arn-all / kinetic_monte_carlo_code Goto Github PK

View Code? Open in Web Editor NEW

This project forked from marcucla/kinetic_monte_carlo_code

1.0 1.0 0.0 1.52 MB

CMake 0.14% C++ 99.37% Shell 0.01% Python 0.48%

kinetic_monte_carlo_code's Introduction

kinetic_monte_carlo_code's People

Contributors

Stargazers

Watchers

kinetic_monte_carlo_code's Issues

Completly disable solute migration to save computing time

Implement non-uniform nucleation probability

Alterate enthalpy of nucleation for segments aligned with solute

Force nucleation of pinned double kinks neighbours

Make all 3 unpinning mechanisms disableable

Dislocations have 3 unpinning mechanisms here : 2 kinks coming from the same side piling up, 2 kinks from opposite side annihilating at the position of the solute, or the thermally activated unpinning from eq.(4) Zhao, Marian, MSMSE, 2018.
To investigate the effect of each mechanism, each of them should be independently enable/disable-able from within the config file.
Status :

Annihilation :
Pile up: allow_stacked_kinks keyword introduced (c612d58)
Thermally activated : allow_debinding keyword introduced (e54b861)

Implement basic testing functionality

When the code is updated, make it easy to test the new build

Allow choosing between cottrell atmosphere and custom atoms positions

It should be possible in the config file

Only the last row of solutes is pinning in JSON mode

Speed up many solutes case

Having a lot of solute atoms (eg. concentration = 0.01) slows down the code a lot.
running a profiler valgrind --tool=callgrind ../../KMC_arnall/bin/DislocationKMC --config test.config reveals the bottleneck is that Simulation::generatePointDefectEventList() makes a huge number of calls to DislocationNetwork::isSoluteOnTheDislocation()
this function has been optimized a bit (many variable definitions in loop) eacd753
a better strategy for neighbours lookup is needed, as ~~70% of runtime is spent there

gprof2dot/gprof2dot.py -n 2 -f callgrind callgrind.out.15812 | dot -Tsvg -o output.svg

Extract dislocation velocity

Write some proper post-processing script for extracting the position / velocity of the dislocation, using existing notebooks and stuff

Alterate nucleation enthalpy for pinned double kink

Allow C atoms positions to be given as a np.ndarray

Write some function that transforms np.ndarray to KMC-readable json file.

Allow two kinks to be pinned by a signe C atom

------       ---------
     |   C   |
     ---------

Move all physical parameters to the config file

Get rid of hard-coded values