CFD Online Logo CFD Online URL
www.cfd-online.com
[Sponsors]
Home > Forums > General Forums > Main CFD Forum

Natural Convection Boundary Conditions

Register Blogs Members List Search Today's Posts Mark Forums Read

Like Tree1Likes
  • 1 Post By FMDenaro

Reply
 
LinkBack Thread Tools Search this Thread Display Modes
Old   September 19, 2018, 04:03
Default Natural Convection Boundary Conditions
  #1
New Member
 
Shyam Sundar Hemamalini
Join Date: Mar 2018
Posts: 3
Rep Power: 3
shyam97 is on a distinguished road
Hi all,

I am simulating 2D natural convection in an open vertical channel of high aspect ratio (12.5) with both sides isothermally heated by ΔT = 5K with water as the operating fluid. I have modelled the regime with large spaces on top and bottom of the channel to visualise inlet and outlet flow.

Gravity of 9.81 m/s^2 is applied in -Y direction. I have used Boussinesq approximation for density since β.ΔT<<1. Initial values of velocity and pressure in the regime are zero.

The boundary conditions I have specified are, as shown in figure 1, isothermal no-slip for the channel walls, adiabatic no-slip for the adjacent and side walls, velocity inlet with v=0, p=0 for the bottom edge and pressure outlet p=0 with flow normal to boundary for the top edge. The simulation performed is transient and laminar, using SIMPLE solver.

Am I correct in using the mentioned boundary conditions? The result agrees with experimental data with 7% error of Nusselt number and averaging out minimises it to 2%. I am confused as to why a pressure boundary condition is required for solving natural convection problems with Boussinesq approximation since the approximation takes out the pressure terms in Navier-Stokes equation and energy equation. The reduced set of equations are described in the Fundamentals of Heat and Mass Transfer by Incropera and Dewitt, shown in figure 2.

Wouldn't the velocity field and temperature field be enough to solve the set of equations implicitly?
Attached Images
File Type: jpg 2D model.jpg (32.3 KB, 15 views)
File Type: png Reduced Eqns.PNG (8.1 KB, 10 views)
shyam97 is offline   Reply With Quote

Old   September 19, 2018, 04:16
Default
  #2
Senior Member
 
Filippo Maria Denaro
Join Date: Jul 2010
Posts: 4,928
Rep Power: 52
FMDenaro has a spectacular aura aboutFMDenaro has a spectacular aura about
Incompressible flows (with Bousinnesq) requires BCs for velocity and temperature, the Dirichlet value for the pressure equation is not necessary.
However, if you fix the pressure outlet value, you have to let free the BC for the velocity. The pressure equation determines a pressure field up to a function of time.

Note that fixing a value for the pressure is a "trick" sometimes used to let the iterative method to converge but if the compatibility condition
shyam97 likes this.
FMDenaro is offline   Reply With Quote

Old   September 19, 2018, 04:33
Default
  #3
New Member
 
Shyam Sundar Hemamalini
Join Date: Mar 2018
Posts: 3
Rep Power: 3
shyam97 is on a distinguished road
Quote:
Originally Posted by FMDenaro View Post
Incompressible flows (with Bousinnesq) requires BCs for velocity and temperature, the Dirichlet value for the pressure equation is not necessary.
However, if you fix the pressure outlet value, you have to let free the BC for the velocity. The pressure equation determines a pressure field up to a function of time.

Note that fixing a value for the pressure is a "trick" sometimes used to let the iterative method to converge but if the compatibility condition
I tried using pressure inlet and pressure outlet (both p=0) but the velocity diverged to a large value after a few time steps to overwhelm buoyancy driven flow. In that case, temperature gradient near the hot wall was almost non-existent.

Yes, your "trick" certainly worked. The combination of velocity inlet + pressure outlet sort of fixes the direction for the solution to proceed towards actual solution and the flow is similar to experimental data.

But I'm confused why pressure inlet + pressure outlet won't work and why velocity inlet + pressure outlet works.
shyam97 is offline   Reply With Quote

Old   September 19, 2018, 05:07
Default
  #4
Senior Member
 
Filippo Maria Denaro
Join Date: Jul 2010
Posts: 4,928
Rep Power: 52
FMDenaro has a spectacular aura aboutFMDenaro has a spectacular aura about
Quote:
Originally Posted by shyam97 View Post
I tried using pressure inlet and pressure outlet (both p=0) but the velocity diverged to a large value after a few time steps to overwhelm buoyancy driven flow. In that case, temperature gradient near the hot wall was almost non-existent.

Yes, your "trick" certainly worked. The combination of velocity inlet + pressure outlet sort of fixes the direction for the solution to proceed towards actual solution and the flow is similar to experimental data.

But I'm confused why pressure inlet + pressure outlet won't work and why velocity inlet + pressure outlet works.

You can set pressure inlet and outlet (leaving free the velocity) but you have to prescribe a pressure difference between inlet and outlet that, for buoyancy-driven flow depends, on the temperature difference that induces the flow.
Have you tried using Neumann condition for the pressure everywhere?
FMDenaro is offline   Reply With Quote

Old   September 19, 2018, 06:00
Default
  #5
New Member
 
Shyam Sundar Hemamalini
Join Date: Mar 2018
Posts: 3
Rep Power: 3
shyam97 is on a distinguished road
Quote:
Originally Posted by FMDenaro View Post
You can set pressure inlet and outlet (leaving free the velocity) but you have to prescribe a pressure difference between inlet and outlet that, for buoyancy-driven flow depends, on the temperature difference that induces the flow.
Have you tried using Neumann condition for the pressure everywhere?
But isn't the pressure difference between inlet and outlet handled by specifying gravity, thereby specifying the hydrostatic pressure difference hρg? I don't see any other form of pressure other than hydrostatic, and the buoyancy-induced localized pressure difference.

No, I did not specify Neumann condition for pressure anywhere directly. The no-slip BC at the walls imply a zero normal pressure gradient. Other than that, inlet is a Dirichlet of velocity and pressure, outlet is a Dirichlet of pressure alone.
shyam97 is offline   Reply With Quote

Reply

Tags
boundary condition, boussinesq approximation, natural convection

Thread Tools Search this Thread
Search this Thread:

Advanced Search
Display Modes

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are On
Refbacks are On


Similar Threads
Thread Thread Starter Forum Replies Last Post
My radial inflow turbine Abo Anas CFX 27 May 11, 2018 02:44
Centrifugal fan-reverse flow in outlet lesds to a mass in flow field xiexing CFX 3 March 29, 2017 11:00
Boundary conditions for external natural convection (chtMultiRegionFoam) Coris OpenFOAM Running, Solving & CFD 4 August 11, 2016 12:33
Waterwheel shaped turbine inside a pipe simulation problem mshahed91 CFX 3 January 10, 2015 12:19
boundary conditions for Natural Convection problem Nav CFX 13 June 6, 2011 08:37


All times are GMT -4. The time now is 20:48.