[mannobot]

Hi,

is it possible to define different viscosities within one domain? I have a model of different chambers and would like to define different behaviour of the fluid in each chamber.

Any help would be much appreciated.

Sincerely

[-mAx-]

Do you have interaction with the 3 fluids? (if yes then you have to use multiphase model)
You can create 3 fluid materials and attribute each one to the desired fluid domain. That means each fluid domain has to be identified separately in your mesh program:
*fluid_domain1 set with fluid1
*fluid_domain2 set with fluid2
*fluid_domain3 set with fluid3

[mannobot]

Hi -mAx-,

thanks for the quick reply. Using the multiphase model is not really common to me. I worked within turbine industry and solely dealt with gas. So in the end FLUENT will calculate the mixture on its own? Assuming I do have an outlet to the environment and I want to track the fluid after exiting the domain do I just have to mesh the outside too, define boundaries and use VOF to get a phase-interphase? What would be an appropriate boundary at the outside? Assuming steady state calculation and a massflow inlet would result into what kind of flow field. How to restrict the domain? For sure there will be needed a plate where the fluid impinges on but what are clever lateral boundaries?

Sincerely

[-mAx-]

can you post a hand sketch of what you want ton compute.

[mannobot]

I have to excuse my unclear definition. The internal is no longer of interest because I think the multiphase solution will make it. In the end the state at the outside is the big unknown. I use ICEM for grid generation. I already modeled the inside of the slot and the cavities which do contain the fluids. Now I want to analyse the behaviour after exiting the slot. The image I uploaded shows the slot, the fluid, and the moving plate.
There are several questions. Do I have to generate a new domain using the solution I obtained at the outlet from the previous calculation as inlet boco or should I discretize the entire environment at once? How would it be possible to define the output from one calculation as input for another?
The other question is how to define shear dependent viscosity within FLUENT. Do I have to write an UDF?

Thank you very much for your advice

[-mAx-]

I think it would be better to treat all the domain with the 3 fluids.

[mannobot]

Thank you. I will try. But what kind of bocos would you recommend at the outside. I will model a cube around the entire slot. Pressure outlet with environmental conditions?

Kind regards

[-mAx-]

where flow the fluids out?

[mannobot]

It would be the same problem modelling a water tap. Where do I have to end the domain surrounding the tap? Assuming I am interested in the behaviour of the flow field close to the point where the fluid impinges. For example the tap should be of diameter 0.02m and the region of interest should be of diameter 0.2m. I would go for a cube with a hole set to massflow inlet. But what about the other faces? Isothermal wall at the downside of the cube but I am not able to set the lateral boundaries (See Figure).
The next thing is how to define different domains. Within my Slot I do have fluid. After exiting there is air too. Do I have to do that in ICEM?

Thanks for your guidance

[-mAx-]

I still don't understand where your fluids flow out.
Do the fluids stay in the cube?
If yes, the I assume your cube is initially filled with air, and then you need to let the air flows out (for a constant volume)
I cannot help you regarding ICEM since I am using Gambit

[mannobot]

There are no walls. The slot is surrounded by environment like watering flowers in the garden using a watering can. But I have to restrict the domain. Otherwise I would get an infinite amount of cells. That is my problem.
I think the other problem does not depend on the mesher. Would you define different domains in Gambit or FLUENT?

Kind regards

[-mAx-]

ah ok...
Then I assume you have to apply someting like pressure outlet (check user guide)far away from the domain you are interessed. It must not affect the solution in this area.
I would define define different domain in Gambit

[mannobot]

Super. I will give it a try. And report about the progress.

Thank you very much.