Abnormal atoms position output when Stiffness is smaller than 1

Started by Haolong, March 13, 2021, 01:30:58 PM

Previous topic - Next topic

Haolong

 Dear Dubbeldam,

There is abnormal position output for atoms in the molecule with extremely low stiffness(1e-20), which is shown as below:

ATOM      1  O   MOL            24.547   0.074   0.871  1.00  0.00           O 
ATOM      2  H   MOL            24.330  -0.629   1.483  1.00  0.00           H 
ATOM      3  H   MOL            24.677  -0.370   0.033  1.00  0.00           H 
ATOM      4  O   MOL            28.454  26.114   3.621  1.00  0.00           O 
ATOM      5  H   MOL            28.578  25.307   4.121  1.00  0.00           H 
ATOM      6  H   MOL            29.160  26.107   2.976  1.00  0.00           H 
ATOM      7  K   MOL          -54312228503.48764424509490.218-8260815321.219  1.00  0.00           K 
ATOM      8 Cl   MOL          59751111813.535-73398710417.3009088062024.881  1.00  0.00          Cl 
ATOM      9  K   MOL          -7045318285.63313304773408.7447403388878.031  1.00  0.00           K 
ATOM     10 Cl   MOL          7750843838.857-14637127284.649-8144771966.366  1.00  0.00          Cl 

The reason why I set the stiffness as 1e-20 is that I want to insert the K and CL ions in pairs and make sure these ions can be distributed freely in the system to simulate the condense system with water molecules and ions, at the meantime, maintain the electric neutrality.

However, it turns out that when the stiffness is lower than 1,e.g: 1e-5,1e-10 and 1e-15, the output of the ions' position becomes higher than the box length of the system.

At first, I think that's because of the wrap-up issue, so I wrap those coordinates up by the length of the box, but the ions aggregate on one side and even overlap with each other, which is shown in the attachment.

Could you please tell me: 1. Can I set the stiffness lower than 1 in RASPA?
                                     2. How can I wrap up properly in this situation?

Thanks a lot!

Kind regards
Haolong


David Dubbeldam

I don't think this is currently possible. The implementation of the CBMC algorithm used for the insertion is the limitation here.