Non-destructive visual characterization of grain features in metallic materials using laser ultrasonics
Materials Research Express · 13.11.2026 · 13 (21) · 216505
DOI: 10.1088/2053-1591/aebccc ↗Kurz gefasst
Fachartikel in der Zeitschrift Materials Research Express, online erschienen am 10.11.2026. Verfasst von Yinuo Rong und 3 weiteren Autor:innen. Die Arbeit nennt 26 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Projects in Science and Technique Plans of Ningbo City, Natural Science Foundation of Zhejiang Province.
Abstract (Originalsprache)
Precise quantitative characterization of microstructural grain features is essential to analyze and tailor the mechanical performance of titanium materials. This study develops a non-contact, rapid nondestructive characterization framework integrating laser ultrasonics and logarithm-transformed ridge regression for statistical grain quantification of commercially pure titanium TA1. Six TA1 specimens with graded grain microstructures are prepared by controlled heat treatment, with electron backscatter diffraction results adopted as ground truth. Complete ensemble empirical mode decomposition (CEEMD) is implemented for ultrasonic signal denoising, followed by extraction of three core ultrasonic descriptors: wave velocity, time-domain attenuation coefficient and frequency-domain attenuation coefficient. Two independent ridge regression models are constructed to synchronously predict the mean grain size and grain size standard deviation. Under leave-one-specimen-out cross-validation, which provides fully independent specimen-level verification free of data leakage, the model yields an mean absolute error (MAE) of 2.74 μ m and R 2 = 0.9123 for mean grain size, alongside MAE = 1.65 μ m and R 2 = 0.8874 for size dispersion. As a supplementary point-level reference, the 80/20 split gives MAE = 2.08 μ m/ R 2 = 0.9469 and MAE = 1.18 μ m/ R 2 = 0.9266. Full-field grain distribution contours are reconstructed via inversion. The proposed methodology delivers a viable technical pathway for rapid nondestructive grain evaluation, showing potential for future industrial deployment in titanium quality inspection and process optimization, pending validation on larger and more diverse datasets.
Förderung
- Projects in Science and Technique Plans of Ningbo City
- Natural Science Foundation of Zhejiang Province
Zitieren
APA
Rong, Y., Shi, H., Li, Q., & Yin, A. (2026). Non-destructive visual characterization of grain features in metallic materials using laser ultrasonics. Materials Research Express, 13(21), 216505. https://doi.org/10.1088/2053-1591/aebccc
Vancouver
Rong Y, Shi H, Li Q, Yin A. Non-destructive visual characterization of grain features in metallic materials using laser ultrasonics. Materials Research Express. 2026;13(21):216505. doi:10.1088/2053-1591/aebccc
BibTeX
@article{Rong2026,
author = {Rong, Yinuo and Shi, Hao and Li, Qiyuan and Yin, Anmin},
title = {Non-destructive visual characterization of grain features in metallic materials using laser ultrasonics},
journal = {Materials Research Express},
year = {2026},
volume = {13},
number = {21},
pages = {216505},
doi = {10.1088/2053-1591/aebccc}
}
EUVIMED