[DFTB-Plus-User] Rotational DOF in gas phase MD
Ran Friedman, Biochemisches Inst.
r.friedman at bioc.uzh.ch
Mon Feb 22 11:30:28 CET 2010
This looks interesting, but isn't it more difficult to implement?
Ran
On Mon, 22 Feb 2010 09:30:58 +0000
Victor Milman <VMilman at accelrys.com> wrote:
> PS There is potentially better thermostat available now, by Leimkuhler (J
>Stat Phys 135 (2009) 261-277) - it appears to be more rigorous in proving
>its ergodicity, for example.
>
> ======================================================
> Victor Milman Senior Fellow
> Accelrys Inc tel: (01223) 228500/228619
> 334 Science Park fax: (01223) 228501
> Cambridge CB4 0WN tel. from abroad: +44 1223 228500
> UK e-mail:
>vmilman at accelrys.com<mailto:vmilman at accelrys.com>
> ======================================================
>
>From: dftb-plus-user-bounces at dftb-plus.info
>[mailto:dftb-plus-user-bounces at dftb-plus.info] On Behalf Of Manuel Dömer
> Sent: 22 February 2010 09:28
> To: dftb-plus-user at dftb-plus.info
> Subject: Re: [DFTB-Plus-User] Rotational DOF in gas phase MD
>
> Dear Ben,
>
> thank you very much for you explanations and hints. I have been using the
>Berendsen thermostat for reasons associated with other simulations on this
>specific project.
> But I definitely will give the Andersen a try. Right now it is especially
>important for me to thermalize the system in a good way. For data
>collection in the production phase, I think I can run NVE dynamics.
>
> Many thanks again for your clarifications!
>
> Best Regards,
>
> Manuel
>
> Dear Ran,
>
> thanks for pointing me to the Bussi thermostat.
>
>
>
>
>
>
> ----------------------------------------------------------------------
>
> Message: 1
> Date: Fri, 19 Feb 2010 12:29:08 +0000
>From: Ben Hourahine
><benjamin.hourahine at strath.ac.uk<mailto:benjamin.hourahine at strath.ac.uk>>
> Subject: Re: [DFTB-Plus-User] Rotational DOF in gas phase MD
> To: User list for DFTB+ related questions
> <dftb-plus-user at dftb-plus.info<mailto:dftb-plus-user at dftb-plus.info>>
> Message-ID:
><4B7E8414.8090601 at strath.ac.uk<mailto:4B7E8414.8090601 at strath.ac.uk>>
> Content-Type: text/plain; charset=ISO-8859-1
>
> Hello Manuel,
>
> which thermostat are you using? This sounds like the known artefacts
> of the Berendsen thermostat (flying icecube effect).
>
> At the moment, as you've observed, only the translational motion is
> removed in the code, but with Berendsen the lowest accessible modes
> always end up accumulating kinetic energy. It would be possible to
> remove the rotational degrees of freedom as well for non-periodic
> calculations, but we haven't done this. If rotation were also removed
> then using Berendsen thermostating would instead lead to the low energy
> collective bend modes of the molecule heating up.
>
> Would the alternative choice of the Andersen thermostat any use? This
> thermostat does not have the same pathology (but has other problems - it
> prevents collection of some types of autocorrelation data and suppresses
> atomic diffusion, but does correctly sample from the NVT ensemble).
>
> As this molecule has bonds between C and H, I'd suggest testing a
> shorter time step is probably a good idea anyway, irrespective of
> any thermostatting choices.
>
> Regards
>
> Ben
>
> Manuel D?mer wrote:
>
> Dear DFTB users,
>
> I have been running an NVT MD of a peptide in gas phase. Time step = 1
> fs. Visual inspection of the trajectory shows a significant amount of
> kinetic energy is being transfered to the rotation of the molecule. Is
> there a way to avoid this, besides probably decreasing the time step?
> I did not find anything related to this in the manual or an earlier
> post.
> I have set KeepStationary = Yes, but this only affects the
> translation, as I understand.
>
> Many thanks in advance and best wishes,
>
> Manuel
> _______________________________________________
> DFTB-Plus-User mailing list
> DFTB-Plus-User at dftb-plus.info<mailto:DFTB-Plus-User at dftb-plus.info>
> http://www.dftb-plus.info/mailman/listinfo/dftb-plus-user
>
> --
> Dr. B. Hourahine, SUPA, Department of Physics,
> University of Strathclyde, John Anderson Building,
> 107 Rottenrow, Glasgow G4 0NG, UK.
