Local URF for DPM source terms (FLUENT)

HadiNHadiN Member Posts: 7


In FLUENT, I am simulating a gas-solid system in steady state (and steady dpm tracking). In some local regions (maybe less that 1% of the cells), the solid-to-fluid mass fraction becomes so high (even though the volume fraction stays below 10%) that I have to use very small URFs for the exchange terms to have a stable simulation.

Is there a way to apply momentum, heat, and mass transfer terms with a locally-varied under relaxation factor?

Thank you




  • RobRob UKForum Coordinator Posts: 8,395

    No. What you may need to do is reduce the parcel mass by injecting more particles (change the injection source or increase stochastic tries). If you look at the affected areas what is causing the particles to cluster?

  • HadiNHadiN Member Posts: 7

    Thank you for the answer.

    I think in the problem that I am solving, having local regions with high concentration of particles is inevitable. This happens especially at the inlet location of particle injection. Also, I am already using N.O. tries larger than one in stochastic tracking.

  • RobRob UKForum Coordinator Posts: 8,395

    Try 10-20 stochastic tries. The high volume fraction isn't always a problem, it's where you get high particle source terms that convergence issues can occur.

  • HadiNHadiN Member Posts: 7

    I have shown a part of my simulation domain below, which is adding a flow of (colder) particles to a main (hot) gas stream. Particles go through an endothermic reaction as well (which explains the nearly constant temperature along particle tracks). Increasing the number of stochastic tries helps with the stability of solution in the main riser, where particles are dispersed; but not the injection zone, especially the cells that I marked.

    In these cells, if I use a large value for DPM URF (e.g. 0.1), the source terms, especially the energy equation source terms, are significantly overpredicted (which seems to be logical since the particle stream heat capacity is 100-400 times larger than the fluid stream heat capacity). For the example below I get a stable solution for a DPM URF of 0.002. For this URF, and in best conditions, I need around 1500 DPM source term updates to reach to 95% of the final value source term.

    That is why I am interested to see if it is possible to use a local URF for DPM source terms.

    Best Regards


  • DrAmineDrAmine GermanyForum Coordinator Posts: 6,649

    There is no local URF. You better need to increase grid resolution as well as the number of particle streams.

  • HadiNHadiN Member Posts: 7

    Thank you for suggestions DrAmine and Rob


Sign In or Register to comment.