How can I get natural convection to run?

Numerical methods and mathematical models of Elmer
Post Reply
hennig
Posts: 4
Joined: 12 Jun 2021, 20:51
Antispam: Yes

How can I get natural convection to run?

Post by hennig »

Hi,

I'm trying to calculate the warming of an eletric cable due to joule loss. This cable is laid in an air filled plastic pipe and, hence, heat transfer via conduction, radiation and natural convection must be considered.

While the calculation of the heat trasfer via radiation and conduction seems to work, I have trouble with getting the natural convection to run.

Maybe someone can give me some advice, how to get the natural convection to run?

Best regards

-Carsten
Attachments
SolverLog.txt
(9.24 KiB) Downloaded 18 times
case.sif
(4.24 KiB) Downloaded 15 times
cable_in_pipe_4.msh
(619.86 KiB) Downloaded 19 times
kevinarden
Posts: 1210
Joined: 25 Jan 2019, 01:28
Antispam: Yes

Re: How can I get natural convection to run?

Post by kevinarden »

I don't think you need radiation on all of your boundaries. The boundaries that share nodes between materials will transfer heat through the shared nodes. The diffused boundary transfers heat out of the boundary and is received by another material that does not share nodes. The idealized radiation transfers heat through the boundary that is not received by another body. For the stokes to have heat convection there needs to be temperature differential and flow boundary conditions on he boundaries of the fluid (in this case no slip condition).
Attachments
case.sif
(4.18 KiB) Downloaded 20 times
velocity.png
(146.45 KiB) Not downloaded yet
hennig
Posts: 4
Joined: 12 Jun 2021, 20:51
Antispam: Yes

Re: How can I get natural convection to run?

Post by hennig »

Hi kevinarden,

Thank you very much for your advices regarding initial temperatures and boundary conditions for navier-stokes (in this case no-slip). Now, the natural convection works fine.

However, I'm still elaborating on the question of using radiation between the surfaces or not. If I calculate with heat transfer via radiation between outer surfaces of the cables and inner surface of the pipe I see a cable conductor temperature of 343 K. Without considering radiation the result is 356 K.

As the result of the calculation incl. radiation is closer to analytic calculatios results I'm tending to consider this as the correct model. But I'm happy to discuss.

BR -Carsten
Attachments
without_radiation.png
(159.88 KiB) Not downloaded yet
with_radiation.png
(159.27 KiB) Not downloaded yet
kevinarden
Posts: 1210
Joined: 25 Jan 2019, 01:28
Antispam: Yes

Re: How can I get natural convection to run?

Post by kevinarden »

If the outer surface of the cable and the inner surface if the the pipe was frozen ice at 0 degrees or 200 degrees gas would you get the same answer with your analytical solution and Elmer analysis?
Post Reply