Learning Forum

[aronjo79]

Hello,nI am having an issue with the Eulerian Wall Film model and the azimuthal droplet rebound. In brief, the current rebound deviation angle is very high.n? The current Eulerian Wall Film model with the droplet impingement models show unrealistic droplet deviation angles (after rebound). This is due to the -90 to +90 degree deviation angle or the azimuthal angle range from the incident direction in both the Stanton-Rutland and the Kuhnke Model.n? Additionally, the equation for calculating the deviation angle has a uniformly distributed random parameter. However, this randomness was not observed for repeated droplet rebound injections from the same location.nIs there a way to switch off or control the deviation or the azimuthal angle for the impingement models? nnThanks for your help,nJon

[Surya Deb]

Hello, nI think you will need to make use of DEFINE_IMPINGEMENT and DEFINE_FILM_REGIME UDFs to specify your own user defined film regime. nPlease check the following links.nDEFINE_IMPINGEMENT : https://ansyshelp.ansys.com/account/secured?returnurl=/Views/Secured/corp/v211/en/flu_udf/flu_udf_sec_define_impingement.html?q=DEFINE_IMPINGEMENTnDEFINE_FILM_REGIME : https://ansyshelp.ansys.com/account/secured?returnurl=/Views/Secured/corp/v211/en/flu_udf/flu_udf_sec_define_film_regime.html. Please check the example which uses the macro Reflect_Particle to reflect the particles which hit the wall.You will need to set up a simple demo model and test the above two UDFs accordingly.nI hope this helps.nRegards,nSDn

[aronjo79]

Hi SD,nThanks for the response. I already tried the DEFINE_FILM_REGIME with FILM_USER_0 example in the Fluent UDF manual. nThe DEFINE_IMPINGEMENT UDF was used to call the FILM_USER_0 from the DEFINE_FILM_REGIME . The only film regime was FILM_USER_0, so that the user film criteria will be met always.nIn the DEFINE_FILM_REGIME with FILM_USER_0, the fluent function Reflect_Particle(tp,f_normal,ND_ND,f,t, NULL, NULL_FACE) resulted in Fluent MPI error. So I changed this function with a velocity reflect UDF to calculate the rebound/reflect droplet velocity without the deviation angle. I updated the droplet velocity at the current and previous time steps. However, the droplets did not rebound after impinging the wall and the trajectory stopped after hitting the wall. nPlease let me know if there is something I am missing.nThanks for your help,nJo

[Surya Deb]

Hi Jo, nDid you test in both Serial and Parallel runs? Can you check if the UDF works with a Serial run?nYou can also use a more generic reflect condition using the examples in the link below for DEFINE_DPM_BC. nYou need to be careful about the signs that are being used.nPlease check this link.nRegards,nSDn

[aronjo79]

Hi SD,nThe UDF with Reflect_Particle(tp,f_normal,ND_ND,f,t, NULL, NULL_FACE) crashes in both serial and parallel run.nThe reflect boundary condition using DEFINE_DPM_BC was first verified for just the DPM simulation . It worked without any issue.nHowever, when reflect UDF code was used in DEFINE_FILM_REGIME with FILM_USER_0 to simulate rebound, the droplets stopped after hitting the wall.nThe Eulerian Wall Film fate for single droplet parcel injected was nEulerian wall film particles: absorbed = 1, splashed = 0, rebound = 1nThough the fate was supposed to be only rebound , there is an absorbed droplet. I suspect the rebound = 1 is only showing because the n_rebound ++ . I guess the droplet is neither rebounding nor being absorbed, since the droplet trajectory stops and there is no film mass.nPlease let me know if there is a way to have only the rebound.nThanks,nJo

[Surya Deb]

Hello, nI am not sure why the Reflect_Particle macro is not working without specifically testing it for the case at hand.nBut you can copy the portion of the reflection from the regular DEFINE_DPM_BC and replace it with the Reflect_Particle macro, essentially doing the same thing. Please check consistency of the macros and the code once you copy.nIn theory, this should work well. The example of the DEFINE_FILM _REGIME UDF is for dey break up. So if the Reflect_particle macro is causing an issue, this can be a workaround. nRegards,nSDn

[aronjo79]

Hi SD,nDEFINE_FILM_REGIME with FILM_USER_0 and new rebound code (from DEFINE_DPM_BC) is working, but is not simulating the rebound. The droplet trajectory stops right after hitting the wall. The droplet mass is not absorbed into the Film mass. nThough the DPM-Film particle update shows both absorbed and rebound at the same time for the same droplet, nothing is updated within the impingement-Film model results.nThe Transcript on the console shows - Eulerian wall film particles: absorbed = 1, splashed = 0, rebound = 1nnPlease let me know if there is a workaround.nnThanks,nJo

[Surya Deb]

Hello, nI would recommend you to diagnose and debug in a systematic way. nWhat is the D_RATIO that you have used? Can you check by setting it to 1?nAlso I would recommend you to print out the mass of the particle at relevant locations in the UDF just to make sure if it is getting splashed, absorbed or reflected in a proper way.nDid you test this with different particle sizes to see if you observe the same behavior?nAlso make sure that your particle type is Inert as I believe the reflection using DEFINE_DPM_BC will only reflect particles of type Inert and will abort any tracking for other particle types.nRegards,nSDn

[aronjo79]

Hi SD,nYes, I tried to debug with all possible values in the UDF. The droplet just stops after hitting the wall.nI used D_RATIO of 1. It did not help. I removed the inert material condition for the particle rebound. Standalone DEFINE_DPM_BC works fine. nWhen I saved the velocity and diameter from the DEFINE_FILM_REGIME with FILM_USER_0, I can see the updated values. However, I also tried to save the values from the DPM scaler update, and the values are not updating there. The path stops after hitting the wall and using the DEFINE_IMPINGEMENT and DEFINE_FILM_REGIME UDFs.nnWere you able to get the particle rebound with the DEFINE_FILM_REGIME(dry_breakup, regime, tp, pp, f, t, f_normal, update) example in the UDF manual?Thanks,nJon

[Surya Deb]

Hi Jo, nI have not tested this at my end yet but let me try to create a simple test case. nI will let you know what I observe.Regards,nSDnn