CP2K BUCKMORSE
elephant gogogo!
Krack Matthias
Hi
Q1: That’s most likely a typo in the paper, i.e. the exponent in the denominator is missing.
Q2: No need to do so, if my reply to Q1 is right.
Q3: Usually, FF summands have a “mol” unit with integer exponents. Most likely, you have a product of two atomic parameters in a term which causes a mol^(1/2) unit for that parameter.
HTH
Matthias
From: cp...@googlegroups.com <cp...@googlegroups.com> on behalf of elephant gogogo! <qumingzi...@gmail.com>
Date: Monday, 12 December 2022 at 10:41
To: cp2k <cp...@googlegroups.com>
Subject: [CP2K:18175] CP2K BUCKMORSE
Dear CP2K users,
Question 1:
recently, I have found such potential in the literature:
It's very similar to the BUCKMORSE potential in the cp2k. But the term in the red circle is a little different,
This term is dipole induced dipole dispersion potential based on the van der Waals interactions. So how could I deal with it?
Question 2:
I have use the GENPOT to write it by myself, but it also shows a mistake, please find it in the attachment.
Question 3:
How could I set such unit mol^1/2, in the cp2k, it shows it must be a integer.
Thank you very much!
Best Regards
Y
--
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 view this discussion on the web visit
https://groups.google.com/d/msgid/cp2k/1454b0fc-13e9-4ed7-a6b3-6db997aa7359n%40googlegroups.com.
elephant gogogo!
To view this discussion on the web visit https://groups.google.com/d/msgid/cp2k/ZRAP278MB08278AB35E64E3E6A96AB676F4E29%40ZRAP278MB0827.CHEP278.PROD.OUTLOOK.COM.
Krack Matthias
Yes, if there is a c value given for each atomic kind, just multiply the corresponding values for each pair interaction.
To view this discussion on the web visit https://groups.google.com/d/msgid/cp2k/CAEB9i%3DV4QL02xw-7Q-PcwnSycd7WJbtwDSFLtpL%2BaPraTgh%3D%3DQ%40mail.gmail.com.
elephant gogogo!
Krack Matthias
It means that your input is still wrong.
I suggest that you provide the reference for the CeO2 force field for which you want to create a CP2K input file and your current input file.
From: cp...@googlegroups.com <cp...@googlegroups.com> on behalf of elephant gogogo! <qumingzi...@gmail.com>
Date: Monday, 12 December 2022 at 14:47
To: cp2k <cp...@googlegroups.com>
Subject: Re: [CP2K:18180] CP2K BUCKMORSE
Thank you a lot, it really helps. But I have met another problem by using williams potential, I have found such sentences in LocalLog files, could you please tell me, what is it mean?
To view this discussion on the web visit https://groups.google.com/d/msgid/cp2k/3d25fd65-9a81-4157-8622-cf48929d63fdn%40googlegroups.com.
Krack Matthias
Hi
There are few things which should be corrected in your input:
- Add the CORE_CHARGE -5.38 and +0.83 for Ce and O in the &SHELL sections (see one my previous replies)
- Remove the SCALED keyword in the &COORD section, because you specify real coordinates and not scaled ones
- Use MULTIPLE_UNIT_CELL in &CELL and &TOPOLOGY sections. I suggest at least 4x4x4 cells because of the RCUT 10.4 A which implies a reasonable cell edge length of at least 20.8 A to avoid self interactions. GMAX has to increase correspondingly.
Optionally, I would also
- drop the keywords KEEP_* to allow for a unconstrained relaxation
- LBGFS could be used instead of CG as OPTIMIZER
- drop the &CHARGE sections
HTH
Matthias
From: cp...@googlegroups.com <cp...@googlegroups.com> on behalf of elephant gogogo! <qumingzi...@gmail.com>
Date: Monday, 12 December 2022 at 15:39
To: cp2k <cp...@googlegroups.com>
Subject: Re: [CP2K:18183] CP2K BUCKMORSE
ok, this is the potential reference. And please find the input file in the attachment. My material is ceo2, I just revise the input file from the example CrO2 in the test folder.
To view this discussion on the web visit https://groups.google.com/d/msgid/cp2k/0b4d4a11-7659-4125-8586-7a411cf1bcccn%40googlegroups.com.
elephant gogogo!
Krack Matthias
That’s correct if C is not a core-shell ion.
To view this discussion on the web visit https://groups.google.com/d/msgid/cp2k/e6fc9daf-8141-48c0-bfd8-31d3d6af6289n%40googlegroups.com.
elephant gogogo!
Krack Matthias
Which problem? There is nothing new or special about that force field. Note, that the rho value for O-O is wrong in that table. It should rather be 0.2192 instead of 0.2912 (a typical typo).
From: cp...@googlegroups.com <cp...@googlegroups.com> on behalf of elephant gogogo! <qumingzi...@gmail.com>
Date: Wednesday, 14 December 2022 at 17:52
To: cp2k <cp...@googlegroups.com>
Subject: Re: [CP2K:18202] CP2K BUCKMORSE
I'm so confused. I run the calculation of ceo2, it works. But when I calculate another material, it still has problem. The input is as the attachment. They use the same potential Williams, I just change the atoms and cell information. Could you please tell me? Nomally, where can we find our mistakes? Thank you very much!
To view this discussion on the web visit https://groups.google.com/d/msgid/cp2k/b81ec919-ae6f-4a13-bf82-f46412d42df0n%40googlegroups.com.
elephant gogogo!
Krack Matthias
You specify SCALED atomic coordinates in the &COORD section without the keyword SCALED and thus the default unit Angstrom is used.
To view this discussion on the web visit https://groups.google.com/d/msgid/cp2k/c9d6d879-1277-40f8-a220-d32b1643577en%40googlegroups.com.