Sensitivity estimation for calculated phase equilibria

Abstract The development of a consistent framework for Calphad model sensitivity is necessary the rational reduction uncertainty via new models and experiments. In present work, theory was developed, closed‐form expression log‐likelihood gradient Hessian multi‐phase equilibrium measurement presented. inherent locality defined metric mitigated through use Monte Carlo averaging. A case study Cr–Ni system used to demonstrate visualizations analyses enabled by developed theory. Criteria based on classical Cramér–Rao bound were shown be useful diagnostic in assessing accuracy parameter covariance estimates from Markov Chain Carlo. applied estimate statistical value phase equilibria measurements comparison with thermochemical measurements, implications reduction.