CFD Online Discussion Forums

CFD Online Discussion Forums (https://www.cfd-online.com/Forums/)
-   OpenFOAM Pre-Processing (https://www.cfd-online.com/Forums/openfoam-pre-processing/)
-   -   DieselEngineFoam (https://www.cfd-online.com/Forums/openfoam-pre-processing/62017-dieselenginefoam.html)

stefanke December 30, 2005 08:02

As far as I known the dieselEn
 
As far as I known the dieselEngineFoam is the only solver that does not currently have a tutorial case or FoamX configuration!

Has anybody out there an existing FoamX config or at least a complete set of dictionaries for the dieselEngine solver?

Any help is very appreciated.

cheers
Stephan

lucchini December 30, 2005 08:54

Hi Stefan, dieselEngineFoam
 
Hi Stefan,
dieselEngineFoam has not a tutorial, but, as you can see in the $FOAM_TUTORIALS/ directory there are the tutorials of dieselFoam and engineFoam.
dieselEngineFoam is dieselFoam with the moving mesh (have a look at the code of both solvers).
So the only difference in a case setup between dieselFoam and dieselEngineFoam is that you need to put in the /constant/ directory the file engineGeometry. To better understand how dieselFoam and engineFoam works you can run both the tutorials (aachenBomb and kivaTest).

In the aachenBomb case focus on the following files:
constant/sprayProperties
constant/injectorProperties
constant/thermophysicalProperties
constant/chemistryProperties

In the kivaTest file have a look to
constant/engineGeometry
system/controlDict
and look at the mesh, and see that the patches requires particular names (piston, liner, cylinderHead).
Then try to merge the two tutorial and have fun!
bye
Tommaso

stefanke December 30, 2005 10:34

I try to "merge" the two cases
 
I try to "merge" the two cases to get a setup for the dieselEngine solver. I use the same mesh as the engineFoam case to get rid of mesh problems.

Nevertheless the calculation fails:

Number of parcels in system | 0
Injected liquid mass....... | 0 mg
Liquid Mass in system...... | 0 mg
SMD, Dmax.................. | 0 mu, 0 mu
Added gas mass = -1.71521e-10 mg
Evaporation Continuity Error| -1.71521e-10 mg
ExecutionTime = 240.61 s


Mean and max Courant Numbers = 0.000929914 31.5694
deltaT = 8.74482e-10
Crank angle = -175.033 CA-deg
deltaZ = 3.57846e-10
clearance: 0.0855261
Piston speed = 0.409209 m/s
volume continuity errors : sum local = 1.10786e-15, global = -5.78031e-18
Solving chemistry
BICCG: Solving for Ux, Initial residual = 0.929704, Final residual = 4.04514e-07, No Iterations 12
BICCG: Solving for Uy, Initial residual = 0.902951, Final residual = 6.88233e-07, No Iterations 12
BICCG: Solving for Uz, Initial residual = 0.909879, Final residual = 4.33674e-07, No Iterations 12
BICCG: Solving for O2, Initial residual = 0.884681, Final residual = 7.73892e-07, No Iterations 12
BICCG: Solving for h, Initial residual = 0.852227, Final residual = 5.31857e-07, No Iterations 12


--> FOAM FATAL ERROR : attempt to use janafThermo<equationofstate> out of temperature range 200 ->
5000; T = 113.962

From function janafThermo<equationofstate>::checkT(const scalar T) const
in file /home/dm2/henry/OpenFOAM/OpenFOAM-1.2/src/thermophysicalModels/specie/lnInclude/ janafTh
ermoI.H at line 73.

FOAM aborting


- Why there is no injected fuel
- How to handle the begin (start) of injection

injector setup:
---------------
injectorType unitInjector;

unitInjectorProps
{
position (0.03 0 0.0995);
direction (0 0 1);
diameter 0.00019;
Cd 0.9;
mass 6e-06;
temperature 320;
nParcels 5000;


Any ideas ?

lucchini December 30, 2005 15:00

Hi, It seems to me there is a
 
Hi,
It seems to me there is a huge Courant Number.

What did you set in the controlDict file for deltaT?

What boundary conditions are you using for T, p, O2, N2, k and epsilon? Are they correct?

