boundary layer

Arvin · October 23, 2013, 10:10

Hey guys, I am really stuck here. I am new to fluent and I am modelling a steady state 2d tank for natural convection problem. Size of the tank being 100 by 100 mm. I have attached a boundary layer as well as face mesh. Total no. of nodes is 1000000. Do I need to check internal continuity?. Despite many attempts, I am not able to get converged solution, not even the plumes from bottom surface. My boundary conditions are 1.Heat flux is supplied at bottom plate. 2.Other three sides are kept at ambient temperature. Working fluid is water. Please do help me out. I want to know the settings that needs to be applied in fluent. Boussinesq or Polynomial for water? Standard or PRESTO? Any other important things to be set?
Thanks. I do hope I get help as soon as possible as I am running out of time.

delaneyluke · October 24, 2013, 10:38

Could you be more specific about what's not converging.
What have you tried so far to improve convergence?
Natural convection is an unsteady phenomenon, have you tried a transient analysis?

Regards
Luke

Arvin · October 24, 2013, 10:55

No, I have not yet tried transient analysis since I am not getting convergence on steady state itself. I have a strong doubt on using Boussinesq approximation. Can you guide me on this?
Thanks.

delaneyluke · October 28, 2013, 15:11

Well the reason you dont get convergence with a steady state solver is because the phenomenon is transient. A transient phenomenon will not converge with a steady state solver.
Whats your question about the Boussinesq approx

Regards
Luke

Arvin · October 29, 2013, 05:42

Hi, thanks Luke. My question is obviously we need to specify density as a function of temperature. And we can do that either by specifying density as Boussinesq approx or by using polynomial and other piecewise functions. But for natural convection problem, generally Boussinesq approx is used. Why is that? It can be used in phenomena involving small temperature changes, so how do we know that? So, how to select the appropriate function of density?
Secondly what is the difference between pressure based solver and density based solver? For natural convection problems why can't we use density based solver?
Thanks.

October 23, 2013, 10:10	boundary layer	#1
Arvin New Member Arvindh Join Date: Oct 2013 Posts: 19 Rep Power: 13	Hey guys, I am really stuck here. I am new to fluent and I am modelling a steady state 2d tank for natural convection problem. Size of the tank being 100 by 100 mm. I have attached a boundary layer as well as face mesh. Total no. of nodes is 1000000. Do I need to check internal continuity?. Despite many attempts, I am not able to get converged solution, not even the plumes from bottom surface. My boundary conditions are 1.Heat flux is supplied at bottom plate. 2.Other three sides are kept at ambient temperature. Working fluid is water. Please do help me out. I want to know the settings that needs to be applied in fluent. Boussinesq or Polynomial for water? Standard or PRESTO? Any other important things to be set? Thanks. I do hope I get help as soon as possible as I am running out of time.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Prism Layer (Boundary Layer) Thickness	Peter-27	STAR-CCM+	12	March 27, 2022 13:25
[snappyHexMesh] No layers in a small gap	bobburnquist	OpenFOAM Meshing & Mesh Conversion	6	August 26, 2015 10:38
Grid Independence of Boundary layer	Luigi_	Main CFD Forum	0	December 14, 2011 14:42
Boundary Layer Question	scottneh	STAR-CCM+	3	September 30, 2010 15:21
[Commercial meshers] Trimmed cell and embedded refinement mesh conversion issues	michele	OpenFOAM Meshing & Mesh Conversion	2	July 15, 2005 05:15

October 24, 2013, 10:38		#2
delaneyluke Senior Member Join Date: Mar 2009 Location: Indiana, US Posts: 186 Rep Power: 17	Could you be more specific about what's not converging. What have you tried so far to improve convergence? Natural convection is an unsteady phenomenon, have you tried a transient analysis? Regards Luke

October 24, 2013, 10:55		#3
Arvin New Member Arvindh Join Date: Oct 2013 Posts: 19 Rep Power: 13	No, I have not yet tried transient analysis since I am not getting convergence on steady state itself. I have a strong doubt on using Boussinesq approximation. Can you guide me on this? Thanks.

October 28, 2013, 15:11		#4
delaneyluke Senior Member Join Date: Mar 2009 Location: Indiana, US Posts: 186 Rep Power: 17	Well the reason you dont get convergence with a steady state solver is because the phenomenon is transient. A transient phenomenon will not converge with a steady state solver. Whats your question about the Boussinesq approx Regards Luke

October 29, 2013, 05:42		#5
Arvin New Member Arvindh Join Date: Oct 2013 Posts: 19 Rep Power: 13	Hi, thanks Luke. My question is obviously we need to specify density as a function of temperature. And we can do that either by specifying density as Boussinesq approx or by using polynomial and other piecewise functions. But for natural convection problem, generally Boussinesq approx is used. Why is that? It can be used in phenomena involving small temperature changes, so how do we know that? So, how to select the appropriate function of density? Secondly what is the difference between pressure based solver and density based solver? For natural convection problems why can't we use density based solver? Thanks.