A General Materials Data Science Framework for Quantitative 2D Analysis of Particle Growth from Image Sequences

IF 2.4 3区材料科学 Q3 ENGINEERING, MANUFACTURING

Integrating Materials and Manufacturing Innovation Pub Date : 2024-02-20 DOI:10.1007/s40192-024-00342-w

Sameera Nalin Venkat, Thomas G. Ciardi, Mingjian Lu, Preston C. DeLeo, Jube Augustino, Adam Goodman, Jayvic Cristian Jimenez, Anirban Mondal, Frank Ernst, Christine A. Orme, Yinghui Wu, Roger H. French, Laura S. Bruckman

{"title":"A General Materials Data Science Framework for Quantitative 2D Analysis of Particle Growth from Image Sequences","authors":"Sameera Nalin Venkat, Thomas G. Ciardi, Mingjian Lu, Preston C. DeLeo, Jube Augustino, Adam Goodman, Jayvic Cristian Jimenez, Anirban Mondal, Frank Ernst, Christine A. Orme, Yinghui Wu, Roger H. French, Laura S. Bruckman","doi":"10.1007/s40192-024-00342-w","DOIUrl":null,"url":null,"abstract":"<p>Phase transformations are a challenging problem in materials science, which lead to changes in properties and may impact performance of material systems in various applications. We introduce a general framework for the analysis of particle growth kinetics by utilizing concepts from machine learning and graph theory. As a model system, we use image sequences of atomic force microscopy showing the crystallization of an amorphous fluoroelastomer film. To identify crystalline particles in an amorphous matrix and track the temporal evolution of the particle dispersion, we have developed quantitative methods of 2D analysis. 700 image sequences were analyzed using a neural network architecture, achieving 0.97 pixel-wise classification accuracy as a measure of the correctly classified pixels. The growth kinetics of isolated and impinged particles were tracked throughout time using these image sequences. The relationship between image sequences and spatiotemporal graph representations was explored to identify the proximity of crystallites from each other. The framework enables the analysis of all image sequences without the requirement of sampling for specific particles or timesteps for various materials systems.</p>","PeriodicalId":13604,"journal":{"name":"Integrating Materials and Manufacturing Innovation","volume":null,"pages":null},"PeriodicalIF":2.4000,"publicationDate":"2024-02-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Integrating Materials and Manufacturing Innovation","FirstCategoryId":"88","ListUrlMain":"https://doi.org/10.1007/s40192-024-00342-w","RegionNum":3,"RegionCategory":"材料科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q3","JCRName":"ENGINEERING, MANUFACTURING","Score":null,"Total":0}

引用次数: 0

Abstract

Phase transformations are a challenging problem in materials science, which lead to changes in properties and may impact performance of material systems in various applications. We introduce a general framework for the analysis of particle growth kinetics by utilizing concepts from machine learning and graph theory. As a model system, we use image sequences of atomic force microscopy showing the crystallization of an amorphous fluoroelastomer film. To identify crystalline particles in an amorphous matrix and track the temporal evolution of the particle dispersion, we have developed quantitative methods of 2D analysis. 700 image sequences were analyzed using a neural network architecture, achieving 0.97 pixel-wise classification accuracy as a measure of the correctly classified pixels. The growth kinetics of isolated and impinged particles were tracked throughout time using these image sequences. The relationship between image sequences and spatiotemporal graph representations was explored to identify the proximity of crystallites from each other. The framework enables the analysis of all image sequences without the requirement of sampling for specific particles or timesteps for various materials systems.

Abstract Image

查看原文本刊更多论文

从图像序列对粒子生长进行二维定量分析的通用材料数据科学框架

相变是材料科学中一个具有挑战性的问题，相变会导致材料性能发生变化，并可能影响材料系统在各种应用中的性能。我们利用机器学习和图论的概念，为粒子生长动力学分析引入了一个通用框架。作为模型系统，我们使用原子力显微镜图像序列来显示无定形氟橡胶薄膜的结晶过程。为了识别无定形基质中的结晶颗粒并跟踪颗粒分散的时间演变，我们开发了二维定量分析方法。我们使用神经网络架构分析了 700 个图像序列，作为正确分类像素的衡量标准，像素分类准确率达到了 0.97。利用这些图像序列对孤立颗粒和撞击颗粒的生长动力学进行了全程跟踪。探索了图像序列和时空图表征之间的关系，以确定晶体之间的距离。该框架可对所有图像序列进行分析，而无需对各种材料系统的特定颗粒或时间步进行采样。

本文章由计算机程序翻译，如有差异，请以英文原文为准。

求助全文

约1分钟内获得全文求助全文

来源期刊

Integrating Materials and Manufacturing Innovation Engineering-Industrial and Manufacturing Engineering

CiteScore

5.30

自引率

9.10%

发文量

审稿时长

39 days

期刊介绍： The journal will publish: Research that supports building a model-based definition of materials and processes that is compatible with model-based engineering design processes and multidisciplinary design optimization; Descriptions of novel experimental or computational tools or data analysis techniques, and their application, that are to be used for ICME; Best practices in verification and validation of computational tools, sensitivity analysis, uncertainty quantification, and data management, as well as standards and protocols for software integration and exchange of data; In-depth descriptions of data, databases, and database tools; Detailed case studies on efforts, and their impact, that integrate experiment and computation to solve an enduring engineering problem in materials and manufacturing.