[CP2K-user] [CP2K:13252] Re: AD_LANGEVIN thermostat for QM/MM with fixed_atoms

[Mayank Dodia]

Hi Noam,

The original files I attached earlier contained no constraints (fixed atoms 
or intramolecular).

What I have noticed is that for AD_LANGEVIN thermostat with:

1) REGION GLOBAL: Does not work in QM, MM  or QM/MM system (with or without 
constraints). For pure QM, no constraints (H2O-32.inp) I get a cholesky 
decomposition error which is gone once I change the thermostat settings.

2) REGION MASSIVE: Fully functional.

Since we need to use a global thermostat for fixed atoms, so effectively 
AD_LANGEVIN doesn't seems to work with constraints and/or REGION GLOBAL 
thermostat setting.

Best,
Mayank
 

On Wednesday, May 6, 2020 at 6:00:13 PM UTC+2, Noam Bernstein wrote:
>
> It’s quite possible that AD_LANGEVIN doesn’t handle constrained atoms 
> correctly. I certainly didn’t have them in mind when I coded it up. Can you 
> confirm that it works OK with REGION GLOBAL if there are no fixed atoms?
>
>     Noam
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.cp2k.org/archives/cp2k-user/attachments/20200506/186155a5/attachment.htm>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: H2O-32.inp
Type: chemical/x-gamess-input
Size: 6078 bytes
Desc: not available
URL: <https://lists.cp2k.org/archives/cp2k-user/attachments/20200506/186155a5/attachment.inp>