General scheme for kMC is very easy. You can look into http://www.msm.cam.ac.uk/phase-trans/2012/Bombac_Thesis.pdf for description of MC methods.

Next question you need to answer is what are you trying to simulate as code I am using is rigid lattice code. If your intent is vacancy diffusion then you can use it, but if you try to solve different system then I am afraid it will not help you much.

I was working with kMC scheme few years ago in view of vacancy migrated processes, Yuo can also find some information here:

http://scd-theses.u-strasbg.fr/2022/01/BIBORSKI_Andrzej_2010.pdf

If I were you I would write simple code to gain intuition in kMC (also known as Residence Time Algorithm though kMC seems to be wider class of approaches than RTA) i.e. 2D Ising system with and and without external magentic field at various temepratures and exchange parameters assumed.

I am thankful for information provided by @ David Bombač and @Andrzej Biborski!

Now I am using Monte Carlo based on Metropolis algorithm. I need to generate a new configuration of the system having 5 degrees of freedom (I have 3D model described by 3 space coordinates and 2 spherical angles) I wonder about the effectiveness of my following thoughts

1. I think to separate trail moves. What this mean? Like to make trail moves for X and Y coordinates, then to calculate the energy of the new configuration and compare it with the old one, then generate trail move for Z etc instead generating trail moves for the all 5 degrees of freedom at once. But I have doubts about this strategy and I need advice from the experts

2. Particularly, for the trail moves for the space-coordinates  is it reasonable to take different step-sizes of each coordinates or not? The argumentation is that the energy of the system depends on z coordinate only, and acceptance of the state, therefore depends on z, therefore having different step-sizes for  trail moves for  X,Y and Z  can reveal some properties of the system, which can not be seen before (just a gues)

Thanks for the help

 Remember that in kMC You need to have whole transition matrix at each time - it means that you must ensure that all possible transitions are computed. If you got the problem where degrees of freedom are "discrete"  it is (at least in theory) no problematic (i.e. you have finite set of angles and X,Y,Z steps) , if no (i.e. you randomize the X,Y,Z step length and/or angle in your example) you are not able to create whole transition matrix (since you have infinite number of possible transitions from  the current microstate). If you use Metropolis scheme you may just random which process is selected  (i.e. with equal probability - but it depends on your particular simulation) and than if one selected, just calculate the energy change for that particularly chosen MC step. If Metropolis is  your case your idea seeems to be reasonable (however you must be careful especially if you are interested in the equilibrium properties) of the system.

According to the second question: it is you who decide which processes are important, interesting etc. - it is why we speak about modelling. We  rather attempt to get as close as possible to the reality. Sometimes it is good strategy to examine the whole hierarchy of models starting from the simplest one.

Hi,

I might be a little bit late, and the Code I propose is not a "simple" one. Nevertheless, if it helps...

We develop an Open Source Code written in C++. MMonCa:

http://www.materials.imdea.org/MMonCa

It has been used for many things so far: 

http://www.materials.imdea.org/research/publications/atomistic-materials-modeling

Hope it helps. 

Best,

Jose

Hi all

Dear Dr. Gomez

The code MMonCa is no longer hosted nor maintained at IMDEA Materials Institute and the mentioned site. I wonder what was happened? because, it was recently available.

with regards

S.M.Zamzamian

Hi all,

I am currently developing a flexible, open-source lattice KMC package that can be used to more easily construct KMC simulations for a wide variety of scientific problems. You can check it out on Github (https://github.com/MikeHeiber/KMC_Lattice). I hope you find it useful, and your suggestions for improvement are always welcome!