Firefly and PC GAMESS-related discussion club



Learn how to ask questions correctly


Re^2: Geometry Optimization does not converged after many iterations

Richard
bonarlaw@liv.ac.uk


On Mon Oct 11 '10 7:13am, Alex Granovsky wrote
----------------------------------------------
Actually, this input seems to converge in ~30 steps anyway with versions 7.1.E and later so I'm not sure what the problem is (can't be a version older than 7.1.G otherwise optimizations would stop after 20 steps unless nstep was specified, which its not (?)).

Richard

>Hi,

>you need to increase the default number of the geometry optimization
>steps. It is controlled by NSTEP of $statpt group. The MAXIT variable
>controls maximum number of SCF cycles, not the geometry optimization steps. So you need to add something like:

>

 $statpt nstep=200 $end

>to your input.

>regards,
>Alex Granovsky

>On Fri Oct 8 '10 4:06pm, Amir Nasser Shamkhali wrote
>----------------------------------------------------
>>Dear All
>>I have a problem in structural optimization of a Mn complex. I have tested many basis sets and DFT functionals. But the RMS of gardient is not lowered than 0.0001000 and the following massage is appeared after 60 iterations:
>>          MAXIMUM GRADIENT =  0.0003210    RMS GRADIENT = 0.0001150
>>          HESSIAN UPDATED USING THE BFGS FORMULA
>>             ACTUAL ENERGY CHANGE WAS  -0.0000135315
>>          PREDICTED ENERGY CHANGE WAS  -0.0000097848 RATIO=  1.383
>>            GDIIS STEP HAS LENGTH         =   0.037744
>>          RADIUS OF STEP TAKEN=   0.03774  CURRENT TRUST RADIUS=   0.21213
>>          TRANSFORMING DISPLACEMENT FROM INTERNALS TO CARTESIANS
>>          THE ROOT MEAN SQUARE ERROR IN ITERATION   1 IS   0.00001480
>>          THE ROOT MEAN SQUARE ERROR IN ITERATION   2 IS   0.00000000
>>          MAX PRIMITIVE INTERNAL COORDINATE DEVIATION IS   0.00259918
>>          INTERNAL COORDINATES SELF-INCONSISTENCY     IS   0.17777718

>> CPU        TIME:   STEP =      0.06 ,  TOTAL =    57879.2 SECONDS (  964.7 MIN)
>> WALL CLOCK TIME:   STEP =      0.09 ,  TOTAL =    62983.4 SECONDS ( 1049.7 MIN)
>> CPU UTILIZATION:   STEP =     67.42° TOTAL =      91.90SCP>1     ***** FAILURE TO LOCATE STATIONARY POINT, TOO MANY STEPS TAKEN *****
>>     UPDATED HESSIAN, GEOMETRY, AND VECTORS WILL BE PUNCHED FOR RESTART

>> Dgauss                                                                          
>> **** THE GEOMETRY SEARCH IS NOT CONVERGED! ****
>>What can I do to solve this problem?
>>I attach my input files (external basis of Degauss density fitting basis is used. Thanks for your attention.


[ Previous ] [ Next ] [ Index ]           Mon Oct 11 '10 12:28pm
[ Reply ] [ Edit ] [ Delete ]           This message read 987 times