# Problems obtaining the correct contours for Lid driven cavity

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

February 3, 2016, 11:56
Problems obtaining the correct contours for Lid driven cavity
#1
New Member

Join Date: Feb 2016
Posts: 6
Rep Power: 9
Hi

I am solving the problem of lid driven cavity in C++. I have used the Finite volume discretization and I have used the Semi Explicit method for solving the problem. However, I am obtaining results that do not match with that of the ones already existing.

The results I plotted are about symmetric about the central vertical line where in that should not be the case. I tried running the code for a finer mesh and higher reynolds number only to encounter the same problem.

I checked my boundary conditions and they are as follows.

u, v = 0 at all the walls except at the top where u = 1
dp/dn = 0 at all the walls
dp'/dn = 0 at all the walls
p' = pressure correction

I have attached the U and V velocity contours and also the pressure contours that I have obtained.

Can someone help me figure out the problem? Any help will be appreciated
Attached Images
 pressure_smaple.jpg (24.8 KB, 11 views) u_velocity_smaple.jpg (28.9 KB, 14 views) v_velocity_smaple.jpg (30.8 KB, 11 views)

 February 3, 2016, 11:58 #2 Senior Member   Filippo Maria Denaro Join Date: Jul 2010 Posts: 6,390 Rep Power: 67 what about the Re number? what main differences you found from the Ghia & Ghia benchmark?

February 3, 2016, 12:28
#3
New Member

Join Date: Feb 2016
Posts: 6
Rep Power: 9
I carried out this simulation at reynolds number 100. I tried for various reynolds number but still got similar plots for all the cases.

I have attached the correct solution of the problem which is also the same as that given by Ghia.

I am unable to figure out the problem causing the u velocity contours to be symmetric about the central vertical line. As compared to the existing result, the nature is same only that the contours are not inclined towards the corner in my solution.
Attached Images
 images.jpg (11.8 KB, 7 views) V&tc_liddriven_fluentu.png (10.4 KB, 10 views)

 February 3, 2016, 12:31 #4 Senior Member   Filippo Maria Denaro Join Date: Jul 2010 Posts: 6,390 Rep Power: 67 if your solution is at Re=100, I see that the contour pattern is too specular as respect with the centerline x=L/2. That seems the convective does not enter into the solution. Are you sure you are gettin a steady state solution?

 February 3, 2016, 12:59 #5 New Member   Join Date: Feb 2016 Posts: 6 Rep Power: 9 Yes, this is the steady state solution that I have obtained. I checked my convection flux equation a lot of times and couldn't find any error in that. I also compared the magnitudes of the convection and diffusion fluxes and found that the diffusion flux was comparatively larger than the convection flux. But I can't figure out any solution to this since the error could be anywhere (even in my pressure correction equations or somewhere else) I cross-checked all my equations and formulas and couldn't get the correct results yet. I think I am missing out on something but I can't figure out what

 February 3, 2016, 13:06 #6 New Member   Join Date: Feb 2016 Posts: 6 Rep Power: 9 Yes, this is the steady state solution that I have obtained. I checked my convection flux equation a lot of times and couldn't find any error in that. I also compared the magnitudes of the convection and diffusion fluxes and found that the diffusion flux was comparatively larger than the convection flux. But I can't figure out any solution to this since the error could be anywhere (even in my pressure correction equations or somewhere else) I cross-checked all my equations and formulas and couldn't get the correct results yet. I think I am missing out on something but I can't figure out what

February 3, 2016, 13:10
#7
Senior Member

Filippo Maria Denaro
Join Date: Jul 2010
Posts: 6,390
Rep Power: 67
Quote:
 Originally Posted by DSS Yes, this is the steady state solution that I have obtained. I checked my convection flux equation a lot of times and couldn't find any error in that. I also compared the magnitudes of the convection and diffusion fluxes and found that the diffusion flux was comparatively larger than the convection flux. But I can't figure out any solution to this since the error could be anywhere (even in my pressure correction equations or somewhere else) I cross-checked all my equations and formulas and couldn't get the correct results yet. I think I am missing out on something but I can't figure out what
Using non-dimensional equations, the convective flux should be O(1), compared to the O(10^-2) of the diffusive one, irrespective of the pressare term.

February 3, 2016, 13:24
#8
New Member

Join Date: Feb 2016
Posts: 6
Rep Power: 9
Quote:
 Originally Posted by FMDenaro Using non-dimensional equations, the convective flux should be O(1), compared to the O(10^-2) of the diffusive one, irrespective of the pressare term.
That is exactly what I have not been able to solve. My convective flux is O(10^-3) of the diffusive one. Hence, the convective flux doesn't contribute much to the equations.

 February 10, 2016, 13:55 #9 New Member   Join Date: Feb 2016 Posts: 6 Rep Power: 9 I'd really appreciate if anyone could help me with this problem. I want to know what could be the possible errors so that I can work on them.

 February 10, 2016, 17:06 #10 Senior Member   Filippo Maria Denaro Join Date: Jul 2010 Posts: 6,390 Rep Power: 67 As I previously wrote, the symmetry could be due to an error in the discretization of the convective terms