> +44 141 548 2325,
>benjamin.hourahine at strath.ac.uk<mailto:benjamin.hourahine at strath.ac.uk>
> The University of Strathclyde is a charitable body,
> registered in Scotland, with registration number SC015263
>
>
>
> ------------------------------
>
> Message: 2
> Date: Fri, 19 Feb 2010 13:37:33 +0100
>From: "Ran Friedman, Biochemisches Inst."
><r.friedman at bioc.uzh.ch<mailto:r.friedman at bioc.uzh.ch>>
> Subject: Re: [DFTB-Plus-User] Rotational DOF in gas phase MD
> To:
>benjamin.hourahine at strath.ac.uk<mailto:benjamin.hourahine at strath.ac.uk>,
>User list for DFTB+ related
> questions
><dftb-plus-user at dftb-plus.info<mailto:dftb-plus-user at dftb-plus.info>>
> Message-ID:
><web-22135993 at idmailbe1b.unizh.ch<mailto:web-22135993 at idmailbe1b.unizh.ch>>
> Content-Type: text/plain;charset=iso-8859-1;format="flowed"
>
> Hi Ben and list readers,
>
> A pretty recent thermostat by Bussi and Parrinello is based on the
>Berendsen
> thermostat but seems to alleviate the artefacts associated with it:
> http://link.aip.org/link/?JCPSA6/126/014101/1
>
> You may want to consider applying it in DFTB.
>
> Ran
>
> On Fri, 19 Feb 2010 12:29:08 +0000
> Ben Hourahine <benjamin.hourahine at strath.ac.uk> wrote:
>
> Hello Manuel,
>
> which thermostat are you using? This sounds like the known artefacts
> of the Berendsen thermostat (flying icecube effect).
>
> At the moment, as you've observed, only the translational motion is
> removed in the code, but with Berendsen the lowest accessible modes
> always end up accumulating kinetic energy. It would be possible to
> remove the rotational degrees of freedom as well for non-periodic
> calculations, but we haven't done this. If rotation were also removed
> then using Berendsen thermostating would instead lead to the low energy
> collective bend modes of the molecule heating up.
>
> Would the alternative choice of the Andersen thermostat any use? This
> thermostat does not have the same pathology (but has other problems - it
> prevents collection of some types of autocorrelation data and suppresses
> atomic diffusion, but does correctly sample from the NVT ensemble).
>
> As this molecule has bonds between C and H, I'd suggest testing a
> shorter time step is probably a good idea anyway, irrespective of
> any thermostatting choices.
>
> Regards
>
> Ben
>
> Manuel D?mer wrote:
> Dear DFTB users,
>
> I have been running an NVT MD of a peptide in gas phase. Time step = 1
> fs. Visual inspection of the trajectory shows a significant amount of
> kinetic energy is being transfered to the rotation of the molecule. Is
> there a way to avoid this, besides probably decreasing the time step?
> I did not find anything related to this in the manual or an earlier
> post.
> I have set KeepStationary = Yes, but this only affects the
> translation, as I understand.
>
> Many thanks in advance and best wishes,
>
> Manuel
> _______________________________________________
> DFTB-Plus-User mailing list
> DFTB-Plus-User at dftb-plus.info
> http://www.dftb-plus.info/mailman/listinfo/dftb-plus-user
>
> --
> Dr. B. Hourahine, SUPA, Department of Physics,
> University of Strathclyde, John Anderson Building,
> 107 Rottenrow, Glasgow G4 0NG, UK.
> +44 141 548 2325, benjamin.hourahine at strath.ac.uk
> The University of Strathclyde is a charitable body,
> registered in Scotland, with registration number SC015263
>
> _______________________________________________
> DFTB-Plus-User mailing list
> DFTB-Plus-User at dftb-plus.info
> http://www.dftb-plus.info/mailman/listinfo/dftb-plus-user
>
>
>
>
>
> ------------------------------
>
> _______________________________________________
> DFTB-Plus-User mailing list
> DFTB-Plus-User at dftb-plus.info
> http://www.dftb-plus.info/mailman/listinfo/dftb-plus-user
>
>
> End of DFTB-Plus-User Digest, Vol 26, Issue 4
> *********************************************
>
>
> ________________________________
> Accelrys Limited (http://accelrys.com)
> Registered office: 334 Cambridge Science Park, Cambridge, CB4 0WN, UK
> Registered in England: 2326316
More information about the DFTB-Plus-User
mailing list