GU: Troubles with restart vibrational calc.

[Brion Jean <jean.brion@univ-reims.fr>]

                 Dear netters,


I meet some troubles when using the Intrinsic vibrational 
frequencies calculation in GAMESS, by the "DECOMP"
option in $FORCE.

I have compared 2 jobs , with (for hessian calculation)
the same molecule, same geometry , same coord. unit:

JOB1: RUNTYPE=OPTIMIZE in $CONTRL
          HSSEND=.TRUE.     in $STATPT
          NVIB=2 METHOD=NUMERIC in $FORCE
          DECOMP=.TRUE.    in $FORCE

JOB2: RUNTYPE=HESSIAN  in $CONTRL
          RDHESS=.TRUE. DECOMP=.TRUE. in $FORCE
          $HESS (taken from JOB1.dat )


The two  corresponding INTRINSIC VIBRATIONAL ANALYSIS 
are DIFFERENT ! 

While the two corresponding set of (CLASSICAL)
VIBRATIONAL FREQUENCIES are identical 


for example: 

              SETS OF VIB. FREQ
JOB1: 64,  92,  152,  214, 229 ... 3322, 3322, 3337, 3337

JOB2: 64,  92,  152, 214, 229, ... 3322,  3322, 3337, 3337

              SETS OF INTRINSIC VIB. FREQ
 STR.      O1-O2      O3-O1   C1-O1   C1-H5  C1-H6½  

JOB1:     1059.8       1097.2     1024.9   3295.7   3315.6

JOB2:     1009.8      494.9     858.7    1397.5    2367.6 


I checked several things:

 - the force constant matrix used in JOB2
    was the last one punched after JOB1 optimization in JOB1.dat 
 - the cartesian force constant matrix in JOB2 is well read
    (JOB1 cartesian hessian=JOB2 cartesian hessian)
 - the geometry is the same in each case, with
    the same orientation
 - the geometry corresponds to the  hessian I used 
 - the coordinate unity in each case

But, I have observed that JOB1 hessian matrix in INTERNAL
COORDINATES was different from which one  computed in JOB2 !
, while internal coordinates are well calculated and have the same 
value in each case...
 Now, INTRINSIC VIBRATIONAL CALCULATION  need this
matrix and this could explain my problem...

I tried to change COORD=ZMAT in COORD=CART
and create a $ZMAT group, but no amelioration was
found.

I tried another molecule (a smaller one: CH3O2) ... and 
the two intrinsic vibrational analysis agree  for JOB1 and JOB2.

Then, my questions are:

- do I forget something in the GAMESS cards ?
- do you know similar problems on other systems with
intrinsic vibrational analysis ?

- does intrinsic vib. analysis need  something else that
the internal coordinates hessian matrix in order to work ?

-do you have any other idea to solve this problem ..?


            Thank you in advance .

		  F. Bohr





**************************************************
*     Dr F. Bohr                                 *
*     Laboratoire de Chimie Physique             *
*     Universite de Reims, Faculte des Sciences  *
*     Moulin de la Housse                        *
*     B.P. 1039                                  *
*     51687 Reims Cedex 2                        *
*     FRANCE                                     *
*     Tel: (33) 26-05-32-33                      *
*     Fax: (33) 26-05-31-47                      *
*     E-mail: jean.brion@univ-reims.fr           *
**************************************************