[cedrik]

Hi everybody,
I 'm trying to perform a simulation of wind on a building in order to know the pressure on a particular part of a building roof.
I follow the simplefoam/motorbike tutorial and I adjust it to my case. But when I try to see the result in paraview, it's wrong!(pictures).
I think it can be a problem of dimensions in my blockMesh but I tried different solutions and I failed.
Could somebody explain me why it doesn't work please ? (I'm a beginner in OF)

[gregjunqua]

what is your inlet condition? (by the way the others)
Don't forget that you are in ABL condition many things need to be change. And then motorbike simulation don't agree with your simulation

what is your domain size compare to the building size?

[cedrik]

Hi,
Thank for your reply.
I'm a very beginner and I guessed I just had to "replace" the motorbike by my own geometry. I'm not sure to understand what you mean by Inlet Condition but I attached some files I hope it's what you mean.

I'am doing research on ABL as you advised to me but, for the moment, I have not the solution.

You said the motorbike tutorial wasn't the good one, could you advise me an other one?

About my domain size : If I call L the maximum lenght of my building, il have 3L between my building and the inlet, 3L on each side and 8L behind.

Thank for your help.

[gregjunqua]

My meaning is that you are in atmospheric boundary layer condition. So you need to apply a log-law (or the power law) on the inlet and change the inlet accordingly to the epsilon (or omega) and to the Turbulent Kinetic Energy (TKE can stay in fixed value).

You can have a try with the boundary condition in the turbineSitting tutorial located in the simpleFoam's tutorial part. There you will have the different inlet (k-epsilon) and also you can use the nut bottom boundary to model grass or road around your buildings.

regards
GJ

[cedrik]

Hello GJ,
Unfortunatly, I have no time to work on it until today. I look at the tutorial you advice me and I think I understand what mean ABLC now but I steel don't see how to change my inlet to have a log-law.
Indeed, I undersatnd why I must do it but as I am a very beginner in C++, I can't see how to set up it in an OpenFoam file...
Thank you very much for your help,
Regards
Cedric

[gregjunqua]

Hi there you needn't to program
just look at tutorials/incompressible/simpleFoam/turbineSiting

it show the different inlet for ABL layer
epsilon : atmBoundaryLayerInletEpsilon
k : uniformFixedValue
nut : nutkAtmRoughWallFunction
U : atmBoundaryLayerInletVelocity

these are a little outdated inlet I suggest you to have a look of this article for parametrize your inlet
Yang, Y., M. Gu, S. Chen, and X. Jin, 2009: New
inflow boundary conditions for modelling the neutral equilibrium atmospheric boundary layer in
computational wind engineering. Journal of wind engineering and industrial aerodynamic, 95,
88–95.
f
And you should lower the constant C_\mu of the bradshaw relationship from 0.09 to 0.03
And change the turbulent kinetic energy inlet accordingly

[cedrik]

Hi,
I'm traying to adjust the turbinesiting tutorial to my case but there still is a file I don't understand everythings.
When I look at the Allrun, I undersand until SnappyHexMesh, after I cut off the topoSet "step" as I think it is useless (I also cut off fvOption). But I can't manage to run my case.
Here is the error message :

Create time

Create mesh for time = 0

Reading field p

Reading field U

Reading/calculating face flux field phi

Selecting incompressible transport model Newtonian
Selecting RAS turbulence model kEpsilon
bounding k, min: 0 max: 1.3 average: 1.30000000001
bounding epsilon, min: 0 max: 0.577954048984 average: 0.0100000000001
kEpsilonCoeffs
{
Cmu 0.09;
C1 1.44;
C2 1.92;
sigmaEps 1.11;
}

No finite volume options present


SIMPLE: convergence criteria
field p tolerance 0.001
field U tolerance 0.0001
field "(k|epsilon)" tolerance 0.0001


Starting time loop

Time = 1

smoothSolver: Solving for Ux, Initial residual = 1, Final residual = 0.0850202022064, No Iterations 2
smoothSolver: Solving for Uy, Initial residual = 1, Final residual = 0.0421649962824, No Iterations 2
smoothSolver: Solving for Uz, Initial residual = 1, Final residual = 0.0820188872721, No Iterations 2
--------------------------------------------------------------------------
MPI_ABORT was invoked on rank 31 in communicator MPI_COMM_WORLD
with errorcode 1.

