This is a script to compute the phase composition of MoNbTaW multiprinciple elements alloy at given temperatures and composition. It requires the list of candidate ordered phases (in data folder) with DFT computed energies, volumes, bulk modulus and B'.
There is a possibility to estimate the uncertainty of the phase diagrams by inclusion of a random noise with controlalble magnitude into data.
- Yury Lysogorskiy (1)
- Alberto Ferrari (1,2)
Affiliations:
- Interdisciplinary Centre for Advanced Materials Simulation, Ruhr-University Bochum, 44801 Bochum, Germany
- Materials Science and Engineering, Delft University of Technology, 2628CD Delft, The Netherlands
See LICENSE file
You will require some packages to run the code. These packages could be installed in a following manner.
- installation with
conda:
conda env create -f environment.yml
conda activate HEA
- installation with
pip:
pip install ase numpy scipy pandas numba pyyaml
Installation was tested on Python 3.7 and 3.8.
File name: input.yaml
# nominal concentration to simulate
nominal_concentration: {Mo: 0.25 , Nb: 0.25, Ta: 0.25, W: 0.25}
# inclusion of ordered structures based on distance to conv_hull (eV)
max_distance_to_convex_hull: 0.002
# list of temperatures to consider
Ts: [50, 300, 1200]
# number of initial configuration guesses
initial_configuration_guesses: 10
# number of noise-attempts. Set it to 0 if you do not want to apply noise
N_noise_attempts: 0
# gaussian noise std for disordered phase energies (eV)
disord_energy_std: 0.003
# gaussian noise std for ordered phase energies (eV)
ord_energy_std: 0.002
# number of parallel process to compute independend configuration attempts
max_workers: 2
# optmizer maxiter
maxiter: 20
# path to the data folder
DATA_PATH: ./data The script requires the input file input.yaml
python hea_phase_simulations.py
As a result, the file total_data.tsv will be generated.
This is table, that for each temperature, initial phase configuration and noise attempt contains:
T: temperature (K),Energy: energy of the phase configuration (eV/atom)noise_attempt: index of the noies attempt (=0 for no noise)Disord_phase_1, ... : Concentrations of disordered phases (from 1 to 4)Phase_1/Mo, ... : Composition of each disordered phase (Mo, Nb, Ta, W for each of the four phases )TaMo2W_1018282, ... : concentration of the given ordered phases
The table is ordered by Energy. If no noise were applied, you could consider the minimal energy for each temperature.
Otherwise, for each T and noise_attempt minimal energy should be considered. Then these phases results should be averaged over noise attempts.