Also: the injectorProperties file for the dieselFoam has the injection law depending on time and not by the crank angle, as you have to set in the dieselEngineFoam case. So if you start a case from -180 and the injection law start from 0 you don't have injected particles till you reach 0 Crank Angles.

I hope this could help you a little bit. Keep on.
Bye
Tommaso

stefanke December 30, 2005 16:02

control dict: -------------
 
control dict:
-------------
adjustTimeStep yes;
maxCo 0.1;
maxDeltaT 1.0;

boundary conditions:
--------------------
T,p,k,epsilon - engineFoam default
O2,N2,Ydefault- zero gradient

injection start:
----------------
You are right, the injection law is depending on time and not by CA but how can I define the injection start time or CA.

Has anyone out there a setup for the dieselEngine solver which works? If so feel free to post! I think a lot of people are insterested in.


Stefan

stefanke January 6, 2006 08:33

is there anybody who can help
 
is there anybody who can help ?

m_ende January 12, 2006 05:40

Hello, what averaged Veloci
 
Hello,

what averaged Velocity and pressure write dieselEngineFoam out ?

stefanke January 12, 2006 06:39

Hi, as you can see from the
 
Hi,

as you can see from the dieselEngine (logSummary.H) code

mean p: spacial mean value of the pressure of the computational domain.

mean u': mean (time+space) turbulent velocity fluctuation

hth

p.s are you able to compute the case (dieselEngine) successfuly? My computation cancels after -66 CA with an error!

stefanke January 12, 2006 06:43

Can anyone help me with this e
 
Can anyone help me with this error.

from CA -180 to -66 all is ok but then the following error occurs:

--> FOAM FATAL ERROR : attempt to use janafThermo<equationofstate> out of temperature range 200 -> 5000; T = 5101.98

From function janafThermo<equationofstate>::checkT(const scalar T) const
in file /home/dm2/henry/OpenFOAM/OpenFOAM-1.2/src/thermophysicalModels/specie/lnInclude/ janafThermoI.H at line 73.

FOAM aborting

time step properties:

adjustTimeStep yes;
maxCo 0.2;
maxDeltaT 0.01;


I use the kivaTest Mesh from the engineFoam tutorial!

m_ende January 12, 2006 08:52

I mean the averaged velocity a
 
I mean the averaged velocity and injector pressure at the begin of calculation ?

The error I think comes from a unstable solution of the Enthalpy equation. Have you chemistry solve on ???

niklas January 12, 2006 09:02

Those values are incorrect for
 
Those values are incorrect for engines, since
the values are based on the average pressure
at the start of the simulation.

Maybe they should be removed for moving meshes,
but I've kept them since I like to have that info
to check that the input is descent.
if you want to check that these values are
physically sound,
estimate the pressure at SOI and set that value as reference in -180/p, run the code and see what you get.

...and for the non-working dieselEngineFoam case,
try setting
momentumPredictor off in the PISO-dictionary
in fvSolution.

N

stefanke January 12, 2006 09:31

Hi Niklas, I have already t
 
Hi Niklas,

I have already tried to set momentumPredicator off but the problem still exists. As I already mentioned above the calculation aborts after -66,5 CA reproducible.

Here are the setup (fvSolution/fvSchemes) I use:


fvSolution:
-----------

solvers
{
rho ICCG 1e-06 0;
U BICCG 1e-06 0;
p ICCG 1e-09 0;
Yi BICCG 1e-06 0;
h BICCG 1e-06 0;
k BICCG 1e-06 0;
epsilon BICCG 1e-06 0;
}

PISO
{
nCorrectors 2;
nNonOrthogonalCorrectors 0;
momentumPredictor off;
}


fvSchemes:
----------

ddtSchemes
{
default Euler;
}

gradSchemes
{
default Gauss linear;
grad(p) Gauss linear;
}

divSchemes
{
default none;
div(phi,rho) Gauss limitedLinear 1;
div(phi,U) Gauss limitedLinearV 1;
div(phiU,p) Gauss linear;
div(phi,k) Gauss limitedLinear 1;
div(phi,epsilon) Gauss limitedLinear 1;
div(phi,Yi_h) Gauss upwind;
div(phi,fu_ft_h) Gauss multivariateSelection
{
fu limitedLinear 1;
ft limitedLinear 1;
h limitedLinear 1;
};
div((muEff*dev2(grad(U).T()))) Gauss linear;
}

