cphf solve2:SCF residual greater than 1d-2: What's this mean?


Just Got Here
We're running NWCHEM 6.1 on a 64-bit Linux RHEL5 system.

I have a user whose calculation stops with multiple occurrences of the error

cphf_solve2:SCF residual greater than 1d-2        0
------------------------------------------------------------------------
------------------------------------------------------------------------
current input line :
133: task dft freq
cphf_solve2:SCF residual greater than 1d-2 0
------------------------------------------------------------------------
------------------------------------------------------------------------
------------------------------------------------------------------------
------------------------------------------------------------------------
This error has not yet been assigned to a category

I have no idea what this means.

The input that produces this is available at

http://www.crc.nd.edu/~msuhovec/files/fc2_plus10.nw

The full output is available at

http://www.crc.nd.edu/~msuhovec/files/fc2_plus10.job.o440490

Any help much appreciated

Mark Suhovecky
HPC System Administrator
Center for Research Computing
University of Notre Dame
suhovecky@nd.edu

Forum Regular
Hi Mark,

I think the problem you are hitting is due to the fact that the Hessian code does not have the constrained DFT terms implemented. So when the CPHF code checks that the orbitals you are giving it are solutions to the Kohn-Sham equations that test fails. If you want to you can still do a frequency calculation but you have to use numerical frequencies. I have listed a simple example of such a calculation below:

 geometry
o 0.0 0.0 0.0
h 0.0 0.0 2.0
end
basis
* library 6-31g**
end
dft
xc b3lyp
odft
mult 2
iterations 100
cdft 1 1 charge -1
end
driver
maxiter 100
end
task dft optimize
task dft freq numerical

I hope this helps,

    Huub

Just Got Here
Huub-

The user tried this- and it worked!

Thanks so much for your help,

Mark


Mark Suhovecky
HPC System Administrator
Center for Research Computing
University of Notre Dame
suhovecky@nd.edu

Just Got Here
I'm using NWChem 6.5 to run molecular properties calculations (I'm more interested in response).

I got a similar problem than Mark had, but in my case the output message was

--------------------------------------------------------------------------
 SCF residual:   4.755368962564806E-002
 ------------------------------------------------------------------------
 cphf_solve2:SCF residual greater than 1d-2                   0
 ------------------------------------------------------------------------
 ------------------------------------------------------------------------
  current input line :
    46: task dft property
 ------------------------------------------------------------------------


The input that produces this is


start dplot_w1_2.4954

permanent_dir ./
scratch_dir ./

geometry  nocenter noautoz noautosym
 O 0.00000000 0.00000000 0.00000000
 H -0.36883624 0.79528131 0.39293982
 H -0.36714416 -0.71845082 0.52149486
 bqO 3.45970300 0.00000000 0.05516905
 bqH 2.49539000 0.00000000 0.00000000
 bqH 3.75249180 -0.09105210 -0.85339719
end
basis
H library aug-cc-pvtz
bqH library H aug-cc-pvtz
O library aug-cc-pvtz
bqO library O aug-cc-pvtz
end
dplot
  TITLE DENSITY
  vectors firstW1g.bsse.movecs
   LimitXYZ
 -5.0 10.0 90
 -5.0 5.0 90
 -5.0 5.0 90
  spin total
  gaussian
  output dens_w1.cube
end
task dplot

dft
xc b3lyp
vectors input firstW1g.bsse.movecs
noscf
end
task dft

property
dipole
response
vectors firstW1g.bsse.movecs
end
task dft property



Thanks in advance for any help.


Sara Gómez
Purdue University Student


Forum >> NWChem's corner >> Running NWChem