LREAL flag

Message

lmiao · #1 Post by **lmiao** » Fri Mar 31, 2006 3:13 am

Hi,

If LREAL=T (using real space calculation) is for rough calculation, how much difference it will have comparing to LREAL=F? I was testing a relaxation job on two clusters, using LREAL=T on one cluster and LREAL=F on the other. The energy of first couple steps from the cluster using LREAL=T is
1 F= -.43820718E+03 E0= -.43820718E+03 d E =-.438207E+03
2 F= -.43858848E+03 E0= -.43858848E+03 d E =-.381297E+00
3 F= -.43879475E+03 E0= -.43879475E+03 d E =-.587572E+00

while the other cluster using LREAL=F gives
1 F= -.43806228E+03 E0= -.43806228E+03 d E =-.438062E+03
2 F= -.43846471E+03 E0= -.43846471E+03 d E =-.402425E+00
3 F= -.43867796E+03 E0= -.43867796E+03 d E =-.615677E+00

Is this a significant difference?

Thank you

#2 Post by **admin** » Tue Apr 11, 2006 1:07 pm

depending on what you intend to calculate, I would say yes.
Please mind that you have to keep to the SAME settings of LREAL,
especially if you want to calculate trends or energy differences