AI-powered prediction of friction and wear in functionalized Epoxy-MWCNT composites

IF 6.1 1区工程技术 Q1 ENGINEERING, MECHANICAL

Wear Pub Date : 2025-09-10 DOI:10.1016/j.wear.2025.206333

Ravisrini Jayasinghe , Maziar Ramezani

{"title":"AI-powered prediction of friction and wear in functionalized Epoxy-MWCNT composites","authors":"Ravisrini Jayasinghe , Maziar Ramezani","doi":"10.1016/j.wear.2025.206333","DOIUrl":null,"url":null,"abstract":"<div><div>This study investigates the self-lubricating properties of epoxy composites reinforced with multi-walled carbon nanotubes (MWCNTs) functionalized with carboxyl (COOH), amine (NH<sub>2</sub>), and silane coupling agents. Advanced artificial intelligence (AI) methods were employed to predict key tribological parameters coefficient of friction (COF) and wear rate as well as to classify wear mechanisms. Wear tests were conducted using a linear reciprocating tribometer under dry sliding conditions with a chromium steel ball as the counter surface. To simulate demanding conditions beyond typical industrial polymer bearings, sliding frequencies of 2, 5, and 8 Hz were applied under normal loads of 5, 10, and 15 N. The resulting data were analyzed using AI tools to identify patterns and predict tribological performance.</div><div>An Artificial Neural Network (ANN) was developed to model the relationships between COF, wear rate, experimental conditions, mechanical properties, and MWCNT composition, achieving high predictive accuracy (R<sup>2</sup> = 0.98 for COF; R<sup>2</sup> = 0.78 for wear rate). A Recurrent Neural Network (RNN) was used to capture the temporal evolution of COF under dominant wear mechanisms, including abrasion, adhesion, fatigue, and severe delamination. Additionally, a Convolutional Neural Network (CNN) accurately classified wear mechanisms from scanning electron microscopy (SEM) images. These AI-driven approaches provide a robust predictive framework for understanding and optimizing the tribological behavior of functionalized MWCNT-epoxy composites.</div></div>","PeriodicalId":23970,"journal":{"name":"Wear","volume":"582 ","pages":"Article 206333"},"PeriodicalIF":6.1000,"publicationDate":"2025-09-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Wear","FirstCategoryId":"5","ListUrlMain":"https://www.sciencedirect.com/science/article/pii/S0043164825006027","RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q1","JCRName":"ENGINEERING, MECHANICAL","Score":null,"Total":0}

引用次数: 0

Abstract

This study investigates the self-lubricating properties of epoxy composites reinforced with multi-walled carbon nanotubes (MWCNTs) functionalized with carboxyl (COOH), amine (NH₂), and silane coupling agents. Advanced artificial intelligence (AI) methods were employed to predict key tribological parameters coefficient of friction (COF) and wear rate as well as to classify wear mechanisms. Wear tests were conducted using a linear reciprocating tribometer under dry sliding conditions with a chromium steel ball as the counter surface. To simulate demanding conditions beyond typical industrial polymer bearings, sliding frequencies of 2, 5, and 8 Hz were applied under normal loads of 5, 10, and 15 N. The resulting data were analyzed using AI tools to identify patterns and predict tribological performance.

An Artificial Neural Network (ANN) was developed to model the relationships between COF, wear rate, experimental conditions, mechanical properties, and MWCNT composition, achieving high predictive accuracy (R² = 0.98 for COF; R² = 0.78 for wear rate). A Recurrent Neural Network (RNN) was used to capture the temporal evolution of COF under dominant wear mechanisms, including abrasion, adhesion, fatigue, and severe delamination. Additionally, a Convolutional Neural Network (CNN) accurately classified wear mechanisms from scanning electron microscopy (SEM) images. These AI-driven approaches provide a robust predictive framework for understanding and optimizing the tribological behavior of functionalized MWCNT-epoxy composites.

查看原文本刊更多论文

基于ai的功能化环氧- mwcnt复合材料摩擦磨损预测

本研究研究了羧基（COOH）、胺（NH2）和硅烷偶联剂功能化的多壁碳纳米管（MWCNTs）增强环氧复合材料的自润滑性能。采用先进的人工智能（AI）方法预测关键摩擦学参数摩擦系数（COF）和磨损率，并对磨损机理进行分类。在干滑动条件下使用线性往复摩擦计进行磨损试验，以铬钢球作为计数器。为了模拟典型工业聚合物轴承以外的苛刻条件，在5、10和15 n的正常载荷下，施加2、5和8 Hz的滑动频率，使用人工智能工具分析所得数据，以识别模式并预测摩擦学性能。利用人工神经网络（ANN）对COF、磨损率、实验条件、力学性能和MWCNT成分之间的关系进行建模，获得了较高的预测精度（COF的R2 = 0.98，磨损率的R2 = 0.78）。使用循环神经网络（RNN）捕捉COF在主要磨损机制下的时间演变，包括磨损、粘附、疲劳和严重分层。此外，卷积神经网络（CNN）从扫描电子显微镜（SEM）图像中准确地分类了磨损机制。这些人工智能驱动的方法为理解和优化功能化mwcnt -环氧复合材料的摩擦学行为提供了强大的预测框架。

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

Wear 工程技术-材料科学：综合

CiteScore

8.80

自引率

8.00%

发文量

280

审稿时长

47 days

期刊介绍： Wear journal is dedicated to the advancement of basic and applied knowledge concerning the nature of wear of materials. Broadly, topics of interest range from development of fundamental understanding of the mechanisms of wear to innovative solutions to practical engineering problems. Authors of experimental studies are expected to comment on the repeatability of the data, and whenever possible, conduct multiple measurements under similar testing conditions. Further, Wear embraces the highest standards of professional ethics, and the detection of matching content, either in written or graphical form, from other publications by the current authors or by others, may result in rejection.