<div dir="ltr">Dear Stefano and Paolo,<div style>I think that information from this thread should be placed into the User Manual and FAQ. :)</div></div><div class="gmail_extra"><br><br><div class="gmail_quote">2013/2/27 Ben Palmer <span dir="ltr"><<a href="mailto:benpalmer1983@gmail.com" target="_blank">benpalmer1983@gmail.com</a>></span><br>
<blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">Dear Stefano,<br>
<br>
I have had a go at converging for Aluminium. I wasn't too sure what to<br>
do with the K-points, but I've had a go anyway. This is what I have<br>
done, step by step, with some results from the calculations.<br>
<br>
First I set k-points to 2, smearing as MV and a width of 0.02 for the<br>
energy convergence. I varied the energy from 10 to 50 (taking 50 as the<br>
'true' value), and selected the first within 1mRy of the 'true' energy.<br>
<br>
113.2226: K points: 2, ecut: 10, energy: -153.984951, time: 4.05s<br>
113.2237: K points: 2, ecut: 15, energy: -153.99949084, time: 4.31s<br>
113.2251: K points: 2, ecut: 20, energy: -154.00496366, time: 4.96s<br>
113.2261: K points: 2, ecut: 25, energy: -154.00841338, time: 5.49s<br>
113.2268: K points: 2, ecut: 30, energy: -154.00937872, time: 6.52s<br>
113.2283: K points: 2, ecut: 35, energy: -154.01004897, time: 8.31s<br>
113.2292: K points: 2, ecut: 40, energy: -154.01057988, time: 8.17s<br>
113.2306: K points: 2, ecut: 45, energy: -154.01066204, time: 10.91s<br>
113.2314: K points: 2, ecut: 50, energy: -154.01083456, time: 12.83s<br>
113.2315: Converged energy cutoff ecutwfc: 35<br>
<br>
I then lowered ecutrho until, and selected the lowest value that fell<br>
within 1mRy of the 'true' energy.<br>
<br>
218.6003: K points: 2, ecutwfc: 35, ecutrho: 252, energy:<br>
-154.01105461, time: 10.98s<br>
218.6014: K points: 2, ecutwfc: 35, ecutrho: 224, energy:<br>
-154.01104417, time: 11.67s<br>
218.6031: K points: 2, ecutwfc: 35, ecutrho: 196, energy:<br>
-154.01069558, time: 8.82s<br>
218.604: K points: 2, ecutwfc: 35, ecutrho: 168, energy:<br>
-154.01054112, time: 9.99s<br>
218.6052: K points: 2, ecutwfc: 35, ecutrho: 140, energy:<br>
-154.01004897, time: 8.46s<br>
218.6064: K points: 2, ecutwfc: 35, ecutrho: 112, energy:<br>
-154.00962358, time: 7.22s<br>
218.6076: K points: 2, ecutwfc: 35, ecutrho: 84, energy:<br>
-154.00782952, time: 5.80s<br>
218.6082: Converged energy cutoff ecutrho: 140<br>
<br>
At this point, I've got ecutwfc = 35 and ecutrho = 140, but I wasn't too<br>
sure how to progress, so I attempted the following. I set a large<br>
number of k-points, 24x24x24, with a narrow smearing of 0.005. I used<br>
the energy cutoffs to then calculate a new reference energy for convergence.<br>
<br>
I increased the smear width and decreased the k-points in quite<br>
arbitrary combinations, and looked for the combination that executed<br>
fastest, while keeping within 1mRy of the new reference energy.<br>
<br>
9956.6004: K-point: 24, Degauss: 0.005, energy: -154.06230709, time:<br>
25m21.81s<br>
9956.6064: K-point: 24, Degauss: 0.01, energy: -154.06231018, time:<br>
27m31.26s<br>
9956.6122: K-point: 24, Degauss: 0.015, energy: -154.06234325, time:<br>
27m28.22s<br>
9956.6191: K-point: 16, Degauss: 0.015, energy: -154.06218801, time:<br>
6m40.66s<br>
9956.6243: K-point: 16, Degauss: 0.02, energy: -154.0622109, time:<br>
6m41.03s<br>
9956.6358: K-point: 12, Degauss: 0.01, energy: -154.06191016, time:<br>
3m47.69s<br>
9956.6473: K-point: 12, Degauss: 0.02, energy: -154.06227228, time:<br>
3m59.45s<br>
9956.672: K-point: 8, Degauss: 0.02, energy: -154.06351667, time: 1m 8.18s<br>
9956.6735: K-point: 8, Degauss: 0.03, energy: -154.06335682, time:<br>
1m20.19s<br>
9956.6754: K-point: 6, Degauss: 0.02, energy: -154.05986685, time:<br>
0m46.42s<br>
9956.6765: K-point: 6, Degauss: 0.03, energy: -154.05968822, time:<br>
0m46.50s<br>
<br>
From this, I'd choose K-points 12x12x12 and smearing width 0.01 or 0.02.<br>
