[CP2K-user] [CP2K:15725] PRESSURE TOLERANCE problems when doing cell opt with strain on 2D slab

Dragon Don shin... at gmail.com
Wed Jul 14 13:19:01 UTC 2021


Dear  Thomas Kühne,
  
  Thanks for your reply. Yes, you are right, This is an anisotropic system 
with different hardness along x and y directions. Do you think it is the 
anisotropy that leads to this issue? If so, what can I do to address it or 
just avoid it.
  Looking forward to your advice.

My  best,
Dragon Don

2021年7月13日 上午6:49:00 (前天) 


在2021年7月14日星期三 UTC+8 上午2:28:26<tkuehne> 写道:

> Dear Dragon Don, 
>
> is this an anisotropic system, e.g. a 2D system, with a different hardness 
> along different directions? 
>
> Best, 
> Thomas Kühne
>
> Am 13.07.2021 um 00:49 schrieb Dragon Don <sh... at gmail.com>:
>
> Dear Thomas Kühne
>
>   I really appreciate your timely help. It's glad to know that the 
> pressure deviation is not that far off but why, I thought 1300 bar(=13 
> MPa)is quite huge.
>   Indeed, I have tested another 400 optimization steps, but the pressure 
> deviation changes slightly, just fluctuating around 1300 bar.
>   It's strange that such convergence issue only happens when the strain is 
> loaded along the y axis (My cell is oblique hexagon as shown below). 
>   And the larger the y strain, the further the pressure deviation reaches, 
> even up to 3000 bar for twice the y strain. 
>   In contrast, the pressure deviation can easily converge to 100 bar for x 
> strain loading. The only difference is there are three cell parameters
>   (a,b,γ) that should be settled in y strain loading, while only b and γ 
> need to be calculated for x strain loading. So, Thomas, what can I do to 
> address this issue. 
>   Looking forward to your advice!
>   
>   
>
> Dragon Don
>
> 在2021年7月12日星期一 UTC+8 下午8:10:32<tkuehne> 写道:
>
>> Dear Dragon Don?, 
>>
>> converging the pressure up to 100 bar is a though call and -1345 bar not 
>> that far off. 
>> If, however, you are not satisfied by the deviation you simply need to 
>> need optimize 
>> for longer. But, be aware that the stress tensor sensitively depends on 
>> your SCF 
>> convergence criterium, so a tight convergence also there is critical! 
>>
>> Best, 
>> Thomas Kühne
>>
>> Am 11.07.2021 um 14:24 schrieb Dragon Don <sh... at gmail.com>:
>>
>> Hi everyone!
>>
>>   Now, I am doing a cell opt with strain by fixing the cell size, while 
>> there is a problem that the pressure tolerance can not converge. Could you 
>> please give me a hand, looking forward for your help!
>>
>>   ps. The pics below are the part of the input and output,respectively.
>>
>>  Don<QQ截图20210711202005.png><QQ截图20210711202109.png>
>>
>> -- 
>> 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 cp... at googlegroups.com.
>> To view this discussion on the web visit 
>> https://groups.google.com/d/msgid/cp2k/f8b86e7f-2203-4f0e-b3e5-dcfb00da0c4dn%40googlegroups.com 
>> <https://groups.google.com/d/msgid/cp2k/f8b86e7f-2203-4f0e-b3e5-dcfb00da0c4dn%40googlegroups.com?utm_medium=email&utm_source=footer>
>> .
>> <QQ截图20210711202005.png><QQ截图20210711202109.png>
>>
>>
>>
>>
>> ==============================
>> Thomas D. Kühne
>> Dynamics of Condensed Matter
>> Chair of Theoretical Chemistry
>> University of Paderborn
>> Warburger Str. 100
>> D-33098 Paderborn
>> Germany
>> tho... at upb.de
>> +49/(0)5251/60-5726 <+49%205251%20605726>
>>
>>
> -- 
> 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 cp... at googlegroups.com.
>
> To view this discussion on the web visit 
> https://groups.google.com/d/msgid/cp2k/25118c4f-b4f6-4358-bcbd-a9e8c7645fdcn%40googlegroups.com 
> <https://groups.google.com/d/msgid/cp2k/25118c4f-b4f6-4358-bcbd-a9e8c7645fdcn%40googlegroups.com?utm_medium=email&utm_source=footer>
> .
>
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.cp2k.org/archives/cp2k-user/attachments/20210714/6296fc04/attachment.htm>


More information about the CP2K-user mailing list