[Seeth]

Hi everyone!
I am new to CFD and Star CCM+ and trying to create a simulation of the heating process of a water tank. So far i have taken a surface model of the tank interior and extruded the internal volume. I have then split the extruded volume into 2 parts to have 1 volume for the water and 1 volume for the air above. I have also split the tank surface into a general interior surface and the heating module surface.
I have created a mesh continuum for the tank surface (only surface remesher) and a mesh continuum for the water and air (surface remesher, prism layer mesher, polyhedral mesher). I have also created a physics continuum for both water and air. Which physics models would be best to choose here?
Do i have to set up a physics continuum for the tank too (i have set one up but do i need it since the tank is only modeled as a surface?)?
Now i would like to set up the heating module surface as a constant heat source (1000 W) and set up the water and air boundaries so that i can see the convection currents in the water, at the water/air interface and in the air.
The heat convection through the solid wall of the tank doesn't really matter to me.
That´s really what I am struggling to do right now, any help is appreciated, thank you very much! I am adding a picture of my setup right now to help a bit with clarity.

[me3840]

No you don't have to model the solid of the tank. Just apply your heat condition to the appropriate surface of the fluid regions. You can delete the tank region, it's not needed.

[Seeth]

Thank you very much, I did that and it seems to work so far. Just have to optimize the mesh I think. Do you know the differences between the convection properties? "2nd" was setup as default, does that make sense in my case?

[me3840]

I think you're referring to the discretization scheme? In that case, yes 2nd order is what you want.

However I will add if you're not familiar with the basic discretization schemes and how they work, you need to do some reading on CFD basics. It's dangerously easy to get results out of a CFD code that are meaningless if you're unsure of what you're doing.

[Seeth]

Yeah, that's what I meant, thank you.
Yes, I know, I am doing that at the same time while working on this. I have a real model of the water tank to validate my results so that will definitely help me.

Since you helped me a lot already, could you give me your opinion on my physics continuum for the water region?

So far i have chosen:
Three Dimensional, Implicit Unsteady, Segregated Flow, Segregated Fluid Temperature, Liquid, IAPWS-IF97 (Water), Cell Quality Remediation, Laminar, Gradients, Gravity

Do i need the Boiling model even though I am not trying to reach a boiling temperature?

[me3840]

It's probably a better idea to do a steady simulation if you're new, they are more forgiving. It's very easy to get bad results from transient if you're not sure what you're doing.


If you don't expect boiling then no, you don't need boiling.

[Seeth]

Okay but my goal is to measure the time necessary to heat the water up to a certain temperature. Don´t I need a transient simulation for that purpose?
Thanks again for the help

[me3840]

Yes, you do need transient simulation for that. However transient simulations are expensive, hard to debug, and unforgiving for new users. You will find you can eliminate mistakes and learn much faster with a steady simulation. When you feel you have a grasp on things then move to transient.

It might also be a good idea to practice your transient simulation on a 2D or coarse 3D mesh (100k cells or so). It will take significantly less time to learn to do things properly that way. Make sure you understand how your solver settings, timestep, and number of inner iterations are affecting your solution, and learn to judge when timesteps are achieving enough convergence.

You will find it very easy to waste time doing long transient simulations like the one you want to do if you are new to things. Best start small and work up.

[Seeth]

Yes, that sounds very reasonable, I will try to follow that kind of workflow.
Thank you very much for all the good advice, it definitely helped me a lot so far!