Simulation of Rayleigh–Benard convection

caja94 · November 5, 2018, 04:20

Hello guys, I wanted to simulate Rayleigh-Benard convection using buoyantBoussinesqPimpleFoam solver.
For geometry i have a cylinder, and I was using gmsh for geometry and mesh creation. Plate on top of that cylinder is at 350K, and the bottom one is at 274 degrees Kelvin.
All case files (0, system, constant) are uploaded and the link is bellow.

Now, simulation is successful, but after I open Paraview to see results, I have one problem. Velocity and pressure fields are generated, and they change over time (10 time steps in this case). They are not the best, because I still need to work on the mesh, but I get some results. The problem comes when I want to see temperature change in cross-section. After 10 time steps, the temperature doesn't change - it is the same on top and bottom plate, and temperature of the fluid doesn't change at all.
Here is how that temperature field looks at time step 10:

I get no errors during simulation, and I really can't figure out why temperature field doesn't change.

Here is my case file:
https://ufile.io/9ngvx

Help would be appreciated. If you have any solution to this problem I would like to hear it.

Best regards,

Petar

arun1994 · April 15, 2019, 09:30

For rayleigh Bénard effect to happen, you need to have a hot plate at the bottom and a cold plate at the top. That is the first thing and the second, you need to run the simulation for a longer time if this doesn't work.

caja94 · April 16, 2019, 02:39

Hey arun, thank you for your answer.

The reason why I put a hot plate at the top is because we had a PhD thesis at our faculty that deals with this case, so I had experimental measurements and simulations in PHOENICS. Everywhere I look, it says that RB convection is with hot plate at the bottom, just as you said. But for some reason, they call that reversed RB at our university.

I did, however, run simulations with more time steps some time ago, and I got some results. I was not satisfied with those, so I'm planning on improving a simulation when I have a little bit of free time.

Thank you for your kind reply.

Best regards,
Petar

November 5, 2018, 04:20	Simulation of Rayleigh–Benard convection	#1
caja94 New Member Petar Cajic Join Date: Jul 2018 Location: City of Nis, Serbia Posts: 5 Rep Power: 7	Hello guys, I wanted to simulate Rayleigh-Benard convection using buoyantBoussinesqPimpleFoam solver. For geometry i have a cylinder, and I was using gmsh for geometry and mesh creation. Plate on top of that cylinder is at 350K, and the bottom one is at 274 degrees Kelvin. All case files (0, system, constant) are uploaded and the link is bellow. Now, simulation is successful, but after I open Paraview to see results, I have one problem. Velocity and pressure fields are generated, and they change over time (10 time steps in this case). They are not the best, because I still need to work on the mesh, but I get some results. The problem comes when I want to see temperature change in cross-section. After 10 time steps, the temperature doesn't change - it is the same on top and bottom plate, and temperature of the fluid doesn't change at all. Here is how that temperature field looks at time step 10: I get no errors during simulation, and I really can't figure out why temperature field doesn't change. Here is my case file: https://ufile.io/9ngvx Help would be appreciated. If you have any solution to this problem I would like to hear it. Best regards, Petar

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
continuity divergence in transient natural convection simulation	wj_wanjie	FLUENT	10	August 9, 2021 05:43
Forced convection between hot air and solid simulation in fluent	calkobtawy	FLUENT	4	May 11, 2018 14:55
Modeling both radiation and convection on surfaces - Ansys Transient Thermal R13	s.mishra	ANSYS	0	March 31, 2012 04:12
Natural Convection Simulation - buoyantSimpleRadiation - Convergence Problem	msarkar	OpenFOAM	32	June 16, 2010 06:27
transient simulation: natural convection problem?	Basics	CFX	3	September 25, 2002 09:42

April 15, 2019, 09:30		#2
arun1994 Member Arun subramanian Join Date: Jun 2016 Location: Florence,Italy Posts: 48 Rep Power: 9	For rayleigh Bénard effect to happen, you need to have a hot plate at the bottom and a cold plate at the top. That is the first thing and the second, you need to run the simulation for a longer time if this doesn't work.

April 16, 2019, 02:39		#3
caja94 New Member Petar Cajic Join Date: Jul 2018 Location: City of Nis, Serbia Posts: 5 Rep Power: 7	Hey arun, thank you for your answer. The reason why I put a hot plate at the top is because we had a PhD thesis at our faculty that deals with this case, so I had experimental measurements and simulations in PHOENICS. Everywhere I look, it says that RB convection is with hot plate at the bottom, just as you said. But for some reason, they call that reversed RB at our university. I did, however, run simulations with more time steps some time ago, and I got some results. I was not satisfied with those, so I'm planning on improving a simulation when I have a little bit of free time. Thank you for your kind reply. Best regards, Petar