Download file (3.27MB)
Download as zip (3MB)

コレクション

引用

説明:

Studying multiple properties of a material concurrently is essential for obtaining a comprehensive understanding of its behavior and performance. However, this approach presents certain challenges. For instance, simultaneous examination of various properties often necessitates extensive experimental resources, thereby increasing the overall cost and time required for research. Furthermore, the pursuit of desirable properties for one application may conflict with those needed for another, leading to trade-off scenarios. In this study, we focused on investigating adhesive joint strength and elastic modulus, both crucial properties directly impacting adhesive behavior. To determine elastic modulus, we employed a non-destructive indentation method for converting hardness measurements. Additionally, we introduced a specimen apparatus preparation method to ensure the fabrication of smooth surfaces and homogeneous polymeric specimens, free from voids and bubbles. Our experiments utilized a commercially available bisphenol A-based epoxy resin in combination with a Poly(propylene glycol) curing agent. We generated an initial dataset comprising experimental results from 32 conditions, which served as input for training a machine learning model. Subsequently, we used this model to predict outcomes for a total of 256 conditions. To address the high deviation in prediction results, we implemented active learning approaches, achieving a 50% reduction in deviation while maintaining model accuracy. Through our analysis, we observed a trade-off boundary (Pareto frontier line) between adhesive joint strength and elastic modulus. Leveraging Bayesian optimization, we successfully identified experimental conditions that surpassed this boundary, yielding an adhesive joint strength of 25.2 MPa and an elastic modulus of 182.5 MPa.

権利情報:

• Creative Commons BY Attribution 4.0 International
  

キーワード: experimental testing, multi-objective optimization, epoxy, adhesive, elastic modulus, machine learning, active learning

刊行年月日: 2024-06-12

出版者: MDPI AG

掲載誌:

• Materials (ISSN: 19961944) vol. 17 issue. 12 2866

研究助成金:

• Japan Science and Technology Agency JPMJCR19J3
• KAKENHI Grant-in-Aid for Scientific Research 23H02031
• MEXT Program: Data Creation and Utilization-Type Material Research and Development Project JPMXP1122714694

原稿種別: 出版者版 (Version of record)

MDR DOI:

公開URL: https://doi.org/10.3390/ma17122866

関連資料:

その他の識別子:

連絡先:

更新時刻: 2025-01-07 08:30:26 +0900

MDRでの公開時刻: 2025-01-07 08:30:26 +0900