Messages

There are packages for linux, including centos 7 at:
https://github.com/iRASPA/RASPA3/releases

You probably used "surface"-rendering. With "volume-rendering" you can get more of a color-gradient.
Bonding is controlled by a general rule, also employed by VMD: The radius < (covalent radius of atom1 + atom2 + 0.56).

No, RASPA does not use parallelism. But, for example, usually each point in an isotherm is computed separately and you can use an "embarrassingly parallel" approach.

No, that is not possible. Usually docking software is used for that. I would look into 'gromacs' for biomolecules as a start.

In principle there is no limit. But Monte Carlo is limited to small molecules, usually molecules of 1-20 atoms. 30 atoms is already considered very, very large.

That is what makes classical simulations more difficult than quantum. But for many systems, forcefields have been published in the literature.

In RASPA2 you input the final result: the partition-factor. In more recent RASPA versions the Log of the partition function to avoid over/underflow.
It is an user input, and up to the user to compute it.

You can look these radii up in handbooks.

You could modify the RASPA code, or label the atoms differently.

I would modify the force field. One of the options is to put a hard-sphere potential around the hydrogen to shield it from overlap.

[pairs]

That is not implemented. Usually, these should be the same, since you are looping of interaction-pairs.

Are you using the official version from https://github.com/iraspa/raspa2 ?
Versions by other people (not us) via python-use reportedly can show this issue.

If so, check that it actually writes output by running MUCH shorter (10 cycles) and check all the output.
In the output-file it should say whether it computes MSD or not.

This is on the TODO list.

In the gallery, then primitives, then IRMOF-1, you can see such an example.

Note that if you do not use an initialization phase, the bias is measured while the adsorption is going up.
Also note that you are comparing CBCMC vs CFCMC. I would guess that CB/CFCMC would probably not plateau like that during initialization.