[Wien] Error related to nlvdw calculations

Peter Blaha pblaha at theochem.tuwien.ac.at
Wed Sep 13 10:33:16 CEST 2017


In addition to what Fabien Tran said:

If you expand the density/potential up to l=10, it makes only sense if 
you also increase the LVNSmax value in case.in1 from 4 to 6 (or 8), 
otherwise these potential components will not even be used in the 
calculations.

Please remember in general: when you have LARGE SPHERES (gt.~2.2) and 
large nonsphericity, an increase of this value could be very important 
for "highest" precision. The effect can in some cases be "larger" than 
increasing RKmax from 8 to 10 !!

For these reasons, init_lapw gives you now a "warning" or a "hint" about 
this, when it detects large spheres. Take it seriously when aiming for 
highest accuracy. To be able to make this automatically, there is a new 
parameter for init_lapw in batch mode:  -lvns X

Peter Blaha

On 09/13/2017 10:10 AM, tran at theochem.tuwien.ac.at wrote:
> Hi,
>
> Maybe it's because you are using lmax=10 in case.in2.
> Does the problem persist if you use the default lmax=6 in case.in2?
> Anyway, lmax=6 is already a very good value, so why are you using
> lmax=10?
>
> FT
>
> On Wednesday 2017-09-13 10:02, Jiawei Zhang wrote:
>
>> Date: Wed, 13 Sep 2017 10:02:46
>> From: Jiawei Zhang <jiaweizhang at chem.au.dk>
>> Reply-To: A Mailing list for WIEN2k users
>> <wien at zeus.theochem.tuwien.ac.at>
>> To: "Wien at zeus.theochem.tuwien.ac.at" <Wien at zeus.theochem.tuwien.ac.at>
>> Subject: [Wien] Error related to nlvdw calculations
>>
>> Dear WIEN2k developers and users,
>> I am trying to do a calculation on some layered materials by including
>> the nlvdw correction into SCF calculation in Wien2k 17.1 running at
>> Linux system (Intel
>> E5-2680, CentOS 7 Linux 64 bit) with intel compilers. The calculations
>> are successful without the calculation of the potential; however, I
>> always get errors when
>> I switch on the potential calculations. In the beginning, when I set
>> "calculation of the potential (T or F)" to "T" in case.innlvdw, I
>> always get the error
>> “Insufficient virtual memory”. According to the suggestion from the
>> previous post, in the line 41 in SRC_nlvdw/calc_vvdw.F, add : lmtmp =0
>> and recompile a
>> selected package : nlvdw. The error “Insufficient virtual memory” is
>> solved. But then I always get the following errors:
>> In case.dayfile,    start (Wed Sep 13 00:35:04 CEST 2017) with nlvdw
>> (200/99 to go)
>>>   nlvdw  > & /dev/null (00:35:04)
>>>   nlvdw (00:36:17) 82.380u 0.204s 1:22.67 99.8% 0+0k 22384+1424io 3pf+0w
>>     cycle 1 (Wed Sep 13 00:37:40 CEST 2017) (200/99 to go)
>>>   lapw0 -nlvdw (00:37:40) 0.925u 0.063s 0:01.06 92.4% 0+0k 6432+56io
>>> 8pf+0w
>> error: command   /home/jiawei/src/WIEN2k_17.1/lapw0 lapw0.def   failed
>>>   stop error
>> Error details:
>> NLVDW END
>> forrtl: severe (64): input conversion error, unit 66, file /home
>> /work/WIEN2k/WIEN2k_17.1/SnSe/exp/PBE_rVV10/include_potential/SnSe/SnSe.r2v_nlvdw
>>
>> Image              PC                Routine            Line
>>  Source
>> lapw0              000000000051E423  Unknown               Unknown
>>  Unknown
>> lapw0              000000000054F798  Unknown               Unknown
>>  Unknown
>> lapw0              000000000054C7EA  Unknown               Unknown
>>  Unknown
>> lapw0              00000000004444C5  MAIN__                   2102
>>  lapw0.F
>> lapw0              00000000004054AE  Unknown               Unknown
>>  Unknown
>> libc.so.6          00000039B041ECDD  Unknown               Unknown
>>  Unknown
>> lapw0              00000000004053B9  Unknown               Unknown
>>  Unknown
>>
>> For the calculations, I set kpoints=605, RKmax=10, and expand lmax to
>> 10 in case.in2.
>> I have tried running in both single mode or parallel mode and I get
>> the same error. Does anyone have any suggestion on how to fix the
>> problem. Thanks in advance!
>> Best regards,
>> Jiawei​
>>
>>
>>
>>
>>
>>
>>>
>
> _______________________________________________
> Wien mailing list
> Wien at zeus.theochem.tuwien.ac.at
> http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien
> SEARCH the MAILING-LIST at:  http://www.mail-archive.com/wien@zeus.theochem.tuwien.ac.at/index.html
>

-- 

                                       P.Blaha
--------------------------------------------------------------------------
Peter BLAHA, Inst.f. Materials Chemistry, TU Vienna, A-1060 Vienna
Phone: +43-1-58801-165300             FAX: +43-1-58801-165982
Email: blaha at theochem.tuwien.ac.at    WIEN2k: http://www.wien2k.at
WWW:   http://www.imc.tuwien.ac.at/TC_Blaha
--------------------------------------------------------------------------


More information about the Wien mailing list