[oliveira1820]

Hello everybody,

I am doing a simulation with the purpose to study the heat transfer between different fluid and solid domains, but my doubt is:

What I´m studying is surrounded by air and of course this will interfere with heat transfer. So what should I do?

Should I consider a big fluid domain with air, where the object is as a solid domain inside the big air domain;

Should I consider a big fluid domain with air but with a empty space that
matches the geometry of my object;

Should I not consider the air around;

Thank you,

Pedro Oliveira

[urosgrivc]

heat transfer from solid to fluid can easily be modeled.
It depends on vhat you vant to simulate:
-Is the case steady state or transient,
-maybe you just vant to get the HTC value,
-do you vant to knov the temperatures of your body, or are these knovn,
- is it forced convection or natural,
-...
You didnt give any details so the ansver is;
yes it can be done but it depends on the case.

[oliveira1820]

It is a transient satet study, and I know how to study heat transfer.

I´m just trying to know how do I consider the air surrounding and include it like this:

https://www.youtube.com/watch?v=vLB71LM9gg8&t=974s

[Gert-Jan]

Do you mean that the solid is hanging in air and is losing heat to the surroundings (or heated by the surrounding air)? And you want to include this heat transfer, without modelling the whole surrounding flow process?

Mostly heat transfer can be ignored since effects are small, but that depends on your exact case which I don't know.
If you want to include it anyway without modelling the whole air space, then you should use combined Neumann-Dirichlet-BC on the outer wall of your solid.
Meaning, use a heat transfer coefficient and a temperature at infinite distance.

As an example, imagine a cube of metal hanging in air of 10°C. Then set 10°C for Temp at infinite distance and use heat transfer coefficients for bottom, top and side walls.

As a rule of thumb, take h=2 W/m2K for the bottomsurface, 7 for the side walls and 5 for the top. This all depends a bit on the conditions.

As mentioned earlier, mostly heat loss or pickup will be small, unless temperature of your solid is very high. Then you might need to include radiation. Just perform a few hand calculations to estimate their effects.

[urosgrivc]

heat disipated in to the enviroment is ofcourse wery important.
imagine a microchip with only 0.1W TDP.
if no heat would be dissipated the temperature would theoreticaly be infinite in steady state.
It depends on the phisics of your problem, vhat your BCs are and all, and I dont exactly understand what your question is if you know how to do it already

[Gert-Jan]

If I study a heat exchanger of 1 MW, then mostly no one is really interested in the heat loss. Only the safety guys who don't want people to burn their fingers.

[sarathaj96]

I had the similar query, where my geometry is inside the air domain and it is a natural convection. Applying pressure outlet at the outer surface of the air domain removes lot of heat from the geometry.

is it right approach to apply HTC and temperature on the outer surface of the air domain ?

[ghorrocks]

is it right approach to apply HTC and temperature on the outer surface of the air domain ?

That depends on what you are modelling and what you are trying to achieve. Please describe what you are trying to do and show some images and we can comment on how the model should be set up.

[sarathaj96]

My Apology, I cannot share the images.
My objective is to monitor the temperature rise for a certain period of time.

Detail about my Geometry
I am simulating a Motor using MRF on the rotor region running at 5000 RPM. I am using a sphere shape air domain to monitor the air flow around the geometry.

First step I ran steady state ( energy equation off) to get flow converged then I applied constant heat loss to the few components and switch to transient analysis by keeping energy on ( flow and turbulent equation off).

Importantly I applied pressure outlet BC on the outer surface of the sphere shaped air domain

[ghorrocks]

This could be due to many things:
* Inaccurate simulation
* Poor convergence
* Not running long enough time
* Boundary condition too close to area of interest
* It is correct and your expectation is wrong
* Wrong boundary condition
* Boundary condition needs to be in a different place