Implicit steger warming fvs

pooja · May 18, 2011, 17:05

Hiiii, i am writing code for the implicit steger warming flux vector splitting for 1d cd nozzle case. i am using gmres method to solve Ax = b.
One strange thing is happening with in the code, as all of us know that for implicit there is no restriction on CFL number, First point is that if i start with CFL value 0.1 and max CFL value 2, with incrementation in CFL given by CFL*1.1 my code is converging nicely upto log rss value -7.034528 and results are matching.
But if i am taking max CFL number more than 2 say 4 the its blow up and go for log rss value -1.234 so on or if further increase in max CFL say upto 10 i will get nan after some iteration. I am really getting fed up with it from last one month then one of my friend told that u can put your problem in cfd forum surely you will get nice ideas. All ideas are welcome and thanks in advance plz help me

praveen · May 19, 2011, 00:10

gmres is not easy to get working for flows with shocks.

From your description I assume you are solving a steady state problem.

You do need to use a smaller cfl to begin with. You can slowly increase this as your residual decreases. So you can try something like

$cfl^n = cfl^{n-1} \frac{Res^{n-1}}{Res^n}$

You can also put an upper bound on the cfl.

Note that without good preconditioner, gmres will not work.

pooja · May 19, 2011, 05:34

Thanx parveen for reply, but for 1d cd nozzle case as i said earlier i am starting with a low value of CFl 0.1 and incrementing with CFL*1.01 .
Shall i have to change the iterative method ?

pooja · May 19, 2011, 06:48

Hey only, one reply from this forum not expected

I am really in very much trouble help me plzzzzzzzzzzz

May 18, 2011, 17:05	Implicit steger warming fvs	#1
pooja New Member dushyant Join Date: May 2011 Posts: 6 Rep Power: 4	Hiiii, i am writing code for the implicit steger warming flux vector splitting for 1d cd nozzle case. i am using gmres method to solve Ax = b. One strange thing is happening with in the code, as all of us know that for implicit there is no restriction on CFL number, First point is that if i start with CFL value 0.1 and max CFL value 2, with incrementation in CFL given by CFL*1.1 my code is converging nicely upto log rss value -7.034528 and results are matching. But if i am taking max CFL number more than 2 say 4 the its blow up and go for log rss value -1.234 so on or if further increase in max CFL say upto 10 i will get nan after some iteration. I am really getting fed up with it from last one month then one of my friend told that u can put your problem in cfd forum surely you will get nice ideas. All ideas are welcome and thanks in advance plz help me

Thread Tools
Show Printable Version Email this Page
Display Modes
Linear Mode Switch to Hybrid Mode Switch to Threaded Mode

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Help. implicit discritization	dearboy	Main CFD Forum	0	November 25, 2010 21:46
Can anyone explain the difference between these implicit concepts?	bearcat	Main CFD Forum	0	February 20, 2010 18:39
Help regarding Steger Warming Scheme	Sunny	Main CFD Forum	2	July 27, 2006 15:53
how implicit is implicit?	Joe cool	Main CFD Forum	7	July 16, 2004 11:10
UDF : Chemical Reaction in Catalytic Converter	adhimac	FLUENT	1	March 27, 2001 13:04

May 19, 2011, 00:10		#2
praveen Super Moderator Praveen. C Join Date: Mar 2009 Location: Bangalore Posts: 222 Blog Entries: 6 Rep Power: 7	gmres is not easy to get working for flows with shocks. From your description I assume you are solving a steady state problem. You do need to use a smaller cfl to begin with. You can slowly increase this as your residual decreases. So you can try something like $cfl^n = cfl^{n-1} \frac{Res^{n-1}}{Res^n}$ You can also put an upper bound on the cfl. Note that without good preconditioner, gmres will not work.

May 19, 2011, 05:34		#3
pooja New Member dushyant Join Date: May 2011 Posts: 6 Rep Power: 4	Thanx parveen for reply, but for 1d cd nozzle case as i said earlier i am starting with a low value of CFl 0.1 and incrementing with CFL*1.01 . Shall i have to change the iterative method ?

May 19, 2011, 06:48		#4
pooja New Member dushyant Join Date: May 2011 Posts: 6 Rep Power: 4	Hey only, one reply from this forum not expected I am really in very much trouble help me plzzzzzzzzzzz