Constitutive Priors for Inverse Design
Abstract
With recent advances in material synthesis and additive manufacturing, material systems can be designed to achieve prescribed mechanical responses. An important class of such problems is the inverse design of elastic networks that attain a target configuration under loading, with applications in robotics, aerospace, and shape-morphing structures. This work presents a framework that formulates inverse design directly within the learned family of constitutive behaviors. Given a collection of stress-strain responses, we construct a constitutive prior, defined as a low-dimensional latent representation of admissible material laws learned directly from these responses. Spatially varying latent variables are then optimized subject to the governing equilibrium equations so that the deformed network matches a target configuration. The constitutive prior is represented using an energy-based, partially input-convex neural network that enforces the constitutive constraints by construction. To improve robustness in the resulting nonconvex optimization problem, the framework combines homotopy-continuation-based optimization with correspondence-free point cloud matching, allowing the target and optimized geometries to have different discretizations. The proposed approach is demonstrated on several inverse design problems for nonlinear elastic networks, and quantitative comparisons with alternative optimization strategies show improved robustness and optimization performance.
Versions
Related
- Well-posed by Design: Learning Constitutive Laws from Velocity Data using Convex Neural Network Potentials
- Design of a combined polarimetric and velocimetric measurement for viscoelastic stress, and its constitutive resolving power
- Using Data-Derived Priors to Guide CNN Architecture Design for NIR Chemometrics
- Gradient-based inverse design of phase-shifting transmission masks for EUV interference lithography
- Inverse design of scalar-tensor EFT from odd-parity effective potentials
BibTeX
@misc{han2026constitutive,
title = {{Constitutive Priors for Inverse Design}},
author = {Jinkyo Han and Bahador Bahmani},
howpublished = {arXiv},
year = {2026},
month = aug,
doi = {10.48550/arxiv.2605.09307},
eprint = {2605.09307},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2605.09307},
note = {CC BY 4.0},
}
RIS
TY - UNPB
TI - Constitutive Priors for Inverse Design
AU - Jinkyo Han
AU - Bahador Bahmani
PY - 2026
DA - 2026/08/21
PB - arXiv
DO - 10.48550/arxiv.2605.09307
UR - https://arxiv.org/abs/2605.09307
AB - With recent advances in material synthesis and additive manufacturing, material systems can be designed to achieve prescribed mechanical responses. An important class of such problems is the inverse design of elastic networks that attain a target configuration under loading, with applications in robotics, aerospace, and shape-morphing structures. This work presents a framework that formulates inverse design directly within the learned family of constitutive behaviors. Given a collection of stress-strain responses, we construct a constitutive prior, defined as a low-dimensional latent representation of admissible material laws learned directly from these responses. Spatially varying latent variables are then optimized subject to the governing equilibrium equations so that the deformed network matches a target configuration. The constitutive prior is represented using an energy-based, partially input-convex neural network that enforces the constitutive constraints by construction. To improve robustness in the resulting nonconvex optimization problem, the framework combines homotopy-continuation-based optimization with correspondence-free point cloud matching, allowing the target and optimized geometries to have different discretizations. The proposed approach is demonstrated on several inverse design problems for nonlinear elastic networks, and quantitative comparisons with alternative optimization strategies show improved robustness and optimization performance.
ER -