Learning Forum

[kiki1]

I need to calculate turbulent natural convection in an enclosed space

Initially, the fluid in space is at a constant temperature

Using the k-epsilon model to deal with turbulent flow

When initializing, it is necessary to determine the turbulent kinetic energy and the turbulent dissipation rate. The relevant formulas all contain the velocity. At this time, should both be set to 0?

[RK]

Hello,
Refer to this tutorial on natural convection : Natural Convection in a Square Cavity Simulation Example | Ansys

[kiki1]

[kiki1]

Thank you for your reply. I'm sorry that I didn't describe the problem of my calculation clearly. What I want to do is a transient calculation. The initialization method of this example is suitable for the steady state, is it also desirable for the transient state?

[Rob]

What do you want the starting (and end) points to represent? In steady we model the equilibrium (time = infinity) condition, in transient it's a little more complex.

[kiki1]

For a fluid in a closed space, I only know its initial temperature, and I can determine its flow state through the boundaries and assumed velocities. I need to calculate the flow field after a period of time, how should I set the initial turbulence parameters, maybe I should first calculate a stable initial field and then further calculate the transient process based on it?

[Rob]

That's one approach. It depends on what starting point you want.
The example I use in training is filling a bucket. If we want to see how the bucket fills we start with an empty bucket and add water. If we want to see a bucket overflowing we'd patch the bucket to around 95% full and then begin the water flow, by the time it's overflowed the initial flow in the bucket shouldn't have too much of an effect on the overflow.
I can't tell you the "best" option for your system, you need to figure out what you want to do, and how it starts.

[kiki1]

Thank you very much for your wonderful reply´╝ü

[Rob]

You're welcome.