Author Topic: Computed residual : 1.#INF0e+000 #  (Read 4315 times)

0 Members and 1 Guest are viewing this topic.

Offline abhi005

  • Regular QuantumATK user
  • **
  • Posts: 18
  • Country: in
  • Reputation: 0
    • View Profile
Computed residual : 1.#INF0e+000 #
« on: November 12, 2014, 08:58 »
i dont know why this residual is coming and what it shows.please help me regarding this.what is this back engine exception?It shows this back engine exception everytime. how can i remove this ?I am also attaching my .py file.
WARNING #
# #
# The computed multigrid residual is greater than the required accuracy. #
# #
# Computed residual : 1.#INF0e+000 #
# Required accuracy : 1.00000e-012 #
# #
################################################################################


** Back Engine Exception : Singular value decomposition failed
** Location of Exception : mathutils.cpp:1603
« Last Edit: November 24, 2014, 13:19 by abhi005 »

Offline kstokbro

  • Supreme QuantumATK Wizard
  • *****
  • Posts: 392
  • Reputation: 13
    • View Profile
    • QuantumWise
Re: Computed residual : 1.#INF0e+000 #
« Reply #1 on: November 13, 2014, 00:16 »
Something is wrong in the calculation then. Try to make the metallic region larger and make the distance between the dielectric and the device larger, maybe 1 Å.

Offline abhi005

  • Regular QuantumATK user
  • **
  • Posts: 18
  • Country: in
  • Reputation: 0
    • View Profile
Re: Computed residual : 1.#INF0e+000 #
« Reply #2 on: November 13, 2014, 14:28 »
thanks for your reply.I have done what you have suggested but still it shows infinity error.I think it is the problem of poission solver settings.I have choose neumann for both A and B.So i tried by changing these condition by setting B Bottom drichlet and now error is not infinity but quite large in the range of e+000.please suggest me in this direction .Am i going in right direction?