Recirculation regions in DPM cyclone separator

    • anonn72

      I have a question about recirculation regions and incomplete particles in the simulation of a cyclone separator. We are simulating a cyclone separator with particle-laden airflow. Modeling parameters include automatic (tetrahedral) mesh, RSM, and two-way coupling with the DPM model and steady tracking.

       I have read many places to be careful of recirculation regions when modeling particle flow. However, I am unsure of how to do so with the vortex cyclone separator. We are using Rosin-Rammler distribution to inject particles 0-75um. Larger particles exit the cyclone, however at least 90% of smaller particles (0-10um) do not exit the cyclone and are marked as incomplete. It may be relevant that our flow is relatively low-velocity, and I have tested and confirmed that the small particles do exit the cyclone with higher velocity. We have tried increasing the maximum number of steps for the particles and this has not solved the issue. Therefore, it makes me believe it may be an issue of recirculation regions. (Note: We have also tried extending the length of the simulation, and this has not helped. The residence time of the incomplete particles is much less than the total flow time.) 

      The main issue issue with this is that we are trying to obtain collection efficiency values in the 0-10um range. Therefore, we cannot do so if almost all of the 0-10um particles injected are incomplete. I have seen the paper "Numerical study of gas–solid flow in a cyclone separator" which suggests that an incomplete particle would fall in reality due to particle-particle interactions, but if we were to assume this our collection efficiency values would be quite unrealistic. 


      Does anyone have information on either

       a) how to adjust our parameters to get the particles to completion 

       b) how to avoid recirculation regions in a cyclone separator 

       c) how to classify/treat incomplete particles in a cyclone separator to obtain collection efficiency values 


      Some details of our DPM tracking parameters: 

      Steady Tracking 

      Update DPM Source Terms Every Flow Iteration (DPM Iteration Interval = 20)

      Maximum Number of Steps: 100,000 (did not improve with 1,000,000)

      Step Length Factor: 5

      Mesh cell size: 0.01

      Particle velocity: 0.75 m/s

      Simulation length: 10s (2000x0.005s time steps)


      Thanks in advance for your help 

    • anonn72

      -- An extra question: does anyone know how to output particle fate vs. particle variables? e.g. particle fate vs. particle velocity or particle fate vs. diameter? I haven't been able to figure out how to do this 

    • DrAmine
      Ansys Employee
      You might better try using unsteady particle tracking. Moreover as additional tracking parameter you will only have number of integration steps and the step size or factor.
Viewing 2 reply threads
  • You must be logged in to reply to this topic.