<br>
My final convergence settings were:<br>
<br>
ecutwfc = 35,<br>
ecutrho = 140,<br>
k points 12x12x12<br>
smearing mv 0.01<br>
<br>
Would this be an acceptable way to chose the settings, or could I speed<br>
up the end part?<br>
<br>
All the best,<br>
<br>
Ben Palmer, Student @ University of Birmingham<br>
<div class="HOEnZb"><div class="h5"><br>
> Dear All,<br>
> My previous post was actually more intended as an answer to Ben<br>
> Palmer question than a comment to<br>
> Ali Kachmar contribution. Sorry.<br>
> best regards,<br>
> stefano<br>
><br>
><br>
> On 02/25/2013 02:58 PM, Stefano de Gironcoli wrote:<br>
>> Dear Ali Kachmar,<br>
>><br>
>> convergence w.r.t. ecutwfc (and ecutrho) and convergence w.r.t.<br>
>> k-points sampling are rather independent issues and can be tested to a<br>
>> large extent separately<br>
>><br>
>> - convergence w.r.t. ecutwfc and ecutrho is a property depending on<br>
>> the highest Fourier components that are needed to describe the<br>
>> wavefunctions and the density of your system. his depends on the<br>
>> pseudopotentials that are present in the calculation and do not depend<br>
>> strongly, for a given set of pseudopotentials, on the particular<br>
>> configuration because it depends mostly on the behaviour of the wfc in<br>
>> the core region which is quite insensitive (in terms of shape) on the<br>
>> environment.<br>
>> So each pseudopotential has a required cutoff. An upperbound to this<br>
>> value can be determined from any system that contains that pseudo.<br>
>> The cutoff needed for a system containing several species is the<br>
>> highest among those needed for each element.<br>
>> Moreover, in US pseudo or PAW the charge density has contributions<br>
>> from localized terms that may (an usually do in USPP) require quite<br>
>> higher cutoff than the one needed for psi**2 (4*ecutwfc) ... hence the<br>
>> possibility to vary and test independently for ecutrho ...<br>
>><br>
>> My recommended strategy to fix ecutwfc and ecutrho is to perform total<br>
>> energy (and possibly, force and stress) covergence test increasing<br>
>> ecutwfc keeping ecutrho at its default vaule (=4*ecutwfc) until<br>
>> satisfactory stability is reached (typically ~1 mry/atom in the<br>
>> energy, 1.d-4 ry/au in the forces, a fraction of a KBar in the stress)<br>
>> ... this fixes the converged value of ecutrho to 4 times the<br>
>> resulting ecutwfc.<br>
>> Now keeping this value for ecutrho one can try to reduce ecutwfc and<br>
>> see how much this can be done without deteriorating the convergence.<br>
>><br>
>> -convergence with respect to k-points is a property of the band<br>
>> structure.<br>
>> I would study it after the ecutwfc/ecutrho issue is settled but some<br>
>> fairly accurate parameters can be obtained even with reasonable but<br>
>> not optimal cutoff parameters.<br>
>><br>
>> There is a big difference between convergence in a band insulator or<br>
>> in a metal.<br>
>><br>
>> In an insulator bands are completely occupied or empty across the BZ<br>
>> and charge density can be written in terms of wannier functions that<br>
>> are exponentially localized in real space.<br>
>> Hence the convergence w.r.t the density of point in the different<br>
>> directions in the BZ should be exponentially fast and anyway quite<br>
>> quick...<br>
>><br>
>> In a metal the need to sample only a portion of the BZ would require<br>
>> an extremely dense set of k points in order to locate accurately the<br>
>> Fermi surface. This induces to introduce a smearing width that smooth<br>
>> the integral to be performed... the larger the smearing width, the<br>
>> smoother the function, and the faster the convergence results...<br>
>> however the larger the smearing width the farther the result is going<br>
>> to be from the accurate, zero smearing width, result that one would<br>
>> desire.<br>
