Physics-Informed Neural Networks
Efficient and interpretable surrogates for structural analysis.
Computational mechanics · Scientific ML
Thonn
Homsnit
PhD Student in Mechanical Science and Engineering
I study how mechanics and machine learning can work together to make structural simulations more reliable, interpretable, and efficient.
// Research focus
Computational
Mechanics.
My research sits at the intersection of computational mechanics and machine learning.
Continuum Mechanics
Moderate to large deformation and hyperelasticity.
Graph Neural Networks
Learning representations for finite element simulations.
// Latest news
What I have
been working on.
Slovak–Japanese One-Day Session 2026
Presented a Hessian-based diagnosis of physics-informed neural network formulations at Meiji University.
Mirai Catalyst Hackathon
Joined an innovation idea competition organized by TSAJ, with the final pitch at Miraikan, Tokyo.
Corotational formulation demonstration
Published a Newton–Raphson method demonstration video.
New PINN preprint
PINN stability and robustness for Euler–Bernoulli beam problems.
// Publications
Research
publications.
-
2026
Optimization and reliability analysis of multi-purpose electric truck chassis under payload positioning variation
T. Khansiriwong, K. Ruangjirakit, J. Srirat, S. Kongwat, T. Homsnit, and P. Jongpradist. Proceedings of the Institution of Mechanical Engineers, Part D, 240(4), 2090–2102.
-
2025
Integrated Structural and Impact Optimization for a Lightweight 6-Wheel Electric Truck Chassis
J. Srirat, J. Theppakdee, J. Tiemkhum, N. Karimee, T. Homsnit, and P. Jongpradist. International Journal of Automotive Technology, 1–12.
-
2025
Investigation of PINN Stability and Robustness for the Euler–Bernoulli Beam Problem
T. Homsnit, K. Kageyama, and T. Kojima. arXiv preprint arXiv:2511.19916.
-
2024
Crashworthiness design of an automotive S-rail using ANN-based optimization to enhance performance and safety
T. Homsnit, P. Jongpradist, S. Kongwat, P. Jongpradist, and C. Thongchom. Structural and Multidisciplinary Optimization, 67(6), 93.
-
2024
Application of artificial neural networks for prediction of crushing performance of a Quadricycle’s S-rail
T. Homsnit, P. Jongpradist, S. Kongwat, and P. Jongpradist. AIP Conference Proceedings, 3086(1), 060001.
-
2023
Optimizing stiffness and lightweight design of composite monocoque sandwich structure for electric heavy quadricycle
T. Homsnit, S. Kongwat, K. Ruangjirakit, P. Noykanna, T. Thuengsuk, and P. Jongpradist. Latin American Journal of Solids and Structures.
-
2022
Safety assessment and crash compatibility of heavy quadricycle under frontal impact collisions
S. Kongwat, T. Homsnit, C. Padungtree, N. Tonitiwong, P. Jongpradist, and P. Jongpradist. Sustainability, 14(20), 13458.
// Education & experience
Academic
journey.
Saitama University · Japan
PhD Student in Mechanical Science
Japanese Government MEXT Scholar. Focused on physics-informed neural networks and computational mechanics.
Virginia Tech · USA
Graduate Student & Teaching Assistant
Engineering mechanics coursework and teaching support for Mechanics of Deformable Bodies.
Gulf Energy Development PCL · Thailand
Performance Engineer
KMUTT · Thailand