Deep-learning for elastic wave elds obtained by time-domain boundary element method and its application to non-destructive testing
-
- SAITOH Takahiro
- Gunma University
-
- MINOWA Riho
- Gunma University
-
- NAKAMURA Maya
- Gunma University
-
- ITO Tsukasa
- Gunma University
Bibliographic Information
- Other Title
-
- 時間領域境界要素法で得られた散乱波動場の深層学習と非破壊評価への応用
Abstract
<p>The goal of ultrasonic non-destructive evaluation (UT) is to detect a defect in materials. In recent years, the deep-learning attracts much attention in a wide variety of engineering fields. Therefore, in this research, the deep-learning for elastic wave scattering by a defect is implemented and is applied to automatic ultrasonic non-destructive evaluation. The elastic wave fields are obtained using the convolution quadrature time-domain boundary element method. Numerical results show that the deep-learning for elastic wave scattering may have a potential to solve the staffing shortage for UT inspectors.</p>
Journal
-
- NCTAM papers, National Congress of Theoretical and Applied Mechanics, Japan
-
NCTAM papers, National Congress of Theoretical and Applied Mechanics, Japan 65 (0), 187-, 2019
National Committee for IUTAM
- Tweet
Details 詳細情報について
-
- CRID
- 1390852581598279424
-
- NII Article ID
- 130008100890
-
- Text Lang
- ja
-
- Data Source
-
- JaLC
- CiNii Articles
-
- Abstract License Flag
- Disallowed