[QE-users] thermo_pw crashes with "can't open a connected unit"

[Vahid Askarpour]

The error seems to be related to lack of bfgs ionic convergence for each structure. Several structures goes through 20 steps but do not reach convergence. The remaining ones go through 0 bfgs steps. 

So I changed nstep in set_thermo_work_todo.f90 from 20 to 50. Now what happens is that the first structure converges in say 25 steps but the other structures  go through 0 bfgs steps. They remain unconverged even though for each of them nstep should be 50. 

Vahid



> On Dec 13, 2018, at 3:23 PM, Vahid Askarpour <vh261281 at dal.ca> wrote:
> 
> This is likely a bug in the thermo_pw code. When I use “lmurn=.TRUE.”, the error disappears. 
> 
> However, I need to use “lmurn=.FALSE.” in order to change all the orthorhombic lattice constants and not just celldm(1).
> 
> Vahid
> 
>> On Dec 13, 2018, at 11:19 AM, Vahid Askarpour <vh261281 at dal.ca> wrote:
>> 
>> Dear QE community,
>> 
>> I have installed the latest thermo_pw along with qe-6.3 and successfully finished QHA calculation for example09 for Si. However, when I redo the calculations for SnSe, I get the following error just as the code starts to do phonon calculations:
>> 
>> Electric Fields Calculation
>> 
>> iter #   1 total cpu time :    11.2 secs   av.it.:   5.8
>> thresh= 1.000E-02 alpha_mix =  0.700 |ddv_scf|^2 =  2.283E+10
>> 
>> %%%%%%%%%%%%%%%%%%
>>    Error in routine seqopn (99):
>>    can't open a connected unit
>> %%%%%%%%%%%%%%%%%%
>> 
>> The error originates from line 430 of io_files.f90 in the Modules folder:
>> 
>> inquire (unit = unit, opened = opnd)
>> if (opnd) call errore ('seqopn', "can't open a connected unit",  abs (unit) )
>> 
>> It appears that a file should not be open but it is.
>> 
>> My input thermo_pw file is:
>> 
>> &INPUT_THERMO
>> what='mur_lc_t',
>> lmurn=.FALSE.
>> deltat=3.
>> /
>> 
>> My other two QE input files are:
>> 
>> &control
>>  calculation = 'scf'
>>  prefix='tinsel',
>>  pseudo_dir = './',
>>  outdir='./tempdir'
>> /
>> &system
>>   ibrav=8,
>>   nat=8 , ntyp= 2, 
>>   celldm(1)=21.34350847, celldm(2)=0.3628777, celldm(3)=0.38586894
>>   ecutwfc = 90, 
>>   occupations='fixed'
>>   nr1=150,nr2=50,nr3=50
>> /
>> &electrons
>>   conv_thr =  1.0d-5
>> /
>> ATOMIC_SPECIES
>> Sn 118.71 Sn.pz-hgh.UPF
>> Se 78.96  Se.pz-hgh.UPF
>> ATOMIC_POSITIONS (crystal)
>> Sn       0.882841736   0.250000000   0.602195343
>> Sn       0.617158264   0.750000000   0.102195343
>> Sn       0.382841736   0.250000000   0.897804657
>> Sn       0.117158264   0.750000000   0.397804657
>> Se       0.142813748   0.250000000   0.976427467
>> Se       0.357186252   0.750000000   0.476427467
>> Se       0.642813748   0.250000000   0.523572533
>> Se       0.857186252   0.750000000   0.023572533
>> K_POINTS AUTOMATIC
>> 2 2 2 0 0 0
>> 
>> And for phonon calculation:
>> 
>> SnSe-Pnma
>> &inputph
>> tr2_ph=1.0d-9,
>> prefix='tinsel',
>> fildyn='SnSe.dyn.xml',
>> ldisp=.TRUE.
>> epsil=.TRUE.
>> nq1=2, nq2=2, nq3=2,
>> /
>> 
>> If you have encountered this error (or possible bug), I would really appreciate hearing from you. 
>> 
>> Thank you,
>> 
>> Vahid
>> 
>> Vahid Askarpour
>> Department of Physics and Atmospheric Science
>> Dalhousie University,
>> Halifax, NS, Canada
>> 
>> _______________________________________________
>> users mailing list
>> users at lists.quantum-espresso.org
>> https://lists.quantum-espresso.org/mailman/listinfo/users
> 
> _______________________________________________
> users mailing list
> users at lists.quantum-espresso.org
> https://lists.quantum-espresso.org/mailman/listinfo/users