STRESS tensor

Axel akoh... at gmail.com
Mon Feb 4 15:21:34 UTC 2008



On Feb 4, 8:47 am, Rad <rad.... at arl.army.mil> wrote:
> Teo,
>
> It is still not printing the pressure/stress tensor. I pulled out a

rad,
please check the files that are created. stress is by default
written to file and not to the regular output.

> copy of cp2k from the cvs respository (not the nightly build) and
> compiled it for the run. When will the nightly build with the updated
> version be available?

why don't you create your own nightly builds? or on-demand builds?

just create a small script and have it executed through cron.
here is what i run (manually) to update and compile on my desktop.
executables for production are then copied over from the exe/
directory (after testing them ok).

#!/bin/sh

PATH=${HOME}/bin:${PATH}
export PATH

cvs update -d
pushd makefiles
make ARCH=Linux-x86-64-intel popt
make ARCH=Linux-x86-64-intel sopt
make ARCH=Linux-x86-64-intel ssgl
make ARCH=Linux-x86-64-g95 sopt
make ARCH=Linux-x86-64-g95 ssgl
make ARCH=Linux-x86-64-g95 sdbg
make ARCH=Linux-x86-64-intel leap2fist
make ARCH=Linux-x86-64-intel fes
popd

test exe/Linux-x86-64-intel/cp2k.popt -nt ChangeLog && perl ./tools/
cvs2cl.pl -b

> Is there a way I can quickly look at the code (for specifice code
> changes) to make sure that I am working with the correct version that

a very useful tool to follow the changes in a cvs archive is 'cvsps'.
it creates "changesets" from cvs commits and you can easily generate
diffs for those changesets. very useful to see what people are doing.

cheers,
   axel.

> prints the tensor in an optimization run?
>
> Thanks
> Rad
>
> On Feb 2, 3:48 pm, Teodoro Laino <teodor... at gmail.com> wrote:
>
> > Ciao Carlo, Rad,
>
> > The printing of STRESS tensor was not enabled in the GEO_OPT module.
> > It is now.. just update and recompile..
>
> > As Rad said you've to specify the "STRESS_TENSOR" keyword and you've to
> > switch on the print_key
>
> > &MOTION
>
> >      &PRINT
> >          &STRESS
> >          &END
> >      &END
> > &END
>
> > Cheers,
> > Teo
>
> > On 2 Feb 2008, at 21:40, Rad wrote:
>
> > > I just ran a geo_opt calculation using a version of cp2k that I
> > > compiled on January 24th and included "STRESS TENSOR"
> > > keyword under force evaulation section. The output didn't have any
> > > stress tenor printed. Here is my input:
>
> > >  &GLOBAL
> > >    BLACS_GRID  SQUARE
> > >    PROGRAM_NAME  CP2K
> > >    PROJECT_NAME CD-NVE
> > > #   TRACE T
> > >    RUN_TYPE GEO_OPT
> > >    PREFERRED_FFT_LIBRARY  FFTW
> > >  &END GLOBAL
> > >  &MOTION
> > >    &GEO_OPT
> > >      OPTIMIZER            CG
> > >      MAX_ITER             200
> > >    &END GEO_OPT
> > >  &END MOTION
> > > &FORCE_EVAL
> > >   STRESS_TENSOR ANALYTICAL
> > >   METHOD Quickstep
> > >   &DFT
> > >     BASIS_SET_FILE_NAME BASIS_MOLOPT
> > >     &MGRID
> > >       NGRIDS 9
> > >       CUTOFF 350
> > >     &END MGRID
> > >     &QS
> > >        MAP_CONSISTENT T
> > >     &END QS
>
> > > Teo - what am I missing in my input?
>
> > > Thanks
> > > Rad
>
> > > On Feb 2, 8:02 am, "carlo antonio pignedoli" <c.pig... at gmail.com>
> > > wrote:
> > >> Dear all,
> > >> is there a way to have the stress tensor in
> > >> a GEO_OPT run for a bulk
> > >> calculation whithin QS DFT?
>
> > >> Ciao
>
> > >> Carlo
>
> > > On Feb 2, 8:02 am, "carlo antonio pignedoli" <c.pig... at gmail.com>
> > > wrote:
> > >> Dear all,
> > >> is there a way to have the stress tensor in
> > >> a GEO_OPT run for a bulk
> > >> calculation whithin QS DFT?
>
> > >> Ciao
>
> > >> Carlo- Hide quoted text -
>
> > - Show quoted text -


More information about the CP2K-user mailing list