VASP training data generation. The training data for the MTPs is generated in these active learning runs.
Moment Tensor Potential (MTP) training runs. Typically 5 training runs with different random initialisations are performed, which are labelled with a, b, c, d and e.
VASP validation runs. VASP active learning runs are performed at 300 K in an NPT ensemble to determine RMSD_MD
ALD-BTE calculations. collision matrices for each grid point are omitted due to their huge size.
MD-BTE calculations. The file with the velocities (named velocities_dyn.dump) requires up to 300 GB; therefore it is omitted. The renormalized force constants require 5 GB of storage (for PT) and are thus also omitted. The calculation is performed in two steps, employing first LAMMPS, and then Dynaphopy.
AEMD simulations
NEMD simulations
In order to align the polymer chains along the z-direction, and to be consistent with our previous publication "Designing Accurate Moment Tensor Potentials for Phonon-Related Properties of Crystalline Polymers" (https://www.mdpi.com/1420-3049/29/16/3724), we relabeld the axis in the following way:
x --> z
y --> y
z --> x
x --> y
y --> x
z --> z
The q-mesh for the phono3py calculation of PE is given as 160×15×10 in the job-submission-script where phono3py is called, but in the paper (and above) it is written as 10×15×160. All calculations (including POSCAR files etc.) are consistent with each other in terms of their axis labels. Just for the paper, the axis relabelling was performed.