{"title":"Experimental, Numerical, and ANN-Based Study of Free Vibration of Glass Fibre and MWCNT-Reinforced Laminated Structures","authors":"Dhaneshwar Prasad Sahu, Ramyaranjan Das, Sukesh Chandra Mohanty","doi":"10.1007/s12221-025-01034-8","DOIUrl":null,"url":null,"abstract":"<div><p>This study develops a predictive model for the natural frequencies of multi-walled carbon nanotube (MWCNT)-reinforced woven glass fibre-metal laminates (FMLs) using Artificial Neural Networks (ANNs). Laminates were fabricated via the open mould technique, and material properties were obtained through uniaxial tensile testing per the ISO 527–5 standard. Experimental modal analysis (EMA) and finite-element simulations in ABAQUS using S58R shell elements with engineering constants showed strong agreement. The ANN was trained on experimental and numerical data to predict the natural frequencies under varying laminate sequences, aspect ratios, and thickness ratios. The obtained results indicate that increasing the aspect ratio, side-to-thickness ratio, and fibre orientation angle reduces the natural frequencies of FML. The first mode of natural frequency increased by 83.80%, 84.85%, and 89.52% for CFCF, CCCF, and CCCC boundary conditions, respectively, compared to CFFF. Conversely, increasing fibre angles led to reductions in the natural frequency of 1.662–7.385% for CFFF, 0.027–1.146% for CFCF, 0.175–0.705% for CCCF, and 0.036–0.148% for CCCC. The ANN model demonstrated high accuracy and efficiency, supporting its use in the design and optimization of advanced composite structures. </p></div>","PeriodicalId":557,"journal":{"name":"Fibers and Polymers","volume":"26 8","pages":"3551 - 3571"},"PeriodicalIF":2.3000,"publicationDate":"2025-06-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Fibers and Polymers","FirstCategoryId":"88","ListUrlMain":"https://link.springer.com/article/10.1007/s12221-025-01034-8","RegionNum":4,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q1","JCRName":"MATERIALS SCIENCE, TEXTILES","Score":null,"Total":0}
引用次数: 0
Abstract
This study develops a predictive model for the natural frequencies of multi-walled carbon nanotube (MWCNT)-reinforced woven glass fibre-metal laminates (FMLs) using Artificial Neural Networks (ANNs). Laminates were fabricated via the open mould technique, and material properties were obtained through uniaxial tensile testing per the ISO 527–5 standard. Experimental modal analysis (EMA) and finite-element simulations in ABAQUS using S58R shell elements with engineering constants showed strong agreement. The ANN was trained on experimental and numerical data to predict the natural frequencies under varying laminate sequences, aspect ratios, and thickness ratios. The obtained results indicate that increasing the aspect ratio, side-to-thickness ratio, and fibre orientation angle reduces the natural frequencies of FML. The first mode of natural frequency increased by 83.80%, 84.85%, and 89.52% for CFCF, CCCF, and CCCC boundary conditions, respectively, compared to CFFF. Conversely, increasing fibre angles led to reductions in the natural frequency of 1.662–7.385% for CFFF, 0.027–1.146% for CFCF, 0.175–0.705% for CCCF, and 0.036–0.148% for CCCC. The ANN model demonstrated high accuracy and efficiency, supporting its use in the design and optimization of advanced composite structures.
期刊介绍:
-Chemistry of Fiber Materials, Polymer Reactions and Synthesis-
Physical Properties of Fibers, Polymer Blends and Composites-
Fiber Spinning and Textile Processing, Polymer Physics, Morphology-
Colorants and Dyeing, Polymer Analysis and Characterization-
Chemical Aftertreatment of Textiles, Polymer Processing and Rheology-
Textile and Apparel Science, Functional Polymers