On Surrogate Modeling of Static Response of AM Short-Fiber Thermoplastics Using Graph Neural Networks

Pharindra Pathak, Vipin Kumar, Trenton M. Ricks, Suhasini Gururaja, Siddhartha Srivastava · 2026

Read On Surrogate Modeling of Static Response of AM Short-Fiber Thermoplastics Using Graph Neural Networks by Pharindra Pathak, Vipin Kumar, Trenton M. Ricks, Suhasini Gururaja, Siddhartha Srivastava (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2606-28996.txt.

Description

Pathak, Kumar, Ricks, Gururaja and Srivastava build a fast stand-in model for short-fiber thermoplastics, light plastic parts reinforced with short fibers, used in aerospace and car structures. Their mechanical response depends on messy inner detail: fiber direction, clustering and pores left by manufacturing. Detailed finite element models can resolve that detail, but a realistic part with thousands of fibers is far too costly to simulate directly.

The team rebuilt three-dimensional microstructures from high-resolution CT scans and split them into small Voronoi cells, each a distinct fiber neighbourhood. Nonlinear simulations of each cell, including matrix damage, produced stress-strain data for training. A hybrid graph neural network with a memory layer then learned both the structure's layout and its history-dependent response. The trained model predicted stiffness and stress-strain behaviour of unseen structures with accuracy near 0.98 against full simulations, at over a hundred times less cost.

Catalog record

CREATORPharindra Pathak, Vipin Kumar, Trenton M. Ricks, Suhasini Gururaja, Siddhartha Srivastava
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE102 KB
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2606.28996

Browse the full library · About the Tower of Babel