Firefly and PC GAMESS-related discussion club


 
Learn how to ask questions correctly  
 
 
We are NATO-free zone
 



Re: Crash with state-specific gradients in MCSCF

Alex Granovsky
gran@classic.chem.msu.su


Dear Bernhard,

there most likely should be a more informative error message
located a few lines above the message you have provided.
The latter only allows to conclude that the problem was
encountered in the MCSCF module.

In addition, could you please provide me the exact input or
output files?

Kind regards,
Alex Granovsky


On Tue Dec 31 '13 6:28pm, Bernhard Dick wrote
---------------------------------------------
>Dear Alex,
>I want to calculate state specific gradients for two states near a conical intersection. I get a crash with the following message:

> ADDRESS 0x00640C48 HAS INITIATED PROGRAM ABORT BECAUSE OF FATAL ERROR(S)

>The options are:

> $SYSTEM TIMLIM=60000 MWORDS=40 KDIAG=0 NOJAC=1 $END
> $P2P p2p=.t. dlb=.t. $END
> $CONTRL SCFTYP=MCSCF RUNTYP=GRADIENT EXETYP=RUN MULT=2 ICHARG=0
>  MAXIT=50 COORD=UNIQUE fstint=.t. gencon=.f. inttyp=hondo icut=11
>  NZVAR=0 MPLEVL=0 $END
> $BASIS GBASIS=cc-pvdz extfil=.t. $END
> $STATPT METHOD=GDIIS NSTEP=20 NOREG=5 METHOD=CONIC $END
> $MOORTH NOSTF=.T. NOZERO=.T. SYMS=.T. SYMDEN=.T. SYMVEC=.T.
>  SYMVX=.T. TOLE=0.0D0 TOLZ=0.0D0 $END
> $SCF DIRSCF=.TRUE. DIIS=.TRUE. $END
> $MCSCF CISTEP=GUGA fullnr=.f. MAXIT=100 ISTATE=1 NTRACK=2
>  fors=.t. acurcy=1d-7 engtol=1d-12 $END
> $MCAVER JSTATE=2 CONIC=2 $END
> $DRT GROUP=C1 ISTSYM=1 NMCC=9 NDOC=1 NALP=1 NVAL=1 FORS=.T. $END
> $trans mptran=2 dirtrf=.t. mode=112 $end
> $GUGDM2 wstate(1)=1,1  cutoff=1d-12 $end
> $GUGDIA ITERMX=200 NSTATE=2 cvgtol=1d-7 $END
> $gugem pack2=.t. cutoff=1d-13 $end
> $GUESS GUESS=MOREAD NORB=12 $END

>When i replace "RUNTYP=GRADIENT" with "RUNTYP=OPTIMIZE", however, the program performs geometry optimization of the conical intersection. This might be similar to the problem reported with respect to MP2 gradient about 3 weeks ago. GENCON=.F. did not help. I think, such an error should never happen even with incorrect keywords.

>best regards,
>Bernhard
>


[ Previous ] [ Next ] [ Index ]           Tue Dec 31 '13 7:13pm
[ Reply ] [ Edit ] [ Delete ]           This message read 704 times