Heat Equation with Convection

General discussion about Elmer
kevinarden
Posts: 2237
Joined: 25 Jan 2019, 01:28
Antispam: Yes

Re: Heat Equation with Convection

Post by kevinarden »

No, Elmer will not take in to account buoyancy of the surrounding air that is not modeled. If the surrounding air is modeled, than it will only include buoyancy if if the stokes equation is included for the surrounding air.
kevinarden
Posts: 2237
Joined: 25 Jan 2019, 01:28
Antispam: Yes

Re: Heat Equation with Convection

Post by kevinarden »

maguirre
Posts: 17
Joined: 03 Nov 2022, 17:59
Antispam: Yes
Location: Buenos Aires, Argentina.

Re: Heat Equation with Convection

Post by maguirre »

Hi everyone!

As you recommended, I added Air to the problem, but new questions arose. Now I have a box of air surrounding a copper sheet, I'll attach the .SIF file for you to check and this is the mesh: https://drive.google.com/file/d/1jRlJMu ... sp=sharing

I added two equations, one for the Air Box (AB) with "Heat Solver" with "convection" as "computed", and "Navier-Stockes Solver" with "convect" on, and another for de Sheet (S) with "Heat Solver" with "convection" as "computed" and "Stat Current Solver" with "Calculate Joule Heating" on.

As body forces, I assigned one for the AB with just the Boussinesq on, and another for the S with Joule Heating on.

As Init. Cond. I assigned 300 K and 101325 Pa for the AB, and 300 K for the S.

And finally, BC I assigned 4 different ones. One is for the "Laboratory Wall" (the faces of the AB) with 300 K and Noslip wall BC on, then two more with the voltage difference for the S, and one more for convection between the S and the AB.

So, here is the main problem, the AB is 30x30x30 at atmospheric pressure (101325 Pa) but at the start of the simulation the pressure distribution it's quite bizarre as you can see in the picture attached. Not only it is far from atmos value but also is greater above than below (the gravity unity vector is 0 0 -1 9.81). I changed quite a few parameters and I can not find what the problem is.

Regards, Matias.
Attachments
pressure_dist.jpeg
(24.05 KiB) Not downloaded yet
case.sif
(4.96 KiB) Downloaded 40 times
Lic. Matias A. Aguirre.
Metrología Física, Metrología y Calidad, INTI.
Avenida General Paz 5445 Edificio 3
B1650WAB San Martin, Buenos Aires.
Rich_B
Posts: 421
Joined: 24 Aug 2009, 20:18

Re: Heat Equation with Convection

Post by Rich_B »

Hello,

The last case.sif is missing the 'Coordinate Scaling = 0.001' statement, which was present in the first sif file.

After running the case with the scaling, it looks like both the StatCurrentSolver and the HeatSolver are reaching 501 iterations without converging. Maybe increase the number of iterations and lower the tolerances.

Rich.
kevinarden
Posts: 2237
Joined: 25 Jan 2019, 01:28
Antispam: Yes

Re: Heat Equation with Convection

Post by kevinarden »

I did some test runs. It appears the default pressure at the boundary is zero. When the boundary pressure is zero any initial pressure is quickly lost. Adding pressure to the boundary retains the initial pressure.
Post Reply