[QE-users] [QE-USER]-restart calculation from md to vc-md/relax and vice versa to maintain zero pressure in NVT MD calculation at fixed temperature using pw.x.

[Jayraj Anadani]

Hello everyone,
I am doing NVT MD calculations using pw.x code. During the simulation, my
pressure was constantly oscillating at a particular value of around 50 GPa.
Many literature review papers* mention this line: "The supercell **volume
was relaxed in such a way to minimize the system's energy." In principle,
during the simulations, one could adjust the supercell volume to make the
external pressure of the inherent **structures equal to zero." I* have
tried a lot of different things to maintain my pressure at zero, but as
soon as I manage to achieve oscillatory zero pressure, some other
parameters seem to shift. For example, I ran "md" at 3000K using the
rescaling method, then restarted and ran "vc-relax" with fixed ion
positions to achieve lower pressure, as shown below.
        " total   stress  (Ry/bohr**3)                   (kbar)     P=
 123.26
          total   stress  (Ry/bohr**3)                   (kbar)     P=
 116.29
          total   stress  (Ry/bohr**3)                   (kbar)     P=
  89.51
          total   stress  (Ry/bohr**3)                   (kbar)     P=
  77.12
          total   stress  (Ry/bohr**3)                   (kbar)     P=
  60.67
          total   stress  (Ry/bohr**3)                   (kbar)     P=
  55.80 "

However, in the same vc-relax output*,** temperature* moved as follows:

"    Ekin =     0.00000000 Ry    T =    0.0 K  Etot =   -13138.95318083
     Ekin =     0.30599294 Ry    T =  253.6 K  Etot =   -13138.94891345
     Ekin =     1.41097581 Ry    T =  714.8 K  Etot =   -13138.96700552
     Ekin =     1.65432345 Ry    T =  878.9 K  Etot =   -13139.19284122
     Ekin =     1.87157146 Ry    T = 1101.3 K  Etot =   -13139.61462781
     Ekin =     1.67673049 Ry    T = 1142.5 K  Etot =   -13140.01208055 "

Is this temperature gain due to electron kinetic energy or why does T begin
at zero? Is there something I am missing? Additionally, I used "vc-md" to
control the temperature and pressure, but after that, I was unable to
restart the file
My exact methodology is:
1.     restart_mode ='from_scratch',----> MD calculation (NVT) at 1000K
with simple rescaling method up to 10 picoseconds
2.    restart_mode = "restart"-----> 'vc-relax' up to pressure convergence
threshold (0.5d0 Kbar) !!! { problem: temperature not maintained shown
above}
My main question is:
Can I restart the calculation from MD to vc-md/relax and again from
vc-md/relax to md ?
And if I run vc-relax or vc-md calculations, my supercell volume fluctuates
slightly, requiring me to switch from an old supercell celldm to a new one
in order to maintain pressure zero..

regards
Jayraj Anadani
Ph.D Scholar, SPU
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.quantum-espresso.org/pipermail/users/attachments/20231120/ad91c470/attachment.html>