[gmx-users] r_list and r_coulomb values while using PME

[David van der Spoel]

OZGE ENGIN wrote:
> So, using twin-range cut-off does not violate the energy conservation for RC method, but it violates for PME method. Is this the case ?
> 
twin range does not conserve energy with cutoff or RF either. however 
you must always compromise between accuracy and computational cost, 
therefore in some cases twin range can be defended.

> 
> -----Original Message-----
> From: David van der Spoel <spoel at xray.bmc.uu.se>
> To: Discussion list for GROMACS users <gmx-users at gromacs.org>
> Date: Sat, 01 Sep 2007 14:48:03 +0200
> Subject: Re: [gmx-users] r_list and r_coulomb values while using PME
> 
> OZGE ENGIN wrote:
>> 'The force is not the derivative of the potential', is it due to the types of calculations performed in PME, using of FFT for reciprocal sum? Because it does not matter while using reaction-field method. 
>> Sorry, this phenomenon is  a little complicated to me :) 
> 
> 
> this is only when you combine twin range (rcoulomb > rlist) and PME. 
> then the forces for a certain distance range are not computed at each 
> time step while the others are. you may want to (re)read the PME papers 
> and/or the gromacs manual about twin range.
>> Thanks for your attention!
>>
>> Oz.
>>
>> -----Original Message-----
>> From: David van der Spoel <spoel at xray.bmc.uu.se>
>> To: Discussion list for GROMACS users <gmx-users at gromacs.org>
>> Date: Sat, 01 Sep 2007 13:37:05 +0200
>> Subject: Re: [gmx-users] r_list and r_coulomb values while using PME
>>
>> OZGE ENGIN wrote:
>>> Hi all,
>>>
>>> I can not understand why it is required to take same values for  both r_list and r_coulomb parameters when using PME for calculation of electrostatic interactions?
>>>
>>> I read archieve, and gromacs manual. In archieve, similar question had been posted, but not replied yet.
>>>
>>> Thanks in advance
>> if rlist < rcoulomb you  will have a twin range interaction, meaning the 
>> atom pair interactions for distance between these two will be computed 
>> every nstlist steps only, while the interaction beyond rcoulomb will be 
>> computed at each time step. this will lead to bad energy conservation, 
>> since the force is not the derivative of the potential.
>>> Oz.
>>>
>>> _______________________________________________
>>> gmx-users mailing list    gmx-users at gromacs.org
>>> http://www.gromacs.org/mailman/listinfo/gmx-users
>>> Please search the archive at http://www.gromacs.org/search before posting!
>>> Please don't post (un)subscribe requests to the list. Use the 
>>> www interface or send it to gmx-users-request at gromacs.org.
>>> Can't post? Read http://www.gromacs.org/mailing_lists/users.php
>>
> 
> 


-- 
David van der Spoel, Ph.D.
Molec. Biophys. group, Dept. of Cell & Molec. Biol., Uppsala University.
Box 596, 75124 Uppsala, Sweden. Phone:	+46184714205. Fax: +4618511755.
spoel at xray.bmc.uu.se	spoel at gromacs.org   http://folding.bmc.uu.se