potential energy per atom

Dear lammps users,

I use this set of directives to calculate potential energy per atom
value in 3C-SiC

group Si type 1
group C type 2

compute peratomsi Si pe/atom
compute pe_si Si reduce sum c_peratomsi

compute peratomc C pe/atom
compute pe_c C reduce sum c_peratomc

thermo_style custom step temp press etotal pe c_pe_si c_pe_c

The result is the same for both c_pe_si and c_pe_c and equals half of pe.
This looks strange.

What I'm doing wrong?

Thanks

German

Dear lammps users,

I use this set of directives to calculate potential energy per atom
value in 3C-SiC

group Si type 1
group C type 2

compute peratomsi Si pe/atom
compute pe_si Si reduce sum c_peratomsi

compute peratomc C pe/atom
compute pe_c C reduce sum c_peratomc

thermo_style custom step temp press etotal pe c_pe_si c_pe_c

The result is the same for both c_pe_si and c_pe_c and equals half of pe.
This looks strange.

What I'm doing wrong?

there is not enough information here to make a qualified statement.

axel.

with SiC meam potential.

I think Damiel is right. Both meam and tersoff potentials for SiC have
interactions between nearest neighbors only.

German

I suggest you first dump the pe/atom values in a dump
file and see if they are what you expect.

Steve

Dear German,

I tried wht the attached "in.meam" in the "examples/meam" direcotry of
"lammps-30Aug12".
I got:

../../src/lmp_crockett < in.meam
LAMMPS (30 Aug 2012)
Reading data file ...
  orthogonal box = (-6 -6 -6) to (5.97232 5.97232 5.97232)
  1 by 1 by 1 MPI processor grid
  128 atoms
64 atoms in group Si
64 atoms in group C
Setting up run ...
Memory usage per processor = 5.10825 Mbytes
Step Temp Press TotEng PotEng pe_si pe_c
       0 0 -76571.819 -636.38121 -636.38121
-322.01796 -314.36325
      10 1807.8862 -150571.49 -636.54126 -666.21959
-331.71497 -334.50463
      20 1932.4467 -120223.52 -636.53498 -668.2581
-336.54629 -331.71181
      30 1951.3652 -100508.4 -636.54771 -668.58139
-339.94967 -328.63172
      40 2172.5974 -110753.34 -636.5617 -672.22715
-339.19303 -333.03412
      50 2056.9149 -105418.07 -636.56468 -670.33108
-335.96935 -334.36173
      60 1947.9564 -111413.04 -636.55015 -668.52788
-335.92205 -332.60583
      70 1994.7712 -109645.76 -636.54225 -669.28849
-336.14335 -333.14514
      80 2126.0903 -97475.831 -636.53557 -671.43755
-338.09941 -333.33813
      90 2065.755 -95858.837 -636.52338 -670.4349
-337.77919 -332.6557
     100 2051.4553 -107068.9 -636.53122 -670.20799
-336.80188 -333.40611
Loop time of 0.14357 on 1 procs for 100 steps with 128 atoms
...

From the the two components of MEAM energy, half of the "pair

potential component" should be attribued to Si and other half to C
atoms, so this component could be the same for Si and C - at least
when the configuration is symmetric w.r.t. Si and C. "Embedding
component" of the energy should, in general, differ for every
individual atom.

Best,
-Bohumir

in.meam (698 Bytes)