[QE-users] how to obtain random initial velocities with CP

[Giuseppe Mattioli]

Hi Paolo and Eduardo
I would like to add some consideration, because I have a similar  
problem. I'm going to simulate the subtraction of one electron from a  
thermalized system. To do this, in previous cp.x versions (up to  
4.3.2, AFAIK), it was possible to: 1) run a cp.x dynamics  
(electrons+ions) with tot_charge=0.0; 2) minimize the electrons with  
tot_charge=+1.0 using the last geometry saved in the previous run; 3a)  
restart a new cp.x dynamics (electrons+ions) by using wavefunctions  
optimized in step 2) and the last velocities saved in step 1)  
(ion_velocities='from_input'), and run a few steps with a short dt to  
assess the equilibration of the dynamics; 3b) restart again  
(ion_velocities='change_step') by using the same dt as in 1).

However, it is apparently not possible in recent versions, as the cp.x  
manual states that

                    'from_input'  : restart the simulation with atomic  
velocities read
                                    from standard input - see card  
'ATOMIC_VELOCITIES'
                                    BEWARE: works only if  
restart_mode='from_scratch',
                                    tested only with electrons_dynamics='cg'

As long as you know what you are doing, the simulation of oxidation or  
reduction processes in thermalized systems was a desirable feature of  
cp.x...
Best
Giuseppe


Quoting Paolo Giannozzi <p.giannozzi at gmail.com>:

> Hi Eduardo
>
> apparently ions_velocities="random" does not set random initial velocities,
> but  velocity rescaling. Not sure why. In any event: I am not sure that
> starting the CP dynamics from ground-state positions and electrons and
> nonzero initial velocities is a good idea.
>
> Paolo
>
> On Sun, Oct 14, 2018 at 8:21 PM Eduardo Menendez <eariel99 at gmail.com> wrote:
>
>> Hi,
>> I am unable to start a MD with CP with random velocities setting an
>> initial temperature, using QE 6.3. Reading the input description I
>> understood that I could start it using              ion_velocities =
>> 'random', tempw= starting T. However, I always start with 0 temperature.
>>
>> This is my output
>> &CONTROL
>>                  calculation = 'cp' ,
>>                 restart_mode = 'restart' ,
>>                   pseudo_dir = './' ,
>>                         outdir='./tmp',
>>                         wfcdir='./tmp',
>>                         prefix='mapi',
>>                         nstep = 500,
>>                            dt = 4.13413733424,
>>             ndr = 51,
>>             ndw = 51,
>>             iprint =50,
>>             isave = 50,
>> /
>> &SYSTEM
>>                          ibrav=0,
>>                        celldm(1)=1.889726132885 ,
>>                           nat=768,
>>                         ntyp=5,
>>                         ecutwfc=35.0  ! 50,
>>                      ecutrho=200.0 ,  ! 180,
>>                     input_dft = 'pbesol',
>>               nr1b=21, nr2b=21, nr3b=21,  ! nr1b=(2*r_cut)/(Lx)*nr1
>>                           nbnd = 1700,
>> /
>> &ELECTRONS
>>                         conv_thr = 1.0e-5,
>> !  options for CP
>>             emass = 400.d0, ! default, Mosconi et al,PCCP 16,16137 (2014),
>> doi: 10.1039/C4CP00569D
>>             emass_cutoff = 2.5d0, ! default
>>             electron_dynamics = 'verlet' ,
>>
>> /
>>  &IONS
>>                 ion_dynamics = 'verlet' , !  'verlet' ,
>> !             ion_temperature = 'not_controlled',
>>              ion_velocities = 'random',
>>                        tempw = 600.0 ,
>>  /
>> &CELL
>>                 cell_dynamics = 'none'
>> /
>> In the output I read
>>    Ionic position will be re-read from restart file
>>    All atoms are allowed to move
>>    Ionic temperature control via canonical velocities rescaling :
>>    temperature required =  600.00000K, tolerance =  100.00000K
>> ......
>>   nfi     ekinc              temph  tempp     etot
>> enthal               econs                econt              vnhh
>> xnhh0   vnhp    xnhp0
>>       5    0.000002129555068    0.0    0.00  -12738.991035174058
>> -12738.991035174058  -12738.991033373713  -12738.991031244159   0.0000
>> 0.0000   0.0000   0.0000
>>       6    0.000005713182507    0.0    0.00  -12738.991057915813
>> -12738.991057915813  -12738.991042309621  -12738.991036596439   0.0000
>> 0.0000   0.0000   0.0000
>>
>> nfi begins at 5 because there were 4 previous steepest descent dynamics
>> (ould I reset the nfi counter? ), which got the wfc from a PWSCF
>> self-consistent run.
>>
>> Thanks,
>>
>> Eduardo Menendez Proupin
>> University of Chile
>> www.gnm.cl/emenendez
>> _______________________________________________
>> users mailing list
>> users at lists.quantum-espresso.org
>> https://lists.quantum-espresso.org/mailman/listinfo/users
>
>
>
> --
> Paolo Giannozzi, Dip. Scienze Matematiche Informatiche e Fisiche,
> Univ. Udine, via delle Scienze 208, 33100 Udine, Italy
> Phone +39-0432-558216, fax +39-0432-558222



GIUSEPPE MATTIOLI
CNR - ISTITUTO DI STRUTTURA DELLA MATERIA
Via Salaria Km 29,300 - C.P. 10
I-00015 - Monterotondo Scalo (RM)
Mob (*preferred*) +39 373 7305625
Tel + 39 06 90672342 - Fax +39 06 90672316
E-mail: <giuseppe.mattioli at ism.cnr.it>