[Thermal_engg]

Hey everyone,
I am trying out a simulation where a solid copper rod is provided a constant heat flux and the rod is kept open to the atmosphere,so that natural convection can be set-up.Geometry making and meshing were fine but while simulating, with every iteration, I get a message of "Reversed Flow in ... cell".
Where might I be going wrong? ....Thnx

[vasava]

The ''Reversed flow'' message is common if you have a flow that is recirculating or non-uniform with in the domain. The message also means that in some of the mesh elements there are faces with flow direction outwards as well as inwards.

In my opinion this is completely natural and nothing to worry.

[Thermal_engg]

While simulating, for a few iterations everything's fine,but after that, a message starts appearing with each iteration ""Temperature limited to 5000k" !!??

What kind of error in giving conditions causes such situtation?
Btw, the temperature value I am working with is 350 K. Thnx

[vasava]

The error is fine during early phase of the simulations. When your initial conditions are not feasible Fluent keeps the calculation going by limiting the parameter (in your case the temperature). The error should go away once as the simulation progresses.

This error is common for areas with bad mesh. You could look where this temperature occurs and the improve the mesh quality.

[upeksa]

If you have a pressure-outlet condition, it's quite common, particularly in some of the first iterations, that you have some "reversed flow" message.

If you want to minimize that, in the termal options of your pressure-outlet boundary, you should set up a realistic reverse flow temperature. I mean: if you think that the outlet temperature will be about 400 K, more or less, that will be your reverse flow temperature. Fluent will use that temperature only in order to calculate those reverse flow. Since you will have reverse flow in the first iterations, that temperature will have nothing to do with your final results. But the more accurate temperature you set, the quicker it gets a solutions.

Don't worry, it's usual, it's not actually a big deal.

Best regards.

[Thermal_engg]

Ok, here's what I did.Initially, I am trying to solve this problem using 2-d mesh.
So I took the cross-section of the cylinder and halved it ,since the flow will be symmetrical.so I got 2-semi-circles.Now I simply meshed it and set up the bondary condns for various zones.For the fluid domain, I chose "pressure outlet" condition.Even after doing so,the simulations I am doing shows "0" velocity in the console with each iteration.
I am attaching a picture of the geometry.
any help is appreciated.Thnx

[Thermal_engg]

After setting some conditions for natural convection, I tried simulating but I have been getting this error : "Error: Divergence detected in AMG solver: pressure correction" .I tried reducing the under-relaxation factors for pressure but it doesn't help.Could anyone suggest some solution please?

[vasava]

Can you share your simulation+CAD+Mesh? We can have a look.

And have you tried mesh adaption?