Author Topic: energyContourIntegralParameters  (Read 2677 times)

0 Members and 1 Guest are viewing this topic.

Offline njuxyh

  • QuantumATK Guru
  • ****
  • Posts: 163
  • Reputation: 0
    • View Profile
energyContourIntegralParameters
« on: December 16, 2010, 13:04 »

when i calculated the the transmission spectra,.
I used SZP basis for H and C, the results consist well with the results published in a paper(also ATK calculation)
later, i used DZP for C and SZP for H, it does not convergence at all.
i have tuned parameters such as mixing parameters, but they does not  work.
until i tuned below papameter, it works (according to the ATK forum hints)
 energy_contour_integral_parameters =energyContourIntegralParameters(
    circle_points = 50,
    integral_lower_bound = 4*Rydberg,
however, the transmission spectra has rather differences with the SZP calculation as well as the publised work.

So, would you please tell me have you faced the similar question?  why different integral_lower_bound inflence the result so much.
if yes, would you give me some hints? Anyway, i feel the scf convergence is a tricky work.
 
THANKS ALOT!

Offline zh

  • QuantumATK Support
  • Supreme QuantumATK Wizard
  • *****
  • Posts: 1141
  • Reputation: 24
    • View Profile
Re: energyContourIntegralParameters
« Reply #1 on: December 23, 2010, 07:44 »
The different results obtained in your calculations may comes from the different setup of basis set for C. If you wants to confirm that it is due to the influence of "integral_lower_bound",  you should a third calculation with the SZP basis for H and C and the same setup of "energyContourIntegralParameters(integral_lower_bound = 4*Rydberg)" mentioned by you, and then do the comparison.