[molpro-user] F12 core/valence correction problem
Simmie, John
john.simmie at nuigalway.ie
Fri May 18 11:00:38 BST 2012
In trying out an F12 core/valence correction calculation to atomisation
energies at the CVTZ level for the first time
I've encountered the following error during the CCSD(T) program
- - - - - - -
CCSD correlation energy -1.393661897713
MP2-F12 correction (uncoupled) -0.112235442734
C*T doubles correction -0.004911738587
U*T singles correction 0.000000000000
f12b correction 0.015798932350
Total F12 correction -0.101348248971
Read error in iow_direct_read; fd=10, l=32768, p=95975236; read returns
-1
2:2:fehler:: 4
(rank:2 hostname:stoney66 pid:31644):ARMCI DASSERT fail.
src/common/armci.c:ARMCI_Error():208 cond:0 Adding F12 terms to K(Cij),
methodcc=6, factor=1.0
Is my input file seriously at fault?
***, C5H6O 2-methylfuran
memory,450.0,m;
basis=cvtz-f12
! C7
! \
! C3 -- C5
! / |
! O2--X1 |
! \ |
! C4 -- C6
!
geometry={dum;
o2,dum,1.0;
c3,o2,rc3o,dum,xoc3;
c4,o2,rc4o,dum,xoc4,c3,180.0;
c5,c3,rc3c5,o2,oc3c5,dum,0.0;
c6,c4,rc4c6,o2,oc4c6,dum,0.0;
c7,c3,rc3c7,o2,oc3c7,dum,180.0;
h8,c4,rc4h8,o2,oc4h8,dum,180.0;
h9,c5,rc5h9,c3,c3c5h9,dum,180.0;
h10,c6,rc6h10,c4,c4c6h10,dum,180.0;
h11,c7,rc7h11,c3,c3c7h11,o2,dihe;
h12,c7,rc7h11,c3,c3c7h11,o2,-dihe;
h13,c7,rc7h13,c3,c3c7h13,o2,180.0}
rc3o=1.36658844 ang
xoc3=54.18584466 deg
rc4o=1.36494428 ang
xoc4=52.85083433 deg
rc3c5=1.36059527 ang
oc3c5=109.98490040 deg
rc4c6=1.35692832 ang
oc4c6=110.61271397 deg
rc3c7=1.48655560 ang
oc3c7=116.45104531 deg
rc4h8=1.07576363 ang
oc4h8=115.88728673 deg
rc5h9=1.07743170 ang
c3c5h9=125.91083270 deg
rc6h10=1.07702941 ang
c4c6h10=126.35049570 deg
rc7h11=1.09131999 ang
c3c7h11=111.00298343 deg
dihe=60.07129155 deg
rc7h13=1.08890722 ang
c3c7h13=109.53298239 deg
{hf;
wf,44,1,0;} ! # electrons, irrep # and multiplicity-1
{ccsd(t)-f12b,df_basis=awcvtz/mp2fit,gem_beta=1.4;
core;}
optg; ! perform geometry optimization
---
Thanks in advance
Emeritus Professor John Simmie::Combustion Chemistry Centre::NUI Galway,
Ireland
More information about the Molpro-user
mailing list