My Community

I'm trying to compute the energy of a homogeneous electron gas (LDA/GGA). The manual says that setting the number of ions to zero in the POSCAR and NELECT=n in the INCAR should do this. I have not had any success though. The code crashes with a segmentation fault -- debugger output:

running on 1 nodes
distr: one band on 1 nodes, 1 groups
vasp.5.2.2 15Apr09 complex
POSCAR found : 1 types and 0 ions
LDA part: xc-table for Pade appr. of Perdew

Program received signal SIGSEGV, Segmentation fault.
0x000000000052002d in latord (tau=(( ) ( ) ( ) ), na=Cannot access memory at address 0x0
) at lattlib.f90:1800
1800 CALL ORDER(NA,TAU(1,1),INDEX)

My INCAR:
NELECT = 1

(More detailed INCAR settings do not seem to help.)

POSCAR:
e in box
1.00000000000000
8.6763554104431080 0.0000000000000000 0.0013935712956462
0.0000000000000000 7.3513852390377004 0.0000000000000000
0.0019699538930299 0.0000000000000000 11.8272707088465339
0
Cartesian

The POTCAR seems to be irrelevant(?)... the code will exit with an error if there is no POTCAR file, but I assume that having zero ions automatically means that there is no ionic potential to speak of.

Any suggestions most appreciated!

there is a special POTCAR for jellium, not available on the server.

I am having the same problem. Where can we find the jellium POTCAR?

Thanks very much.