AI Business LensTHE BUSINESS OF AI, FOR PEOPLE WHO TEACH IT OR LEARN FROM IT
Phys.org — Physics · September 12, 2026

AI extracts interpretable constitutive laws directly from solid-mechanics data

Materials ScienceMechanical EngineeringPhysics
THE AI ANGLE
Autonomously discovering interpretable constitutive equations from experimental data

Researchers have developed GraphED, a graph-based AI framework that directly extracts concise, physically interpretable constitutive equations from solid-mechanics experimental data by simultaneously optimizing mathematical topology and material parameters. Validated on alloy steels, lithium metal, and filled rubbers, the method outperforms established empirical formulations such as the Johnson–Cook model in predictive accuracy. This development marks a major shift for solid mechanics and materials curricula, moving beyond the historical constraint of fitting data to predefined mathematical equations.

THE TEACHING ANGLE
Instructors can examine the shift from traditional constitutive modeling—which relies on physical intuition to predetermine equation structures before fitting parameters—to data-driven frameworks that autonomously search for both mathematical forms and parameterizations.

Read the original at phys.org   Generate teaching or study materials

Instructors get discussion guides, assignments, and mini-cases. Students and readers get a plain summary, class prep, and an exercise. All built from the full article. Three are free with an account.

More in Materials Science