directions. Thermal images of the microstructure are obtained by numerical methods. In this way the dataset is generated using finite element methods. Each dataset sample is labelled according to grain orientation angles ranging from 0° to 90°, segmenting the images into 90 discrete classes. A semantic segmentation network is trained to get grain orientation directly from the thermal map of the microstructure. The accuracy of the gain orientation prediction by the Semantic segmentation establishes the potential of thermal imaging combined with data driven modeling as a powerful tool for microstructure characterization.

Zitieren

APA

Bhemani, N., & Balasubramaniam, K. (2026). Prediction of Grain Orientation from Thermal Maps Using a Data-Driven and AI Approach. e-Journal of Nondestructive Testing, 31(7). https://doi.org/10.58286/33512

Vancouver

Bhemani N, Balasubramaniam K. Prediction of Grain Orientation from Thermal Maps Using a Data-Driven and AI Approach. e-Journal of Nondestructive Testing. 2026;31(7). doi:10.58286/33512

BibTeX
@article{Bhemani2026,
  author = {Bhemani, Nishi and Balasubramaniam, Krishnan},
  title = {Prediction of Grain Orientation from Thermal Maps Using a Data-Driven and AI Approach},
  journal = {e-Journal of Nondestructive Testing},
  year = {2026},
  volume = {31},
  number = {7},
  doi = {10.58286/33512}
}
Kostenlos starten →