[lammps-users] Fix- Phonon

Dear Steve and Kong

Regarding fix_phonon, I am using LAMMPS 16 Dec 2010 version for that,
while making user-phonon, I am getting the following errors showing at
line 351 of fix_phonon.cpp.

make[1]: Entering directory /vijay/lammps-16Dec10/src/Obj_linux' Makefile:93: modify.d: No such file or directory gcc -O -DLAMMPS_GZIP -DMPICH_SKIP_MPICXX -DFFT_FFTW -M modify.cpp > modify.d make[1]: Leaving directory /vijay/lammps-16Dec10/src/Obj_linux’
make[1]: Entering directory /vijay/lammps-16Dec10/src/Obj_linux' gcc -O -DLAMMPS_GZIP -DMPICH_SKIP_MPICXX -DFFT_FFTW -c fix_phonon.cpp fix_phonon.cpp: In member function ‘virtual void LAMMPS_NS::FixPhonon::end_of_step()’: fix_phonon.cpp:351: error: ‘MPI_Scatterv’ was not declared in this scope make[1]: *** [fix_phonon.o] Error 1 make[1]: Leaving directory /vijay/lammps-16Dec10/src/Obj_linux’
make: *** [linux] Error 2

I have blocked the line 351 and again recomplied, it was through but
while running the example problem it is giving error after the time
step of 999900. If I select lesser run time the problem runs fine but
for longer time its giving the error of “Singular matrix in complex
GaussJordan”.

make[1]: Entering directory `/vijay/lammps-16Dec10/src/Obj_linux’
gcc -O -DLAMMPS_GZIP -DMPICH_SKIP_MPICXX -DFFT_FFTW -M fix_phonon.cpp

fix_phonon.d
make[1]: Leaving directory /vijay/lammps-16Dec10/src/Obj_linux' make[1]: Entering directory /vijay/lammps-16Dec10/src/Obj_linux’
gcc -O -DLAMMPS_GZIP -DMPICH_SKIP_MPICXX -DFFT_FFTW -c fix_phonon.cpp
gcc -O -DLAMMPS_GZIP -DMPICH_SKIP_MPICXX -DFFT_FFTW -c modify.cpp
gcc -O -L/vijay/mpich-1.2.7/lib -L/usr/local/lib angle_charmm.o
angle_cosine.o…verlet.o write_restart.o -lmpich -lpthread
-lfftw -lstdc++ -o …/lmp_linux
size …/lmp_linux
text data bss dec hex filename
2388306 2804 276 2391386 247d5a …/lmp_linux
make[1]: Leaving directory `/vijay/lammps-16Dec10/src/Obj_linux’

LOG File details:

999400 0.006791104 0.0020357428 0.0050068508
999500 0.0057438667 0.0032590117 0.0057719534
999600 0.0046363269 0.0033060662 0.0053344592
999700 0.0078564222 0.0021085079 0.0055456926
999800 0.0050788373 0.0014833157 0.003705307
999900 0.0020899096 0.0011981872 0.0021125226
ERROR on proc 0: FixPhonon: Singular matrix in complex GaussJordan!

Any commnets will be highly appreciated.

warm regards

Vijay Kumar Sutrakar

I’m not familiar with fix phonon - it’s not part of the
current LAMMPS distribution. So I suggest contacting
the author, if you haven’t already.

Steve