Incompressible Navier-Stokes equations for a backward facing step

werder85 · January 3, 2011, 12:57

Hi!
I'm trying to write a program which calculates the velocities and the pressure for a backward facing step using a MAC grid and artificial compressibility for the continuity equation.

I'm using the non-dimensionalized equations, as a result the speed at the inflow boundary is 1.0. The problem is: which values should the velocities inside the grid have (and the pressures)? In a first, naive approach I have initialized them with zero but after just two or three time steps the values get out of control because of the big difference between the values at the inflow boundary and the values situated just to the right of the inflow boundary.

otd · January 3, 2011, 17:19

You might try setting up the narrow part as a channel flow (analytic solution is in Schlichting, Boundary Layer Theory. After the step (maybe one or two step heights downstream), initialize the velocities as a new channel flow starting and flowing to the outlet. Of course the recirculation region just downstream from the step is the toughest part. Don't have a good suggestion to initialize that.

I don't have experience using artificial compressibility. But a rule of thumb for starting calculations is that "if the IC values are smooth and grossly similar to the expected solution, that's good enough." Just spend a lot of time with small relaxation factors and small time steps to reduce the residuals, then gradually increase them as the flow settles out. NOTE that this is not the only "rule of thumb for starting calculations".

Good luck.

OTD

January 3, 2011, 12:57	Incompressible Navier-Stokes equations for a backward facing step	#1
werder85 New Member Join Date: Dec 2010 Posts: 14 Rep Power: 15	Hi! I'm trying to write a program which calculates the velocities and the pressure for a backward facing step using a MAC grid and artificial compressibility for the continuity equation. I'm using the non-dimensionalized equations, as a result the speed at the inflow boundary is 1.0. The problem is: which values should the velocities inside the grid have (and the pressures)? In a first, naive approach I have initialized them with zero but after just two or three time steps the values get out of control because of the big difference between the values at the inflow boundary and the values situated just to the right of the inflow boundary.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
How to write k and epsilon before the abnormal end	xiuying	OpenFOAM Running, Solving & CFD	8	August 27, 2013 16:33
Backward facing step	sega	OpenFOAM Running, Solving & CFD	8	June 2, 2011 01:07
Backward facing step code	Josyula	Main CFD Forum	0	March 31, 2009 07:33
Backward facing step	ganesh	Main CFD Forum	0	February 8, 2007 03:40
Navier Stokes Equations	J.J.	Main CFD Forum	2	June 29, 2000 10:31

January 3, 2011, 17:19		#2
otd Member private Join Date: Mar 2009 Posts: 74 Rep Power: 17	You might try setting up the narrow part as a channel flow (analytic solution is in Schlichting, Boundary Layer Theory. After the step (maybe one or two step heights downstream), initialize the velocities as a new channel flow starting and flowing to the outlet. Of course the recirculation region just downstream from the step is the toughest part. Don't have a good suggestion to initialize that. I don't have experience using artificial compressibility. But a rule of thumb for starting calculations is that "if the IC values are smooth and grossly similar to the expected solution, that's good enough." Just spend a lot of time with small relaxation factors and small time steps to reduce the residuals, then gradually increase them as the flow settles out. NOTE that this is not the only "rule of thumb for starting calculations". Good luck. OTD