laplacianSchemes
{
default Gauss linear corrected;
laplacian(muEff,U) Gauss linear corrected;
laplacian(muEff,ft) Gauss linear corrected;
laplacian(muEff,fu) Gauss linear corrected;
laplacian(((alphah*mut)+alpha),h) Gauss linear corrected;
laplacian((rho|A(U)),p) Gauss linear corrected;
laplacian(rhoD,k) Gauss linear corrected;
laplacian(rhoD,epsilon) Gauss linear corrected;
}

interpolationSchemes
{
default linear;
interpolate(HbyA) linear;
}

snGradSchemes
{
default corrected;
}

fluxRequired
{
p;
}

other:
------

chemistry off
no fuel injection


Do you have a clue what is going wrong here?

niklas January 12, 2006 10:26

Yes I think I have a clue what
 
Yes I think I have a clue whats going on.
you're doing non-linear CFD on a complex mesh,
it crashes easily then.

try with nNonOrthogonalCorrectors 1
and upwind for all variables.

I would reduce the convergence criteria for Yi to 10e-7 also.

N

stefanke January 12, 2006 11:12

Niklas, as recommended I ch
 
Niklas,

as recommended I changed the follwing parameters:

Yi BICCG 1e-07 0;
nNonOrthogonalCorrectors 1;
momentumPredictor off;

div(phi,rho) Gauss upwind;
div(phi,U) Gauss upwind;
div(phiU,p) Gauss upwind;
div(phi,k) Gauss upwind;
div(phi,epsilon) Gauss upwind;
div(phi,Yi_h) Gauss upwind;

all other parameters are unchanged.

but the problems still exists http://www.cfd-online.com/OpenFOAM_D...lipart/sad.gif

niklas January 12, 2006 11:17

Wow, that was fast... I sur
 
Wow, that was fast...

I sure hope you didnt restart from the latest
result, it sounds like it already has gone wrong there.

Remove the last result directory and restart from an earlier solution.
Also start from -180 with these settings and see if the problem persists.

N

stefanke January 12, 2006 11:37

Niklas, thanks for you supp
 
Niklas,

thanks for you support so far!

Sorry for my ignorance but what do you mean with "... and upwind for all variables"

set upwind for all convection terms set all interpolation schemes to upwind?

what`s about: div(phi,fu_ft_h) Gauss multivariateSelection, set this to upwind too?

btw: have you got a setup which works and you are able to email me?

>Remove the last result directory and restart from an earlier solution.

Ok I started with -80 CA but it crashes as before at -66.5 CA! I am gonna restart the calculation from -180CA but I think this doesn`t help ...

stefanke January 12, 2006 11:37

Niklas, thanks for you supp
 
Niklas,

thanks for you support so far!

Sorry for my ignorance but what do you mean with "... and upwind for all variables"

set upwind for all convection terms set all interpolation schemes to upwind?

what`s about: div(phi,fu_ft_h) Gauss multivariateSelection, set this to upwind too?

btw: have you got a setup which works and you are able to email me?

>Remove the last result directory and restart from an earlier solution.

Ok I restarted from -80 CA but it crashes as before at -66.5 CA! I am gonna restart the calculation from -180CA but I think this doesn`t help ...

stefanke January 12, 2006 15:52

the same problem occurs when I
 
the same problem occurs when I restart my calc from -180 CA http://www.cfd-online.com/OpenFOAM_D...lipart/sad.gif

I have changed my mesh to a simple wedge geomerty and now the calc proceed successfully. It seems to be a mesh problem.

I use the provided KivaTest mesh from the egineFoam tutorial. I think this mesh should work!

Niklas (other peoples are welcome too) what do you think about this?

niklas January 13, 2006 09:32

That mesh is pretty bad and i
 
That mesh is pretty bad and
it looks like it needs some outercorrectors...

The problem is in the highly distorted cells,
which requires some extra corrections for the
enthalpy.

use this
PISO
{
nOuterCorrectors 3;
nCorrectors 1;
nNonOrthogonalCorrectors 1;
momentumPredictor off;
}

Should work.

N

stefanke January 13, 2006 09:40

thanks I will try your PISO se
 
thanks I will try your PISO setup.

another problem is that there is no combustion (signifiant pressure or temperature gain) altough I enabled the chemistry (15species, 39reations).


All times are GMT -4. The time now is 10:28.