TURBOMOLE Users Forum
TURBOMOLE Modules => Aoforce and Numforce => Topic started by: greisen on July 07, 2011, 06:30:57 PM
-
Hi,
I have done a geometry optimization using cosmo and would like to calculate the hessian. To include the cosmo- treatment I am using the NumForce
NumForce -ri -c -cosmo > NumForce.log
but it gives me the error
$hessian : could not find file <hessian> !
how do I specify it to calculate the hessian?
Thx,
-
Hi,
in older version you should find an error message in your NumForce.log.
In the current version you will get this message:
PLEASE NOTE: -cosmo does not make sense in combination with -c.
For a frequency calculation the structure should be converged
If you are interested in the Hessian, -cosmo is not needed
In this case the full relaxed cosmo charges are used.
Cheers,
Arnim
-
Hi,
The geometry is converged and I do not get any error message in the NumForce.log file. I have tried with and without the -cosmo option but still not hessian is calculated - I have look in the log-files generated during the run but I was not able to locate the error there
any advise how to debug this?
Cheers
-
Hi,
in that case it is quite likely, that one (or more) of the steps did not produce a proper gradient.
It should be possible to find that step in the output file, but you can also go to numforce/KraftWerk
and check the *.log files there.
Best regards,
Arnim
-
Hi,
this just happened to me because there was not enough memory to do the jobs simultaneously.
Would be nice if Numforce would check up on that and print some error before starting aoforce.
I will switch to a machine with more memory. Can I tell Numforce to do the jobs one after another, but mpi-parallelized, instead of running a bunch of jobs at once?
Cheers, Martin
-
Hey,
I think I have got the same error message
$hessian : could not find file <hessian> !
MODTRACE: no modules on stack
HSSRD : error reading $hessian !
force ended abnormally
I did an excited state optimization using jobex -level cc2 with $scfconv 8
after that I started NumForce with NumForce -level cc2 -central -d 0.02 > force.out
and in the control file I wrote
$ricc2
cc2
geoopt model=cc2 (a 1)
$excitations
irrep=a multiplicity= 1 nexc= 3 npre= 6 nstart= 9
xgrad states=(a 1)
spectrum states=all operators=xdiplen,ydiplen,zdiplen
exprop states=all operators=diplen
...
but I still get the error message and there is no hessian file in my directory
I already checked the KraftWerk log file but there is no error message.
Has someone a solution for this problem?
-
Hi,
I had the same problem with an excited state frequency calculation with Turbomole6.1. After a lot of unsuccessful efforts I tried the 6.3 version and now the problem with the hessian file seems to have vanished...
Best regards
Christof
-
Hello,
I am having the same error with NumForce where in aoforce.out file, the followings error are printed
$hessian : could not find file <hessian> !
========================
internal module stack:
------------------------
force
========================
HSSRD : error reading $hessian !
force ended abnormally
My calcuation is with TURBOMOLE7.1. I am looking forward to any suggestion on how to solve this issue.
Regards,
Iban
-
Hi,
the error message just tells you that there has been a problem during the numerical 2nd derivative calculation, and the resulting Hessian file has not been created. The cause can come in many ways. Non-converged single-point energy calculations or, if excited states are used, TDDFT instabilities if you are near a conical intersection, or something else...
I'd recommend to contact the support team to find out what happened in your case, because without inspection of the individual jobs in the numforce/KraftWerk directory it is not possible to find the reason for the failure.
It often helps to look at the KraftWerk directory and check if there are log or results files with zero length.
Regards,
Uwe