Dresden 2020 – wissenschaftliches Programm
Die DPG-Frühjahrstagung in Dresden musste abgesagt werden! Lesen Sie mehr ...
Bereiche | Tage | Auswahl | Suche | Aktualisierungen | Downloads | Hilfe
MM: Fachverband Metall- und Materialphysik
MM 26: Topical Session: Data Driven Materials Science - Machine Learning for Materials Properties
MM 26.1: Vortrag
Dienstag, 17. März 2020, 14:15–14:30, BAR 205
From Atom Probe Tomography to CALPHAD modeling: Estimating Tc from local concentration fluctuations — •Marvin Poul, Sebastian Eich, and Guido Schmitz — Universität Stuttgart, Stuttgart, Germany
One way to determine the extent of the miscibility gap and the associated critical solution temperature Tc in binary alloys from Atom Probe Tomography (APT) is to prepare nano-layer stacks, anneal them and determine the respective layer concentrations, which mark the boundaries in the phase diagram. Since this relies on diffusion, it can be problematic when Tc and atomic mobilities are low, such as in Cu/Ni, leading to long annealing times.
This work proposes a novel methodology based on statistical mechanics to extract Tc from histograms of thermodynamically inherent local concentration fluctuations annealed above Tc, i.e. in the region of complete miscibility. The same formalism allows to extract relative chemical potential differences from two or more samples with different mean concentration. Given enough data to span the full concentration range it is even possible to non-parametrically recover the excess free energy of mixing gex(c), which allows a direct approach to a CALPHAD parametrization.
The approach is benchmarked using Embedded Atom Monte Carlo simulations of Cu/Ni and applicability to experimental histograms from APT is discussed.