Design and Validation of an Artificial Intelligence-Driven Digital Twin for Real-Time Monitoring and Control in Polymer Composite Manufacturing

Notice

This is an unedited manuscript accepted for publication and provided as an Article in Press for early access at the author’s request. The article will undergo copyediting, typesetting, and galley proof review before final publication. Please be aware that errors may be identified during production that could affect the content. All legal disclaimers of the journal apply.

Year : 2026 | Volume : 14 | 02 | Page :
    By

    Manish Kumar Jha,

  1. PhD, Department of Mathematics (Computer Science Research), Patliputra University, Patna, Bihar, India

Abstract

Polymer Matrix Composites (PMCs) have become indispensable in high-performance sectors such as aerospace and automotive engineering, offering exceptional strength-to-weight ratios that outperform traditional metals in many demanding applications. However, the reliability of manufacturing PMCs via Vacuum-Assisted Resin Transfer Molding (VARTM) is frequently undermined by stochastic process variabilities. Unpredictable fluctuations in thermal history, preform permeability, resin rheology, and ambient conditions often lead to some defects; namely voids, dry spots, and incomplete cure; which compromise structural integrity, increase scrap rates, and necessitate costly post-process remediation or rejection. To address these persistent challenges, this study proposes and validates a comprehensive Digital Twin (DT) framework that tightly integrates high-fidelity physics-based modeling with real-time Artificial Intelligence (AI) control. The system continuously synthesizes in-situ data from a distributed sensor network (including thermocouples for thermal profiling, pressure transducers for flow monitoring, and fiber optic strain sensors for compaction state) with a numerical model that accurately captures Darcy flow in anisotropic porous media and autocatalytic cure kinetics. A Convolutional Neural Network (CNN), built on a customized VGG-16 architecture and trained on thousands of augmented multi-modal sensor images, identifies complex defect signatures in real time, while a Model Predictive Control (MPC) algorithm dynamically optimizes vacuum pressure gradients and thermal cycles to prevent defect formation. Experimental validation involving thirty controlled manufacturing trials demonstrated that the DT-driven closed-loop control significantly outperformed conventional open-loop processing. Specifically, the average void volume fraction was reduced from 3.2% ± 0.7% to 0.9% ± 0.3% (p < 0.001), and part rejection rates dropped from 18% to 6.3%. Mechanical characterization further revealed an 18.1% increase in tensile strength, attributed to minimized stress concentrations, improved fiber-matrix interfacial adhesion, and more uniform cure distribution. These findings underscore the transformative potential of AI-enhanced digital twins in shifting composite manufacturing from a largely craft-based, trial-and-error practice to a precise, data-driven, and adaptive science; with clear benefits in reducing waste, improving yield, enhancing sustainability, and enabling higher confidence in safety-critical applications such as aerospace structures.

Keywords: Digital twin; Artificial intelligence; Polymer composites; Vacuum-assisted resin transfer molding; Real-time monitoring; Defect detection; Model Predictive Control; Smart manufacturing.

How to cite this article:
Manish Kumar Jha. Design and Validation of an Artificial Intelligence-Driven Digital Twin for Real-Time Monitoring and Control in Polymer Composite Manufacturing. Journal of Polymer & Composites. 2026; 14(02):-.
How to cite this URL:
Manish Kumar Jha. Design and Validation of an Artificial Intelligence-Driven Digital Twin for Real-Time Monitoring and Control in Polymer Composite Manufacturing. Journal of Polymer & Composites. 2026; 14(02):-. Available from: https://journals.stmjournals.com/jopc/article=2026/view=239666


References

[1] Grieves M. Digital twin: Manufacturing excellence through virtual factory replication. White Paper. 2014;1:1-7.

[2] Marsh G. Airbus A350 XWB update. Reinforced Plastics. 2010;54(6):20-24.

[3] Friedrich K, Almajid AA. Manufacturing aspects of structural automotive polymer composites. Composite Structures. 2013;95:10-23.

[4] European Commission. Factories of the Future: Multi-Annual Roadmap for the Contractual Public-Private Partnership under Horizon 2020. Brussels; 2013.

[5] Advani SG, Sozer EM. Process Modeling in Composites Manufacturing. 2nd ed. Boca Raton: CRC Press; 2010.

[6] Wang J, Zhang Z, Liang G. Voids in fiber-reinforced polymer composites: A review on their formation, characteristics, and effects on mechanical performance. Journal of Composite Materials. 2019;53(15):2045-2075.

[7] Poodts E, Minak G, Zucchelli A. Impact of process parameters on the void content of VARTM manufactured composites. Advanced Composite Materials. 2013;22(5):331-346.

[8] Potter K. Resin Transfer Moulding. 1st ed. London: Chapman & Hall; 1997.

[9] Amenabar I, Mendikute A, López-Arraiza A, Lizaranzu M, Aurrekoetxea J. Comparison and analysis of non-destructive testing techniques suitable for delamination inspection in wind turbine blades. Composites Part B: Engineering. 2011;42(5):1298-1305.

[10] Teti R, Vasilakos C. Tool wear monitoring techniques for turning: A review. International Journal of Machine Tools and Manufacture. 2010;50(12):1037-1052.

[11] Brokel H, Weber J, Hüllmann R. Towards a Digital Twin for Liquid Composite Molding. Procedia CIRP. 2019;85:305-310.