>> Therefore different shapes fro the smearing functions have been<br>
>> proposed to alleviate this problem and<br>
>> Marzari-Vanderbilt and Methfessel-Paxton smearing functions give a<br>
>> quite mild dependence of the (k-point converged) total energy as a<br>
>> function of the smearing width thus being good choices for metals.<br>
>><br>
>> My recommended strategy for fix the k-point sampling in metals is<br>
>> 1) chose the smearing function type (mv or mp, recomended)<br>
>> 2) for decreasing values of the smearing width (let's say from an high<br>
>> value of 0.1 ry = 1.36 eV to a low value of 0.01 - 0.005 ry =<br>
>> 0.136-0.068 eV if feasable) CONVERGE the total energy w.r.t to<br>
>> smearing well within the global desired tolerance (of 1 mry/atom, for<br>
>> instance)<br>
>> 3) by examining the behaviour of the CONVERGED Energy vs smearing<br>
>> width curve E(sigma) identify the smearing width for which E(sigma) is<br>
>> within tolerance w.r.t. E(sigma==0) keeping in mind that for<br>
>> methfessel-paxton E(sigma) ~ E(0) + A*sigma**4 + o(sigma**6) while for<br>
>> marzari-vanderbilt the dependence is more likely E(sigma) ~ E(0)<br>
>> +A*sigma**3 + o(sigma**4).<br>
>> 4) select that value of the smearing width and the smallest set of<br>
>> k-points for which this is converged.<br>
>><br>
>> HTH<br>
>><br>
>> stefano<br>
>><br>
>><br>
>><br>
>> On 02/24/2013 06:54 PM, Ali KACHMAR wrote:<br>
>>> Hi,<br>
>>><br>
>>> as far as I know, there is no any techinques for choosing ecut and<br>
>>> k-points. Please have a look at the pwscf archive and make up a<br>
>>> conclusion.<br>
>>><br>
>>> Best,<br>
>>> Ali<br>
>>><br>
>>>> Date: Sat, 23 Feb 2013 19:55:51 +0000<br>
>>>> <a href="mailto:From%3Abenpalmer1983@gmail.com">From:benpalmer1983@gmail.com</a><br>
>>>> <a href="mailto:To%3Apw_forum@pwscf.org">To:pw_forum@pwscf.org</a><br>
>>>> Subject: [Pw_forum] Technique for converging Ecut and K-points?<br>
>>>><br>
>>>> Hi everyone,<br>
>>>><br>
>>>> I just wanted to ask if users have any techniques for choosing ecut and<br>
>>>> k-points? I've read that one way would be to start with a high number<br>
>>>> of k-points and high energy cutoff, and use that energy as an almost<br>
>>>> true value. Then adjust k-points and energy cutoff from a lower<br>
>>>> number/cutoff until it converges to the true value. Would you try to<br>
>>>> converge energy cutoff first, or k-points? Does it matter which you<br>
>>>> converge first?<br>
>>>><br>
>>>> Thanks<br>
>>>><br>
>>>> Ben Palmer<br>
>>>> Student @ University of Birmingham<br>
>>>> _______________________________________________<br>
>>>> Pw_forum mailing list<br>
>>>> <a href="mailto:Pw_forum@pwscf.org">Pw_forum@pwscf.org</a><br>
>>>> <a href="http://pwscf.org/mailman/listinfo/pw_forum" target="_blank">http://pwscf.org/mailman/listinfo/pw_forum</a><br>
>>><br>
>>><br>
>>> _______________________________________________<br>
>>> Pw_forum mailing list<br>
>>> <a href="mailto:Pw_forum@pwscf.org">Pw_forum@pwscf.org</a><br>
>>> <a href="http://pwscf.org/mailman/listinfo/pw_forum" target="_blank">http://pwscf.org/mailman/listinfo/pw_forum</a><br>
>><br>
> _______________________________________________<br>
> Pw_forum mailing list<br>
> <a href="mailto:Pw_forum@pwscf.org">Pw_forum@pwscf.org</a><br>
> <a href="http://pwscf.org/mailman/listinfo/pw_forum" target="_blank">http://pwscf.org/mailman/listinfo/pw_forum</a><br>
<br>
_______________________________________________<br>
Pw_forum mailing list<br>
<a href="mailto:Pw_forum@pwscf.org">Pw_forum@pwscf.org</a><br>
<a href="http://pwscf.org/mailman/listinfo/pw_forum" target="_blank">http://pwscf.org/mailman/listinfo/pw_forum</a><br>
</div></div></blockquote></div><br><br clear="all"><div><br></div>-- <br>С уважением,<br>Дмитрий Коротин<br><br>Best regards,<br>Dr. Dmitry Korotin<br><br>Institute of Metal Physics<br>S. Kovalevskaya, 18<br>620990 Yekaterinburg<br>
Russia
</div>