Re: [CP2K:5392] How can I get the temperature converged in MD calculation (NVT) for system having H atoms?

[Ari Paavo Seitsonen]

Dear Jingyun Ye,

  If you start from a more or less optimised structure about half of the initial kinetic energy first gets converted into an increase of potential energy, and the thermostat is a bit slower: Your time constant now is 100 fs, but this does not mean that ionic temperature would reach this temperature within the time constant (small value means faster equlibration, at least in principle; if you use a value that does not couple to the system - too small for example - then there will not occur heat transfer). Rather it is a "characteristic frequency". If one needs to equilibrate the initial part faster, one can try with the CSVR thermostat and a smaller time constant, and later either increase the time constant or then shift to the Nosé-Hoover thermostat chain.

  By the way, 0.1 fs is quite small a time step - do you need such a small value? Well, it should certainly be a safe choice for any kind of system.

    Greetings,

       apsi

2014-06-11 22:15 GMT+02:00 Jingyun Ye <jingyu...@gmail.com>:

Hi, all

I am running a MD calculation for searching the most stable structure at room temperature. But the temperature goes from 300K to 150K after 1000 steps and the energy goes up. My question is how can I get the temperature converged? Is there any thing wrong in my input? What is the reasonable TIMESTEP for the system having H atoms in it? Small TIMECON and big TIMECON, which one is more efficient for heat transfer? 

I included the input and results. 

Thanks very much.

&MOTION

 &MD

  ENSEMBLE NVT

  STEPS 10000

  TIMESTEP 0.1

  TEMPERATURE 300

  &THERMOSTAT

   TYPE NOSE

    &NOSE

     TIMECON 100

    &END NOSE

  &END THERMOSTAT

 &END MD

&END MOTION

--
You received this message because you are subscribed to the Google Groups "cp2k" group.
To unsubscribe from this group and stop receiving emails from it, send an email to cp2k+uns...@googlegroups.com.
To post to this group, send email to cp...@googlegroups.com.
Visit this group at http://groups.google.com/group/cp2k.
For more options, visit https://groups.google.com/d/optout.

--

-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-
  Ari Paavo Seitsonen / Ari.P.S...@iki.fi / http://www.iki.fi/~apsi/
  Institut für Chemie der Universität Zürich
  Tel: +41 44 63 55 44 97  /  Mobile: +41 79 71 90 935

[Jingyun Ye]

Dear Ari

I really thank you for your suggestion. I tried TIMESTEP=0.1, and TIMECON=10 or 1 which controls the temperature very well. But do you think I need to test the TIMESTEP by doing NVE calculation first? 

Actually, I tried NVE calculation by using TIMESTEP=1.0, 0.5 or 0.1. Which one do you think more efficient for my system? I can not see the difference. Why does the temperature drop down to 150 K? 

Here is the input:

&MOTION

 &MD

  ENSEMBLE NVT

  STEPS 10000

  TIMESTEP 0.1

  TEMPERATURE 300

&END

&END

The results shows bellow. 

TIMESTEP=1.0

TIMESTEP=0.5

TIMESTEP=0.1

[Jingyun Ye]

Dear Ari

I am sorry that the picture are not well ordered. I have tried to edit it 4 times, but always failed to post. 

On Wednesday, June 11, 2014 5:07:24 PM UTC-4, Ari Paavo Seitsonen wrote:

[Jingyun Ye]

Hi, all

I am running a MD calculation for searching the most stable structure at room temperature. But the temperature goes from 300K to 150K after 1000 steps and the energy goes up. My question is how can I get the temperature converged? Is there any thing wrong in my input? What is the reasonable TIMESTEP for the system having H atoms in it? Small TIMECON and big TIMECON, which one is more efficient for heat transfer? 

I included the input and results. 

Thanks very much.