CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   FLUENT (http://www.cfd-online.com/Forums/fluent/)
-   -   How to choose a good model for multiphasis flow ? (http://www.cfd-online.com/Forums/fluent/47916-how-choose-good-model-multiphasis-flow.html)

Yasmail AKARIOUH April 15, 2008 03:40

How to choose a good model for multiphasis flow ?
 
hello

i wish to simulate a truck in a multiphasis flow with a 2D geometry. the principal phase is air and the other one is water liquid. the multophase model is mixture. the problem with this modelisation is that the turbulent viscosity increase enormously. indeed, there is a swirl which is created by the truck. residuals converge for very little water particles ( average 1e-5m). however, residuals become "crazy" for larger water particles. i have modified under-relaxation number but the calculus is diverging. i have tested this model with k-espilon and k-omega viscous model. all of them diverge.

some one has ever done a simulation like mine ?

do you have any idea about dirverging reasons ?

thank you for your answers.

yasmail

vk April 18, 2008 12:52

Re: How to choose a good model for multiphasis flo
 
you may try transient run with large time step instead od steady state. Hope this may give better results.

Yasmail AKARIOUH April 21, 2008 02:12

Re: How to choose a good model for multiphasis flo
 
time step increasing doesn't any result. the matter of this model is the water-liquid drop size. indeed when i use a diameter inferior to 10^-5 m, residuals converge. however for a 10^-3 m drop the model is totally instable. it's due to the increasing of the turbulent viscosity. this message appears during calculation " turbulent viscosity limited to viscosity ratio of 1.000000e+05 in 202602 cells". i ever try to reduce the under-relaxation number for viscosity, but it does not work. i submit this problem to fluent-support, but they do not give me an answer. they probably hope i will forget.


All times are GMT -4. The time now is 05:37.