Euler-Euler modeling of non-newtonian fluid in which software ??

navidoo · December 8, 2018, 07:16

Hi everybody
I am trying to simulate a non-newtonian fluid which transfers solid particles by COMSOL and I have several problems.
First of all, it is not prepared in COMSOL to select non-newtonian fluid in Euler-Euler model.
Second, my geometry is two 3D concentric cylinder which only the interior one rotates. In Euler model the option for wall movement disappears !
Now if you are familiar with Comsol or you have experience in such a simulation in other softwares ,would you please help me to solve it or do you advise to change my software?

CeesH · December 10, 2018, 08:54

Well, you might have to write your own code for that...

There is software that can do it, in principle. For example ANSYS FLUENT. But you are unlikely to get sensible results. In brief, bubbles will lead to accelerations (strain) at small scales (in their wake for example), which will modify your local viscosity on a scale much smaller than your grid. Non-newtonian viscosity models implemented in software such as FLUENT do not account for these small-scale strains, that may have a big impact on your rheology. Instead, it uses the strain-rates of the resolved scale, which are not representative and hence will lead to poor estimations of the true rheological behavior, and consequentially, multiphase interactions. The same applies for turbulence in non-newtonian flows, hence FLUENT has also disabled the use of non-newtonian rheology together with turbulence (the user can re-enable it, at own risk).

Apparently COMSOL chose to disable the combination of rheology and eulerian multiphase, since the models are not designed to work together and hence are highly likely giving garbage results if forced to. A lot of developments are still required to include the effects of unresolved strains originating from gas-liquid interaction and turbulence on rheology. For turbulence, some steps are being made (such as Garilov et al. 2014, turbulence model for power law fluids), but Eulerian multiphase + rheology (or better, multiphase + turbulence + rheology) is a highly unexplored field. If you are going to dive into it, do so carefully.

navidoo · December 10, 2018, 15:58

Thank you very much.
That was helpful.

December 8, 2018, 07:16	Euler-Euler modeling of non-newtonian fluid in which software ??	#1
navidoo New Member navid Join Date: Oct 2018 Posts: 3 Rep Power: 7	Hi everybody I am trying to simulate a non-newtonian fluid which transfers solid particles by COMSOL and I have several problems. First of all, it is not prepared in COMSOL to select non-newtonian fluid in Euler-Euler model. Second, my geometry is two 3D concentric cylinder which only the interior one rotates. In Euler model the option for wall movement disappears ! Now if you are familiar with Comsol or you have experience in such a simulation in other softwares ,would you please help me to solve it or do you advise to change my software?

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Modeling Solid and Fluid at the same using Interface BC..Help needed	Prince Jassal	Main CFD Forum	1	June 12, 2013 05:07
Solid Fluid interactions with non newtonian fluid	daniebae	ANSYS	12	May 27, 2013 03:57
Is 2-way FSI valve modeling with seperate fluid domains possible?	Maurosso	FLUENT	4	March 7, 2013 08:56
Modeling heat transfer fluid to solid	Elvis1991	FLUENT	6	December 7, 2012 09:08
fluid dynamics software sought for 3D models	Woon San Joe	Main CFD Forum	1	March 11, 2004 23:00

December 10, 2018, 08:54		#2
CeesH Senior Member Cees Haringa Join Date: May 2013 Location: Delft Posts: 607 Rep Power: 0	Well, you might have to write your own code for that... There is software that can do it, in principle. For example ANSYS FLUENT. But you are unlikely to get sensible results. In brief, bubbles will lead to accelerations (strain) at small scales (in their wake for example), which will modify your local viscosity on a scale much smaller than your grid. Non-newtonian viscosity models implemented in software such as FLUENT do not account for these small-scale strains, that may have a big impact on your rheology. Instead, it uses the strain-rates of the resolved scale, which are not representative and hence will lead to poor estimations of the true rheological behavior, and consequentially, multiphase interactions. The same applies for turbulence in non-newtonian flows, hence FLUENT has also disabled the use of non-newtonian rheology together with turbulence (the user can re-enable it, at own risk). Apparently COMSOL chose to disable the combination of rheology and eulerian multiphase, since the models are not designed to work together and hence are highly likely giving garbage results if forced to. A lot of developments are still required to include the effects of unresolved strains originating from gas-liquid interaction and turbulence on rheology. For turbulence, some steps are being made (such as Garilov et al. 2014, turbulence model for power law fluids), but Eulerian multiphase + rheology (or better, multiphase + turbulence + rheology) is a highly unexplored field. If you are going to dive into it, do so carefully.

December 10, 2018, 15:58		#3
navidoo New Member navid Join Date: Oct 2018 Posts: 3 Rep Power: 7	Thank you very much. That was helpful.