RMC config strain analyzer
Python script for strain field analysis for RMC6F configuration. Details about strain analysis theoretical description can be found in the following paper,
https://doi.org/10.1107/S1600576719000372
1. For microstrain analysis (Eqn. 11 in the paper mentioned above), we need a single RMC6F configuration as the input. For the deformation gradient tensor analysis, we need the initial and fitted RMC6F configurations as inputs.
The script has been embodied into the RMCProfile release package so that it can be simply executed as,
rmc_strain RMC6F_CONFIG [OPTIONS]
For example, if we want to analyze the microstrain, we can type something like,
rmc_strain -i ceriaNano_NP.rmc6f -t rms
If we want to analyze deformation gradient tensor, we can type something like,
rmc_strain -i ceriaNano_NP.rmc6f -t dgt -r ceriaNano_NP_init.rmc6f
One can see a list of options by typing
rmc_strain -h
The full list of [OPTIONS] is presented below,
- -h
Show current help information.
- -i
[RMC6F config name] Input RMC6F configuration.
- -t
[rms/dgt] Analysis type.
- -r
[Reference RMC6F config file] If ‘dgt’ analysis is selected, one needs to provide the reference RMC6F configuration for computing the deformation gradient tensor.
- -v
Show version information.
The program will then ask some questions interactively during execution,
a) If ‘rms’ type of analysis is selected, the program will ask whether to do the shell analysis. Then it will ask whether this is for bulk when executing shell analysis. The program was originally designed for analyzing nanoparticle, and for the analysis we need to figure out the center and radius of the particle first. Sometimes, we may also want to run the program against bulk as well just to make sure, with some confidence, what we obtain from nanoparticle shell analysis is real. However, for bulk model, it may be tricky to figure out the center and radius, so we need extra input here.
If we do want to do it for bulk, we need to first generate a nanoparticle model from the bulk configuration. To do this, we can use the ‘bulk_shells.py’ script in the ‘NP_Shells’ folder. The usage is similar to ‘np_shells.py’, and here we only need to generate one shell (by specifying ‘by thickness’ for particle generation and particle radius as the shell thickness), which is actually a nanoparticle in the center. Then we want to copy the ‘shell_0.rmc6f’ and ‘np_shell_gen.log’ to the same directory where we execute the ‘rmc_strain.py’ script.
Then the program will ask the minimum number of cells in each shell. We may want to figure out an estimation based on the total number of cells information printed out in the log file (see list above).
b) If ‘dgt’ type of analysis is selected, the program will ask for the cutoff (in angstrom) for the local deformation analysis. Usually, 10 angstrom should be good enough. This analysis will take a while since figuring out all neighbors for all atoms is time consuming.
The output is described as follows, assuming the input fitted RMC6F configuration is with the name of ‘ceriaNano_NP.rmc6f’.
Output if ‘rms’ mode selected,
ceriaNano_NP_#.log – Overall microstrain result
ceriaNano_NP_#_shells.log – Microstrain for various shells.
where ‘#’ represents the smallest integer that is not already existing in the output file names.
Output if ‘dgt’ mode selected,
ceriaNano_NP_dgt.out – Deformation gradient tensor for each single atom.
ceriaNano_NP_rot.out – Rotation tensor for each single atom.
ceriaNano_NP_strain.out – Strain tensor for each single atom.
ceriaNano_NP_strain_invar.out – Strain invariants for each single atom.
Explanation for the output quantities can be again found in the paper, https://doi.org/10.1107/S1600576719000372.