Physics-Informed Neural Networks for Multiphysics Analysis of Biomedical Polymer Composite Systems

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 | 04 | Page :
By

Vijaykumar Bhanuse,

Shailesh Solanki,

Shilpa S. Ruikar,

Anusha ATMK,

Prakash Kumar Sarangi,

V V N Sarath,

  1. Associate Professor, Department of Instrumentation and Control Engineering, Vishwakarma Institute of Technology, Pune, Maharashtra, India
  2. Associate Professor, Department of Science, School of Sciences, Noida international University, Greater Noida, Uttar Pradesh, India
  3. Associate Professor, Department of Computer Science and Engineering, Krishna Institute of Science and Technology, Krishna Vishwa Vidyapeeth “Deemed to be University”, Karad, Satara, Maharashtra, India
  4. Associate Professor, Department of Allied Health, Meenakshi College of Allied Health Sciences, Meenakshi Medical College Hospital & Research Institute, Meenakshi Academy of Higher Education and Research, Kanchipuram, Tamil Nadu, India
  5. Associate Professor, Department of Computer Science and Engineering (AI&ML), Vardhaman College of Engineering, Shamshabad, Hyderabad, Telangana, India
  6. Associate Professor, Department of Mechanical Engineering, Pragati Engineering College, Kakinada District, Andhra Pradesh, India

Abstract

Physics-Informed Neural Networks (PINNs) offer an effective model of solving coupled multiphysics equations in biomedical polymer composite systems, which are data-driven. In the given work, the PINN method is presented where equations of elasticity, mass diffusion, and heat transfer are integrated to model the complex processes that take place in composite biomaterials. The neural network loss is specified to include the governing partial different equations which enables both the system responses and the physical behavior to be learned at the same time. The proposed structure reduces the use of the massive experimental data and ensures consistency on the physical level. Compared to conventional numerical methods, numerical data can be more accurately predicted and those that are based on standard numerical methods are more predictable and stable, which proves that PINNs are a potentially valuable tool to work with new biomedical materials and optimize their design. It has been demonstrated in the paper that Physics-Informed Neural Network is useful in simulating coupled multi-physics physics in biomedical polymer composite systems. The combination of the process of elasticity, heat transfer and diffusion in one learning model allows the proposed method to be very accurate, stable as well as computationally efficient unlike the traditional methods. The fact that the PINNs can enforce the physical law, and no longer depends on large datasets is what makes them particularly suitable in more complex biomedical problems. Subsequent research will be done on scaling, time-dependent, and nonlinear behavior framework, experimental validation, and scaling, which will allow real-time analysis and design of advanced composite materials in biomedical applications.

Keywords: Physics-Informed Neural Networks, Multiphysics Modeling, Biomedical Polymer Composites, Heat Transfer, Mass Diffusion, Elasticity.

How to cite this article: Vijaykumar Bhanuse, Shailesh Solanki, Shilpa S. Ruikar, Anusha ATMK, Prakash Kumar Sarangi, V V N Sarath. Physics-Informed Neural Networks for Multiphysics Analysis of Biomedical Polymer Composite Systems. Journal of Polymer & Composites. 2026; 14(04):-.
How to cite this URL: Vijaykumar Bhanuse, Shailesh Solanki, Shilpa S. Ruikar, Anusha ATMK, Prakash Kumar Sarangi, V V N Sarath. Physics-Informed Neural Networks for Multiphysics Analysis of Biomedical Polymer Composite Systems. Journal of Polymer & Composites. 2026; 14(04):-. Available from: https://journals.stmjournals.com/jopc/article=2026/view=253036

References

  1. Tang Y, Fu Z, Raos G, Ma F, Zhao P, Hou Y. Molecular dynamics simulation of adhesion at the asphalt-aggregate interface: a review. Surf Interfaces. 2024;44:103706.
  2. Li Y, Chen R, Zhou B, Dong Y, Liu D. Rational design of DNA hydrogels based on molecular dynamics of polymers. Adv Mater. 2024;36:2307129.
  3. Kalateh F, Kheiry M. A review of stochastic analysis of the seepage through earth dams with a focus on the application of Monte Carlo simulation. Arch Comput Methods Eng. 2024;31:47-72.
  4. Vijayan DS, Sivasuriyan A, Devarajan P, Stefańska A, Wodzyński Ł, Koda E. Carbon fibre-reinforced polymer (CFRP) composites in civil engineering application. Buildings. 2023;13:1509.
  5. Schiavo M. Numerical impact of variable volumes of Monte Carlo simulations of heterogeneous conductivity fields in groundwater flow models. J Hydrol. 2024;634:131072.
  6. Gawusu S, Ahmed A. Analyzing variability in urban energy poverty: a stochastic modeling and Monte Carlo simulation approach. Energy. 2024;304:132194.
  7. Shahid M, Monastyreckis G, Zeleniakiene D. Numerical study on elastic properties of natural fibres in multi-hybrid composites. Polymers (Basel). 2025;17:3031.
  8. Zhang Y, Li Y, Luan X, Meng B, Liu J, Lu Y. Effects of void characteristics on the mechanical properties of carbon fiber reinforced polyetheretherketone composites: micromechanical modeling and analysis. Polymers (Basel). 2025;17:1721.
  9. Arzovs A, Judvaitis J, Nesenbergs K, Selavo L. Distributed learning in the IoT-edge-cloud continuum. Mach Learn Knowl Extr. 2024;6:283-315.
  10. Sincak PJ, Prada E, Miková L, Mykhailyshyn R, Varga M, Merva T, et al. Sensing of continuum robots: a review. Sensors (Basel). 2024;24:1311.
  11. Charjan SS. Investigation of some magnetized cosmological model with cosmological constant Λ-term in bimetric theory of gravitation. Adv Phys Lett. 2025;14(1):1-10.
  12. Guo R, Mao L, Xin Z, Ding L. Experimental characterization and micro-modeling of transverse tension behavior for unidirectional glass fibre-reinforced composite. Compos Sci Technol. 2022;222:109359.
  13. Jalili D, Jadidi M, Keshmiri A, Chakraborty B, Georgoulas A, Mahmoudi Y. Transfer learning through physics-informed neural networks for bubble growth in superheated liquid domains. Int J Heat Mass Transf. 2024;232:125940.
  14. Hussain A, Sakhaei AH, Shafiee M. Machine learning-based constitutive modelling for material non-linearity: a review. Mech Adv Mater Struct. 2024:1-19.
  15. Nakka R, Harursampath D, Pathan M, Ponnusami SA. A computationally efficient approach for generating RVEs of various inclusion/fibre shapes. Compos Struct. 2022;291:115560.
  16. Tian W, Xu L, Qi L, Chao X. Minimum potential method appropriate to generate 2D RVEs of composites with high fiber volume fraction. Compos Struct. 2023;318:117070.
Ahead of Print Subscription Original Research
Volume 14
04
Received 01/05/2026
Accepted 30/06/2026
Published 21/08/2026
Publication Time 112 Days

Login

My IP

PlumX Metrics