[12] Tao F, Zhang M, Nee AYC. Digital twin in industry: State-of-the-art. IEEE Transactions on Industrial Informatics. 2019;15(4):2405-2415.

[13] Kritzinger W, Karner M, Traar G, Henjes J, Sihn W. Digital Twin in manufacturing: A categorical literature review and classification. IFAC-PapersOnLine. 2018;51(11):1016-1022.

[14] Liu M, Fang S, Dong H, Xu C. Review of digital twin about concepts, technologies, and industrial applications. Journal of Manufacturing Systems. 2021;58:346-361.

[15] Hojjati M. Automation of composite manufacturing processes: A review. Polymers. 2019;11(10):1554.

[16] Grieves M, Vickers J. Digital Twin: Mitigating Unpredictable, Undesirable Emergent Behavior in Complex Systems. In: Kahlen F-J, Flumerfelt S, Alves A, editors. Transdisciplinary Perspectives on Complex Systems. Springer; 2017. p. 85-113.

[17] Tao F, Cheng J, Qi Q, Zhang M, Zhang H, Sui F. Digital twin-driven product design, manufacturing and service with big data. The International Journal of Advanced Manufacturing Technology. 2018;94:3563-3576.

[18] Jones D, Snider C, Nassehi A, Yon J, Hicks B. Characterising the Digital Twin: A systematic literature review. CIRP Journal of Manufacturing Science and Technology. 2020;29:36-52.

[19] Liu C, Xu X, Hu D. Multiobjective optimization of composite laminated structures using digital twins. Composite Structures. 2021;255:112932.

[20] Nielsen D, Pitchumani R. Closed-loop flow control in resin transfer molding using real-time numerical process simulations. Composites Science and Technology. 2002;62(2):283-298.

[21] Zhang C, Liu J. Digital twin-enabled reconfigurable modeling for smart manufacturing. International Journal of Production Research. 2020;58(10):3012-3028.

[22] Sacco C, Radl B, Woo J, Harne RL, Kim HA. Machine learning in composites manufacturing: A case study of Automated Fiber Placement defect detection. Composite Structures. 2020;250:112514.

[23] Schubel PJ, Warrior NA, Costen GS. Monitoring liquid composite moulding using an acoustic emission sensor. Composites Part A: Applied Science and Manufacturing. 2012;43:1595–1602.

[24] Heider D, Piovoso MJ, Gillespie Jr JW. Intelligent control of the VARTM process. Journal of Composite Materials. 2004;38(22):2065-2090.

[25] Schmachtenberg E, Gerber C, Seidel A. Process control in resin transfer moulding (RTM). Polymer Testing. 2005;24(3):330-336.

[26] Humfeld KD. Machine learning for thermal history control in autoclave composite processing. Composites Part A: Applied Science and Manufacturing. 2022;155:106818.

[27] Tang T, Dai X, Liu Y, Zhao J. MQTT-based internet of things for data acquisition and control in industrial applications. IEEE Access. 2019;7:82345-82354.

[28] Murukeshan VM, Chan PZ, Ong LS, Seah LK. Cure monitoring of composite using fiber Bragg grating sensors. Sensors and Actuators A: Physical. 2000;79(2):153-161.

[29] Kamal MR, Sourour S. Kinetics and thermal characterization of thermoset cure. Polymer Engineering & Science. 1973;13(1):59-64.

[30] Rasmussen CE. Gaussian processes in machine learning. In: Advanced Lectures on Machine Learning. Berlin: Springer; 2004. p. 63-71.

[31] Maihtripala DHSC, Dwivedy SK. A review on surrogate modeling techniques for optimization of composite structures. Composite Structures. 2021;262:113642.

[32] Simonyan K, Zisserman A. Very deep convolutional networks for large-scale image recognition. arXiv preprint arXiv:1409.1556. 2014.

[33] Nielsen D, Leith J. Model predictive control of resin transfer molding for optimum fill time and void minimization. Composites Part A: Applied Science and Manufacturing. 2018;115:102-114.

[34] Greenhalgh E. Failure Analysis and Fractography of Polymer Composites. Cambridge: Woodhead Publishing; 2009.

[35] Weiss K, Khoshgoftaar TM, Wang D. A survey of transfer learning. Journal of Big Data. 2016;3(1):1-40.

[36] Bang J, Kim S, Kim H. Deep learning-based inspection of defects in industrial manufacturing: A review. Applied Sciences. 2019;9(20):4416.

[37] Zhang L, Wang Z. Intelligent fault diagnosis of rolling bearings using a new convolutional neural network and transfer learning. IEEE Access. 2020;8:99690-99700.

[38] Lee J, Bagheri B, Kao HA. A cyber-physical systems architecture for industry 4.0-based manufacturing systems. Manufacturing Letters. 2015;3:18-23.

[39] Shi W, Cao J, Zhang Q, Li Y, Xu L. Edge computing: Vision and challenges. IEEE Internet of Things Journal. 2016;3(5):637-646.

[40] Elkington M, Bloom D, Ward C, Chatzimichali A, Potter K. Hand layup: understanding the manual process. Advanced Manufacturing: Polymer & Composites Science. 2015;1(3):138-151.

 

Ahead of Print Subscription Original Research
Volume 14
02
Received 31/01/2026
Accepted 20/02/2026
Published 03/04/2026
Publication Time 62 Days
34

Login


My IP

PlumX Metrics