NOTE: invoking MPI_ABORT causes Open MPI to kill all MPI processes.
You may or may not see output from other processes, depending on
exactly when Open MPI kills them.
--------------------------------------------------------------------------
[0]
[0]
[0] --> FOAM FATAL ERROR:
[0] No coarse levels created, either matrix too small for GAMG or nCellsInCoarsestLevel too large.
Either choose another solver of reduce nCellsInCoarsestLevel.
[0]
[0] From function GAMGSolver::GAMGSolver(const word& fieldName,const lduMatrix& matrix,const FieldField<Field, scalar>& interfaceBouCoeffs,const FieldField<Field, scalar>& interfaceIntCoeffs,const lduInterfaceFieldPtrsList& interfaces,const dictionary& solverControls)
[0] in file matrices/lduMatrix/solvers/GAMG/GAMGSolver.C at line 121.
[0]
FOAM parallel run exiting

Do you understand where the error could be?

[cedrik]

Hi, I worked on it the all day and I finally managed to have results for my own stl.
After that, I searched for renseignement on "U : atmBoundaryLayerInletVelocity" and there are somethings I can't anderstand and I hope you could help me.

As you can see in the pictures I attached, the results are diferents when I use Excel with the U atmBoundaryLayerInletVelocity equation and when I watch the same thing with Paraview.

Moreover I can't understand why there is no correspondance between the Href/Uref theoric (I put in the ABLConditions file in OpenFoam) and the Href/Uref I visualise paraview.

I hope I clear enought.
Regards,
Cedric

[elvis]

Hi,

http://openfoamwiki.net/index.php/Ma...banAreaNiigata
http://openfoamwiki.net/index.php/Ma...ngInCityBlocks

[cedrik]

Hi Elvis,
Thanks for your reply.
I look at the 2 examples you adviced me. I guess it can be usefull to compare with my model when it will work but I still can't understand why I get this diffences between the theory (with excel) and the simulation (in OpenFoam) as I assume it the same equation at the base.

[cedrik]

Hi everybody,
Now, I manage to perform a case with my building thank to your help.
I have just an other question concerning the boundaries : why the tutorial "turbinesitting" doesn't use the condition "symmetrtPlane" on the sides and the top boundaries?
Thanks for your replies,
Regards,
Cedric

[cedrik]

Hi,
I had no understand what "symettryPlane" condition mean in OpenFoam, I use an other soft where it means that the conditions one the one side of the boundary is the same that on the other side, whereas in OF it means just a "basic" symmetryPlane which make a symetrie. It is usefull if you have a symetry on your domain for exemple but it's no use to define a boundary limit.

[gregjunqua]

if you use symettryPlane you need to link it to one other face. But it's only good for geometrical design (in order to save memory and time). slip condition is a good condition.

you should have a look on the thesis : Wind resource in complex terrain with OpenFoam June 2011 by Benjamin Martinez (it s almost the same as turbine sitting tutorial)

It's a very good thesis, everything is well explained from the boundary to the resolution. It helped me a lot.

[cedrik]

Hi Greg,
Thanks for your reply about my boundary condition. I found the these you advice me toread, it's indeed very interresting for my studies!
I'm trying to reproduce a study somebody else has done on an hospital, I manage to reproduce the general comportement but there are détails I still can't reproduce (picture).
In your opinion, where does the problem come from? The precision of the snappyHexMesh (I tried to incress it several times but I still can't have the comportement I joined my file) or an other parameter I don't see yet?

[donalder]

hi!!cedrik! i can't found the these above advised by gregjunqua! could you sent it to me use email ? thanks