CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   OpenFOAM Running, Solving & CFD (http://www.cfd-online.com/Forums/openfoam-solving/)
-   -   How to improve solution time for the SonicTurbFoam app (http://www.cfd-online.com/Forums/openfoam-solving/59697-how-improve-solution-time-sonicturbfoam-app.html)

shaun May 1, 2007 03:16

Ladies and Gents, I am solving
 
Ladies and Gents, I am solving a case which involves the RAE2822 airfoil. I have used the sonicTurbFoam solver because my flowfield has a freestream value of Mach 0.73, Re no of 6.5 * 10^6, 2.79deg AoA, Tinf = 300K. On the top surface of the airfoil I know speeds will go supersonic (approx 1.2 Mach). I have used a pressure-inlet-outlet for the farfield-in boundary and pressureTransmissiveOutlet for the outlet. Current timestep is 1e-06 and typical max Courant number is 0.42. I estimate I will need approx 2.85 seconds of total flow time to allow the flow from the inlet to pass to the outlet 5 times (147(m)/253.45(m/s/)). I am running parallel on 4 cpu's (two dual core Intel Xeon's, 4Gb RAM).

Given below is an output of a typical timestep report.

Time = 0.098347

Courant Number mean: 0.00275724 max: 0.426394
diagonal: Solving for rho, Initial residual = 0, Final residual = 0, No Iterations 0
DILUPBiCG: Solving for Ux, Initial residual = 4.10529e-05, Final residual = 2.62615e-09, No Iterations 1
DILUPBiCG: Solving for Uy, Initial residual = 3.03639e-05, Final residual = 3.48547e-08, No Iterations 1
DILUPBiCG: Solving for h, Initial residual = 0.000784216, Final residual = 2.38317e-08, No Iterations 1
DILUPBiCG: Solving for p, Initial residual = 0.00150568, Final residual = 9.3766e-09, No Iterations 2
DILUPBiCG: Solving for p, Initial residual = 6.48232e-06, Final residual = 2.45758e-09, No Iterations 1
DILUPBiCG: Solving for p, Initial residual = 7.69953e-08, Final residual = 3.94493e-11, No Iterations 1
diagonal: Solving for rho, Initial residual = 0, Final residual = 0, No Iterations 0
time step continuity errors : sum local = 2.35889e-09, global = 8.84185e-10, cumulative = -6.03889e-07
DILUPBiCG: Solving for p, Initial residual = 8.8288e-06, Final residual = 7.66856e-10, No Iterations 2
DILUPBiCG: Solving for p, Initial residual = 1.837e-08, Final residual = 1.05416e-11, No Iterations 1
DILUPBiCG: Solving for p, Initial residual = 5.82397e-10, Final residual = 5.82397e-10, No Iterations 0
diagonal: Solving for rho, Initial residual = 0, Final residual = 0, No Iterations 0
time step continuity errors : sum local = 1.42302e-08, global = 1.77953e-10, cumulative = -6.03711e-07
DILUPBiCG: Solving for epsilon, Initial residual = 9.97207e-09, Final residual = 9.97207e-09, No Iterations 0
DILUPBiCG: Solving for k, Initial residual = 5.876e-10, Final residual = 5.876e-10, No Iterations 0
ExecutionTime = 5465.25 s ClockTime = 5536 s


As you can see, I don't think there is much more I can do to get this going any quicker. At this rate it will take some time to approach 2.85secs of total flow time with a timestep of 1e-06 secs. If I double this value I obviously get a doubling of courant number to approx 0.8. When I did this I would get spikes in courant number where it would go above 1. Is it best to keep courant number around 0.5 for this type of problem? Should I try the rhoSimpleFoam solver even though it supposed to be setup for low-mach flows (i.e. Should I try my luck with this as it is a steady-state app?)

Thanks for reading all this. If there are any further questions, please don't hesitate to ask them.

Cheers

Shaun.D

olesen May 2, 2007 06:51

Hi Shaun, If you are seekin
 
Hi Shaun,

If you are seeking a steady-state solution, a SIMPLE solver will definitely be the more economical solution. The rhoSimpleFoam solver can be coaxed a bit to improve the Mach range (up to approx 1.3), but for your case more modifications would be helpful.
If you cannot afford the normal support contract, perhaps OpenCFD will quote you for a smaller job. Or try to find other forum users to help fund this development.
/mark

shaun May 11, 2007 03:34

Thanks Mark for your help. I h
 
Thanks Mark for your help. I have some more time to play with OF now, so I will run the problem in rhoSimpleFoam and see what results I get.

Cheers

Shaun.D


All times are GMT -4. The time now is 23:11.