Asif Rashid, Farzad Vatandoust, Akshar Kota, Shreyes N. Melkote
{"title":"Bead geometry prediction in wire arc directed energy deposition using physics-informed machine learning and low-fidelity data","authors":"Asif Rashid, Farzad Vatandoust, Akshar Kota, Shreyes N. Melkote","doi":"10.1016/j.addma.2025.104881","DOIUrl":null,"url":null,"abstract":"<div><div>Wire Arc Directed Energy Deposition (Wire Arc DED) is a promising metal additive manufacturing technique, yet accurate bead geometry prediction remains a challenge due to the complex thermal and geometric interactions in the process. In this study, we present a coupled Physics-Informed Neural Network (PINN) framework to predict the bead geometry by integrating the governing process physics and experimental data, thereby addressing the limitations of both computationally expensive numerical models and purely data-driven approaches. The model employs a sequential two-step workflow, where a thermal model first predicts temperature evolution, which subsequently informs a geometry model for predicting the bead geometry. Results indicate that a high-fidelity PINN model with high spatiotemporal resolution captures the intricately coupled thermal and geometric variations inherent to bead deposition with good predictive accuracy albeit at a higher computational cost, while a low-fidelity PINN model with lower spatiotemporal resolution offers a computationally efficient alternative with marginally higher errors. The incorporation of measured bead geometry data significantly enhances prediction accuracy, with a minimal amount of low-fidelity data sufficing to refine predictions effectively. Moreover, the model generalizes well across different bead locations along the deposition length, demonstrating reliable performance. The high-fidelity PINN model, using a temporal step size of 0.2 s, achieves an average height prediction error of 8.38 % and width error of 1.09 % after approximately 12.7 hours of training on four H100 GPUs. In contrast, the low-fidelity model, with a coarser temporal step size of 0.5 s, reaches nearly the same accuracy (8.33 % height error, 1.56 % width error) with just 2.7 h of training on a single H100 GPU. This corresponds to a 79 % reduction in training time and substantially lower hardware requirements, highlighting the scalability and efficiency of the proposed hybrid modeling approach.</div></div>","PeriodicalId":7172,"journal":{"name":"Additive manufacturing","volume":"109 ","pages":"Article 104881"},"PeriodicalIF":11.1000,"publicationDate":"2025-07-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Additive manufacturing","FirstCategoryId":"5","ListUrlMain":"https://www.sciencedirect.com/science/article/pii/S2214860425002453","RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q1","JCRName":"ENGINEERING, MANUFACTURING","Score":null,"Total":0}
引用次数: 0
Abstract
Wire Arc Directed Energy Deposition (Wire Arc DED) is a promising metal additive manufacturing technique, yet accurate bead geometry prediction remains a challenge due to the complex thermal and geometric interactions in the process. In this study, we present a coupled Physics-Informed Neural Network (PINN) framework to predict the bead geometry by integrating the governing process physics and experimental data, thereby addressing the limitations of both computationally expensive numerical models and purely data-driven approaches. The model employs a sequential two-step workflow, where a thermal model first predicts temperature evolution, which subsequently informs a geometry model for predicting the bead geometry. Results indicate that a high-fidelity PINN model with high spatiotemporal resolution captures the intricately coupled thermal and geometric variations inherent to bead deposition with good predictive accuracy albeit at a higher computational cost, while a low-fidelity PINN model with lower spatiotemporal resolution offers a computationally efficient alternative with marginally higher errors. The incorporation of measured bead geometry data significantly enhances prediction accuracy, with a minimal amount of low-fidelity data sufficing to refine predictions effectively. Moreover, the model generalizes well across different bead locations along the deposition length, demonstrating reliable performance. The high-fidelity PINN model, using a temporal step size of 0.2 s, achieves an average height prediction error of 8.38 % and width error of 1.09 % after approximately 12.7 hours of training on four H100 GPUs. In contrast, the low-fidelity model, with a coarser temporal step size of 0.5 s, reaches nearly the same accuracy (8.33 % height error, 1.56 % width error) with just 2.7 h of training on a single H100 GPU. This corresponds to a 79 % reduction in training time and substantially lower hardware requirements, highlighting the scalability and efficiency of the proposed hybrid modeling approach.
期刊介绍:
Additive Manufacturing stands as a peer-reviewed journal dedicated to delivering high-quality research papers and reviews in the field of additive manufacturing, serving both academia and industry leaders. The journal's objective is to recognize the innovative essence of additive manufacturing and its diverse applications, providing a comprehensive overview of current developments and future prospects.
The transformative potential of additive manufacturing technologies in product design and manufacturing is poised to disrupt traditional approaches. In response to this paradigm shift, a distinctive and comprehensive publication outlet was essential. Additive Manufacturing fulfills this need, offering a platform for engineers, materials scientists, and practitioners across academia and various industries to document and share innovations in these evolving technologies.