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/)
-   -   Intake boundary conditions error (http://www.cfd-online.com/Forums/openfoam-solving/57790-intake-boundary-conditions-error.html)

kieranwood85 March 3, 2009 06:34

Hi, I am trying to solve a
 
Hi,

I am trying to solve a flow for a supersonic engine intake (much like the one on the Blackbird aircraft) and am having problems specifying the BCs. All the ones I have tried bring up this error after a few iterations.

[0]
[0]
[0] Maximum number of iterations exceeded#0 Foam::error::printStack(Foam:http://www.cfd-online.com/OpenFOAM_D...part/proud.gifstream&) in "/home/kieran/OpenFOAM/OpenFOAM-1.5/lib/linuxGccDPOpt/libOpenFOAM.so"
#1 Foam::error::abort() in "/home/kieran/OpenFOAM/OpenFOAM-1.5/lib/linuxGccDPOpt/libOpenFOAM.so"
#2 Foam::hThermo<foam::puremixture<foam::consttranspo rt<foam::speciethermo<foam::hc onstthermo<foam::perfectgas> > > > >::calculate() in "/home/kieran/OpenFOAM/OpenFOAM-1.5/lib/linuxGccDPOpt/libbasicThermophysicalMode ls.so"
#3 Foam::hThermo<foam::puremixture<foam::consttranspo rt<foam::speciethermo<foam::hc onstthermo<foam::perfectgas> > > > >::correct() in "/home/kieran/OpenFOAM/OpenFOAM-1.5/lib/linuxGccDPOpt/libbasicThermophysicalMode ls.so"
#4 main in "/home/kieran/OpenFOAM/OpenFOAM-1.5/applications/bin/linuxGccDPOpt/rhoCentralFoa m"
#5 __libc_start_main in "/lib/tls/i686/cmov/libc.so.6"
#6 Foam::regIOobject::readIfModified() in "/home/kieran/OpenFOAM/OpenFOAM-1.5/applications/bin/linuxGccDPOpt/rhoCentralFoa m"
[0]
[0]
[0] From function specieThermo<thermo>::T(scalar f, scalar T0, scalar (specieThermo<thermo>::*F)(const scalar) const, scalar (specieThermo<thermo>::*dFdT)(const scalar) const) const
[0] in file /home/dm2/henry/OpenFOAM/OpenFOAM-1.5/src/thermophysicalModels/specie/lnInclude/ specieThermoI.H at line 78.


My case needs a supersonic far-field inlet condition (which I am just using fixedValue for p, T, and U), and then two outlets- one for the flow around the outside of the intake (using zeroGradient for p,T,U at the moment) - and one for the flow captured by the intake which needs to reach a lower total pressure in order to induce a shockwave (currently using inletOutlet for p and zeroGradient for T and U). For the second outlet I have also tried ouletInlet and fixedValue but get similar errors. Also am using the rhoCentralFoam solver.

Does anyone know of a more appropriate BC for this case?

Thanks in advance

Kieran

kieranwood85 March 4, 2009 21:58

After looking at the flow befo
 
After looking at the flow before the error occurs it appears that two flows are being created giving a non physical solution. 1) the farfield inlet conditions set by fixed value BCs and 2) a pressure driven flow backwards into the domain from the inletOutlet BC.

I guess when these two flows meet the solver cannot handle the difference in total pressure and temperature well and so crashes with a thermoPhysicalModel error above.

Is there a way to specify an outlet BC that maintains flow out of the domain and never acts as an inlet? A an average mass flux would be useful but how to specify it on the patch in the p, T, and U files?

kieranwood85 March 13, 2009 07:37

Hi, I am still having probl
 
Hi,

I am still having problems solving my case and have tried all of the boundary condition combinations I could think of. Perhaps if I upload a picture of my case it will describe my problem more. Here is a picture of the mesh I am using.

http://www.cfd-online.com/OpenFOAM_D...es/1/11647.jpg

It is an axi-symmetric case where there is a conical section inside a cowl. There is only a 2D slice shown but the case is 3D (90 degree segment with symmetry planes applied).

The problem is that I cannot force any fixed values on the outlet2 since they are unknown (well I have an approximate expected value) and will be determined by the shock wave that forms between the cone and cowl lip.

Instead what I need is to force the flow to always be out of the domain parallel to the X axis and never an inflow. Hopefully the flow will then choke itself and converge to a solution.

My current boundaries (although the solution still crashes with an error similar to above) are,

for p
inlet
{
type fixedValue
value uniform 101370
}
outlet1
{
type zeroGradient
}
outlet2
{
type zeroGradient
}
+ some symmetry planes

for T
inlet
{
type fixedValue
value uniform 166.6
}
outlet1
{
type zeroGradient
}
outlet2
{
type zeroGradient
}
+ some symmetry planes

for U
inlet
{
type fixedValue
value uniform (526.26 0 0)
}
outlet1
{
type zeroGradient
}
outlet2
{
type pressureDirectedInletVelocity;
inletDirection uniform (1 0 0);
value uniform (223 0 0);
}
+ some symmetry planes

The initial internal fields are set the same as the inlet values. The 223 is an approximate value of the expected velocity but as far as I know this is only an initial boundary value and will not affect the solution.

Also i'm using rhoCentralFoam.

If anyone knows a better way to set the boundaries to make the solution reach a steady state I would be grateful.

Thanks

Kieran

mahmood.asgari September 10, 2010 06:59

My idea....
 
You said : I know this is only an initial boundary value and will not affect the solution.BUT NO!!!
It is very important . you can change initial internal fields by subsonic and change back pressure for solving! because i had this problem later.


All times are GMT -4. The time now is 22:26.