Project title:

Machine learning for predicting Lattice thermal conductivity


Neeraj Kulhari

Defense Year: 2024-2025

To obtain the anharmonic force constants, which depends on the cutoff distance and lattice symmetry, may require a few hundred/thousand single-point DFT force calculations. We propose the employment of machine-learning interatomic potentials (MLIP) to substitute the computationally demanding DFT calculations in obtaining the anharmonic force constants using PANNA code.