## Firefly and PC GAMESS-related discussion club

Learn how to ask questions correctly

** Re^3: Restarting VIbrational Calculation when $FORCE has been omitted **
sanya

sanya@photonics.ru

OK, let's start again.The output file you've attached is just a single-point hessian run (and EXECUTION OF FIREFLY TERMINATED NORMALLY, in contrary to what you say). Since the geometry was obviously not optimized, this is not a stationary point. Hence, the warning "THIS IS NOT A STATIONARY POINT ON THE MOLECULAR PES THE VIBRATIONAL ANALYSIS IS NOT VALID !!!" appeared.

I'm afraid you didn't perform geometry optimization, i.e., the calculation with RUNTYP=OPTIMIZE, before running hessian calculation. If so, you should first optimize the geometry. Set RUNTYP=OPTIMIZE, add NSTEP=1000 and HSSEND=.T. to the $STATPT group, and hessian will be calculated automatically after the geometry convergence is achieved. By the way, in hessian-only runs (i.e., with RUNTYP=HESSIAN), $STATPT group is ignored.

If your geometry optimization ended without convergence (no ***** EQUILIBRIUM GEOMETRY LOCATED ***** in the output), visualize your output with a proper viewer and see whether any geometry optimization steps were made at all. Analyze your output carefully. In case of abnormal termination (EXECUTION OF FIREFLY TERMINATED ABNORMALLY), post your output here.

If several steps were made, but the calculation ended abruptly (say, due to power failure), look for the last entry "COORDINATES OF ALL ATOMS ARE (ANGS)" in the output. This is the last geometry you need. Use it for restart (just replace the atomic coordinates in the $DATA group with these new coordinates). Restart until geometry converges (hessian will be calculated automatically with the converged geometry).

One more improvement may help you to accelerate the calculation, especially for larger molecules:

$SCF DIRSCF=.TRUE. $END

By the way, I'm really surprized that FireFly understands the HyperChem atomic coordinate format

C 6 1.193982, -3.321735, -0.4646571

Fri May 4 '12 1:06am

This message read **1296** times