Sbir fast-track: automated creation of material models for multi-polymer 3D printing

The National Science Foundation awarded Neural Solid, Inc. a $1.555 million cooperative agreement on August 1, 2026, under the NSF Technology, Innovation, and Partnerships program (CFDA 47.084) for automated creation of material models for multi-polymer 3D printing.

Neural Solid will develop a software platform that uses physics-constrained machine learning algorithms to automate the generation of constitutive models — the mathematical descriptions of how materials behave under stress. The platform integrates directly into engineering simulation software manufacturers already use, automating a process that currently requires months of expert labor per material. Phase I focuses on extending an existing framework, previously validated for hyperelastic and elastoplastic single-component materials, to capture rate-dependent viscoelastic behavior in multi-material polymer systems for 3D printing. By shortening the cycle from material discovery to manufacturable product, the technology addresses key product development bottlenecks in multi-material additive manufacturing, where engineers combine different materials in a single printed part to achieve tailored mechanical properties.

The award runs through July 31, 2029, with place of performance in Ithaca, New York.

Abstract

The broader/commercial impact of this small business innovation research (sbir) fast-track project is to transform how engineers develop mathematical descriptions of the physical behaviors of materials or mixtures of materials as they experience stress. Every material requires months of expert labor to characterize before IT can be reliably used in product design and simulation. This is especially limiting the application of multi-material three-dimensional (3D) printing, a technology that allows manufacturers to combine different materials in a single printed part to achieve tailored mechanical properties. This project addresses key product development bottlenecks by creating a software platform that uses physics-constrained machine learning algorithms to automate the generation of material behavior models. The platform integrates directly into the engineering simulation software that manufacturers already use, lowering the barrier to adoption. By shortening the cycle from material discovery to manufacturable product, this technology strengthens domestic manufacturing competitiveness, reduces dependence on international supply chains, expands the workforce capable of performing advanced engineering tasks, and accelerates innovation. This small business innovation research (sbir) fast-track project addresses the challenge of automated constitutive model discovery with an initial focus on multi-material polymer systems used in 3D printing. Constitutive models are essential inputs for finite element analysis, the computational tool engineers use to simulate and design structural components. This project applies physics-informed machine learning to automate that process. Phase i will focus on extending an existing framework, which was previously validated for hyperelastic and elastoplastic single-component materials with high predictive accuracy to capture rate-dependent viscoelastic behavior in multi-polymer photocured resin systems fabricated using polymer-jetting 3D printing technology. Key research objectives include demonstrating model generalization to previously untested material compositions and reducing required experimental characterization by >50%. Phase II will expand the material library to include additional polymer blends and UV-cured adhesives, develop a graphical user interface for non-expert users, and conduct controlled benchmark studies quantifying time savings and usability relative to traditional workflows. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the foundation's intellectual merit and broader impacts review criteria.- subawards are planned for this award.