[Pw_forum] volume discrepancy when adjusting (ecfixed, q2sigma, qcurtz)

[Riping WANG]

Dear Forum,

I have some questions on setting the parameter set (ecfixed,q2sigma,qcurtz)
in cp.x of espresso-4.2.1.
I use SiO2 of example18 in source code.

I found that when I adjust (ecfixed,q2sigma,qcurtz) the average volume at
equilibrium state varies.
How do think of this volume disagreement?
What is the criteria to select the best (ecfixed,q2sigma,qcurtz) value?



I summarized my results in detail following.


In this code, the original value is  (ecfixed=16,q2sigma=2,qcutz=150).
I did adjustment with all other parameter fixed as the same to check the
pressure, volume, and total energy at equilibrium state of 10Kbar-10K.
I adjust ecfixed first, then q2sigma, then qcutz.

1. adjusting ecfixed.
ecfixed (with q2sigma=2,qcutz=150)        average  volume
volume vibarion
17.1
1340                     small
17.0                                                                1350
                    small

16.9
1360                     small
16.8
1370                     small
16
1330                     larger
19
1390                     larger
All above have similar pressure vibration.
The larger ecfixed, the lower total energy.

How do think of this volume discrepancy? Is this satisfying with inaccuracy?
What is the criteria to select the best ecfixed value?
Now, without anyother base, personally, I think that ecfixed=16.8 has small
volume vibration and it has the largest volume among those with small volume
vibration.
Maybe ecfixed=16.8 it is.

2. adjust q2sigma
q2sigma (with ecfixed=16.8,qcutz=150)           volume
volume vibration
1.6
1420                           larger
1.8                                                                1345
                      larger
2.0
1370                           small
2.2                                                                1380
                     larger
2.4                                                                1370
                      larger
2.6                                                                1330
                      larger
All above have similar pressure vibration.
The smaller q2sigma, the lower total energy.

I think that q2sigma=2 is the best because it has smallest volume vibration.

Right?

3. adjusting qcutz
qcutz (with ecfixed=16.8,q2sigma=2)           volume                 volume
vibration
50
1310                           small
80                                                                1335
                      small
120
1360                           small
150                                                              1370
                   small
180                                                              1380
                    small
400                                                              1370
                    small
All above have similar pressure vibration.
The smaller qcutz, the lower total energy.

They have similar results except the volume discrepancy.
So what is the criteria to select ?



In sum, regarding the volume discrepancy, how to select the best value for
(ecfixed,q2sigma,qcurtz) ?


Thanks for you attention.
Any comment is grateful.


WANG Riping
2010.1.7


-- 
******************************************************************************
WANG Riping
Ph.D student,
Institute for Study of the Earth's Interior,Okayama University,
827 Yamada, Misasa, Tottori-ken 682-0193, Japan
Tel: +81-858-43-3739(Office), 1215(Inst)
E-mail: wang.riping.81 at gmail.com
******************************************************************************
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.quantum-espresso.org/pipermail/users/attachments/20110107/9b746908/attachment.html>