Materials modeling

Materials modeling uses computational or mathematical representations to predict and explain materials’ structures, properties, processing, and performance across scales.

Connect

Density functional theory: A quantum-mechanical method that estimates a system’s electronic structure from its electron density. It predicts bonding, energies, and electronic properties from atomic arrangements.
Interatomic potential: A mathematical function that estimates the potential energy of atoms as a function of their positions. It makes large-scale atomistic simulations feasible without solving electronic structure directly.
Computational materials design: The use of computation to identify or optimize materials for specified properties and functions. It turns predictions into candidate materials tailored to target performance.
First-principles calculation: A calculation derived from fundamental physical laws with minimal empirical input. It contrasts with models fitted to experimental data or simpler approximations.
Model validation: The assessment of whether a model adequately represents reality for its intended use. Agreement with independent measurements determines whether predicted properties can be trusted.
Molecular dynamics: A simulation method that calculates atomic trajectories by integrating equations of motion. It models how atoms move and materials respond over time.
Crystal structure: The ordered arrangement of atoms, ions, or molecules in a crystalline solid. It supplies the geometry underlying many calculations of material properties.
Materials discovery: The identification of previously unknown materials with useful compositions, structures, or properties. Model-driven screening can narrow the search across vast chemical spaces.
Empirical model: A model whose parameters or relationships are fitted to observations rather than derived solely from theory. Its speed and data dependence differ from first-principles approaches.
Transfer learning: A machine-learning approach that adapts knowledge from one task or dataset to another. It may reduce the data needed to model new chemistries or environments.