barostat problem NPT

470 views
Skip to first unread message

simin pahlavi

unread,
Jul 6, 2018, 10:51:37 AM7/6/18
to cp2k
Dear Cp2k usrs,
I am new in CP2K and I am trying to simulate 190 water molecule in T=800 k and P=4500 bar by the means of AIMD.
I have a question about the barostat in CP2k. According the previous articles I have used CSVR barostat with timecon 2000 fs.
As I have attached the plot of instantaneous pressure, the fluctuation in pressure are about -4500 to 4500.
I have tried to change the timecon to the smaller value after 4000 fs relaxation to 1000 but the fluctuation did not change that much.
I appreciate it if you recommend me a convenient barostat and thermostat for the system in hand.
Thanks in advance 
simin
pressure.png
water.inp

Matt W

unread,
Jul 6, 2018, 3:06:21 PM7/6/18
to cp2k
I don't really see a problem.

I'd suggest running a classical simulation (should be water setups in the tests) to get an idea of equilibriation of an NPT system, and what the fluctuations would look like.

Matt

Yingchun Zhang

unread,
Jul 8, 2018, 1:14:05 AM7/8/18
to cp...@googlegroups.com
Hi simin,

I usually don't run AIMD under NPT ensemble. But I remember another thermostat should be added into the barostat section. And I recommend use CELL_REF with slightly larger dimension, say 15 A for your case in a NPT simulation.

BR,
Yingchun

Mohammad Hellani

unread,
Feb 7, 2019, 9:30:37 AM2/7/19
to cp2k
Hello,
Can you please tell me how you have calculated the pressure?
i Added in the print of Motion section STRESS_TENSOR but it is writing 1/3 Trace of Stress Tensor.
is 1/3 Trace of STress Tensor the same as Pressure?
Thanks for your help.
Mohammad

Thomas Kühne

unread,
Feb 7, 2019, 9:57:05 AM2/7/19
to cp...@googlegroups.com
Dear Simin, 

due to the fact that the fluctuations are proportional to the square root of the number of 
particles your pressure as a function of time seems to me very reasonable. I hence, 
would only suggest to substantially reduce the timecon to a much lower value, e.g. 
50 - 100fs. Also, please keep in mind that the region in the thermostat section is by 
default global, which to the best of my knowledge, only couples to 3 degrees of freedom 
of the whole system! So better use massive to equilibrate your system. 

Best, 
Thomas

--
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 https://groups.google.com/group/cp2k.
For more options, visit https://groups.google.com/d/optout.
<pressure.png><water.inp>

==============================
Prof. Dr. Thomas D. Kühne
Dynamics of Condensed Matter
Chair of Theoretical Chemistry
University of Paderborn
Warburger Str. 100
D-33098 Paderborn
Germany

Reply all
Reply to author
Forward
0 new messages