Fluids

Fluids

Topics relate to Fluent, CFX, Turbogrid and more

Spray combustion modelling using pdf and dpm

    • anu4anusai
      Subscriber

      I am modelling spray combustion in gas turbine combustors using dpm for spray (pressure swirl model) and non premixed combustion by generating pdf. I would like to know whether we can use parallel computing for dpm?


      If yes, how do we fix the number of streams?


      How do we conclude the convergence of the solution? My residuals are now fluctuating between e-2 and e-3.

    • Rob
      Ansys Employee

      You can use parallel, but check the model settings against how your hardware (memory) is set up.


      If you mean the number of parcels, that'll depend on the model. To increase have a look at stochastic tracking.


      Convergence is measured in the same way as normal. With more particle streams you may find the model is more stable. Also read the manual section on DPM numerics.

    • anu4anusai
      Subscriber

      My residuals are fluctuating at e-2 and e-3 and not coming beyond that.

    • anu4anusai
      Subscriber

      Residuals


      sum of dpm mass source


      dpm evaporated mass


      mass in domain

    • Rob
      Ansys Employee

      Keep going. I think you've turned transient particles on so you'll need to wait until the monitors etc level off.

    • anu4anusai
      Subscriber

      I dont seem to have convergence for any of the residuals as well as the other parameters. I am not able to understand if any mistakes are there in the case setup.



         




Viewing 5 reply threads
  • You must be logged in to reply to this topic.