QuantumATK Forum
QuantumATK => General Questions and Answers => Topic started by: postnikov on December 7, 2010, 02:53
-
In my current calcultion, the obtained transmission are all zero.
I have try to change:
1) all kinds of bases, DZP,SZP etc.
2) Mesh-cutoff,
3) integral_lower_bound and circle_points
4) green_function_infinitesimal
How can I get the right transmission?
-
First, you had better check the log file of your job. Particularly, see whether the total charge is converged to zero or not. If the total charge is converged to nearly zero, the 3rd and 4th parameters mentioned by you need to be tuned.
-
First, you had better check the log file of your job. Particularly, see whether the total charge is converged to zero or not. If the total charge is converged to nearly zero, the 3rd and 4th parameters mentioned by you need to be tuned.
You are right. I first check my structure, then read my log file to check the charge. However, everything is OK, shown as follow.
# sc 45 : q = 277.45785 e Etot = -1292.40783 Ry dRho = 1.1254E-02 dEtot = -4.0064E-03 Ry
# sc 46 : q = 277.43865 e Etot = -1292.40513 Ry dRho = 7.4815E-03 dEtot = 2.6987E-03 Ry
# sc 47 : q = 277.43211 e Etot = -1292.40419 Ry dRho = 3.2435E-03 dEtot = 9.4009E-04 Ry
# sc 48 : q = 277.41565 e Etot = -1292.40180 Ry dRho = 6.6191E-03 dEtot = 2.3908E-03 Ry
# sc 49 : q = 277.45588 e Etot = -1292.40747 Ry dRho = 1.5931E-02 dEtot = -5.6712E-03 Ry
# sc 50 : q = 277.25674 e Etot = -1292.37649 Ry dRho = 8.4306E-02 dEtot = 3.0977E-02 Ry
# sc 51 : q = 277.35372 e Etot = -1292.39248 Ry dRho = 4.2263E-02 dEtot = -1.5981E-02 Ry
# sc 52 : q = 277.40286 e Etot = -1292.39985 Ry dRho = 2.0517E-02 dEtot = -7.3707E-03 Ry
# sc 53 : q = 277.43945 e Etot = -1292.40509 Ry dRho = 1.4680E-02 dEtot = -5.2436E-03 Ry
# sc 54 : q = 277.35574 e Etot = -1292.39272 Ry dRho = 3.4311E-02 dEtot = 1.2374E-02 Ry
# sc 55 : q = 277.35565 e Etot = -1292.39271 Ry dRho = 6.3075E-05 dEtot = 7.3768E-06 Ry
Zh, thanks, any other suggestion?
-
Anders Blom， could you give me some advice about such a problem?
I notice someone have the same problem, however, no method is given.
-
The convergence pattern is not very stable, so maybe you would need to improve on the numerics a bit, maybe longer electrode. But without seeing the input structure it's impossible to guess. I would generally recommend trying ATK 10.8 which has much improved convergence.
Another reason for the transmission being zero is of course that the system is not conducting... Is T(E) identically zero or just very small?