Problem calculating partial DOS

Message

keith · #1 Post by **keith** » Mon May 14, 2007 8:58 am

I am trying to calculate the partial DOS of a large cell of MgO that contains a grain boundary. I have already optimised the cell size and relaxed the atomic coordinates. I use the optimised POSCAR with the following input:

----------------------------------------------
SYSTEM = MgO GB
ENCUT = 400
GGA = 91
IALGO = 48
LREAL= A
NELMIN = 4
BMIX=2.
MAXMIX=50

ISPIN=1
LORBIT = 10

ISYM=0
NSW = 0
IBRION = -1
ISIF=0
POTIM= 0.5
NPAR = 4

ISMEAR = 1
SIGMA= 0.1
--------------------------------------------

I get the following error on the standard output. The code does not exit and continues to run with no further output.

-------------------------------------------
running on 128 nodes
distr: one band on 32 nodes, 4 groups
vasp.4.6.26 15Dec04 gamma-only
POSCAR found : 2 types and 320 ions
LDA part: xc-table for Ceperly-Alder, standard interpolation
POSCAR, INCAR and KPOINTS ok, starting setup
WARNING: wrap around errors must be expected
FFT: planning ... 1
reading WAVECAR
internal ERROR RSPHER:running out of buffer 19184 48 8 2 19499
internal ERROR RSPHER:running out of buffer 19184 48 8 2 19499
internal ERROR RSPHER:running out of buffer 19184 48 8 2 19499
internal ERROR RSPHER:running out of buffer 19184 48 8 2 19499
internal ERROR RSPHER:running out of buffer 19440 36 8 2 19499
internal ERROR RSPHER:running out of buffer 43192 72 8 2 43310
internal ERROR RSPHER:running out of buffer 43192 72 8 2 43310
internal ERROR RSPHER:running out of buffer 30808 62 8 2 31270
internal ERROR RSPHER:running out of buffer 19440 36 8 2 19499
internal ERROR RSPHER:running out of buffer 30808 62 8 2 31270
internal ERROR RSPHER:running out of buffer 19440 36 8 2 19499
internal ERROR RSPHER:running out of buffer 19440 36 8 2 19499
internal ERROR RSPHER:running out of buffer 36768 52 8 2 36904
internal ERROR RSPHER:running out of buffer 43192 72 8 2 43310
internal ERROR RSPHER:running out of buffer 36768 52 8 2 36904
internal ERROR RSPHER:running out of buffer 30808 62 8 2 31270
internal ERROR RSPHER:running out of buffer 30808 62 8 2 31270
internal ERROR RSPHER:running out of buffer 43192 72 8 2 43310
internal ERROR RSPHER:running out of buffer 36768 52 8 2 36904
internal ERROR RSPHER:running out of buffer 36768 52 8 2 36904
internal ERROR RSPHER:running out of buffer 72440 70 8 2 72475
internal ERROR RSPHER:running out of buffer 72440 70 8 2 72475
internal ERROR RSPHER:running out of buffer 72440 70 8 2 72475
internal ERROR RSPHER:running out of buffer 72440 70 8 2 72475
--------------------------------------

I looked in the forum and there is a post that attributes this problem to a bug that has been rectified some time ago. I am running a version of VASP that should not have this bug (vasp.4.6.26). Can anyone help with this problem?

#2 Post by **admin** » Tue May 15, 2007 2:33 pm

1) please check if the optimized output geometry of the structural optimization is reasonable (pressure, XDATCAR,...)
2) if it is, please do the following:
search for the following lines in nonlr.F
IF (IRMAX > NONLR_S%IRMAX) THEN
! IRMAX is the maximum global number, could be improved !!!!
NONLR_S%IRMAX =IRMAX *1.1
LREALLOCATE=.TRUE.
ENDIF
IF( IRALLOC > NONLR_S%IRALLOC) THEN
! more safety on parallel machines increase by 20 %
NONLR_S%IRALLOC =IRALLOC*1.2
LREALLOCATE=.TRUE.
ENDIF
and change
NONLR_S%IRALLOC =IRALLOC*1.1 to
NONLR_S%IRALLOC =IRALLOC*1.2
and
NONLR_S%IRALLOC =IRALLOC*1.2 to
NONLR_S%IRALLOC =IRALLOC*1.3

(or, if necessary to higher values)

askhetan · #3 Post by **askhetan** » Wed Jan 30, 2013 1:36 pm

and then do we have to recompile the whole software? i am facing the same error of no buffer in that file.

#4 Post by **admin** » Wed Jan 30, 2013 3:24 pm

Of course you need to recompile the modified source code.