[Aliiiii]

Hi all,

I’m trying to run a Fluid-Solid-Interaction simulation of a series (2-4) of vertical beams in a shear flow. The problem is very similar to the beamInCrossFlow tutorial and the main difference is that I’d like to run the simulation for a Couette flow (linear velocity profile) rather than pipe flow (parabolic velocity profile).

In order to impose the right BC, I went through the cavity tutorial and adopted the new BC for my problem accordingly.

These are the BC that I use for my problem (2D):

U:

boundaryField
{
top
{
type fixedValue;
value uniform (1 0 0);
}
bottom
{
type fixedValue;
value uniform (0 0 0);
}
inlet
{
type zeroGradient;
}
outlet
{
type zeroGradient;
}
interface
{
type movingWallVelocity;
value uniform (0 0 0);
}
frontAndBack
{
type empty;
}
}




P:
boundaryField
{
top
{
type zeroGradient;
}
bottom
{
type zeroGradient;
}
inlet
{
type fixedValue;
value uniform 0;
}
outlet
{
type fixedValue;
value uniform 0;
}
interface
{
type extrapolatedPressure;
value uniform 0;
}
frontAndBack
{
type empty;
}
}


pointMotionU

{
interface
{
type fixedValue;
value uniform (0 0 0);
}
outlet
{
type fixedValue;
value uniform (0 0 0);
}
inlet
{
type fixedValue;
value uniform (0 0 0);
}
bottom
{
type fixedValue;
value uniform (0 0 0);
}
top
{
type slip;
}
frontAndBack
{
type empty;
}
}



controlDict:


application fsiFoam;

startFrom latestTime;

startTime 0;

stopAt endTime;

endTime 4;

deltaT 0.001;

writeControl timeStep;

writeInterval 10;

purgeWrite 0;

writeFormat ascii;

writePrecision 6;

writeCompression uncompressed;

timeFormat general;

timePrecision 6;

runTimeModifiable yes;

adjustTimeStep no;

maxCo 0.2;

functions
(
beamReport
{
type pointHistory;
functionObjectLibs
(
"libpointHistory.so"
);

refHistoryPoint (0.45 0.15 -0.15);

region solid;
}

// I comment out this part to avoid parabolic initial velocity Ipipe // flow)
// setInletVelocity
// {
// type setInletVelocity;
// functionObjectLibs
// (
// "libsetInletVelocity.so"
// );
// }

forces
{
type forces;
functionObjectLibs ( "libforces.so" );
outputControl timeStep;
outputInterval 1;
patches (interface);
pName p;
UName U;
rhoName rhoInf;
log true;
rhoInf 1000;
CofR (0.5 0.1 0);
}

);


I also changed the 3D mesh of the original problem (beamInCrossFlow) to a 2D mesh.


This is how the initial configuration of the problem looks like (file attached - initial).

The problem that I run into is that the simulation dies after the beams reach their max deflection and this is the error that I get:

Time = 2.44, iteration: 6
Current fsi under-relaxation factor (Aitken): 4.85002e-23
Maximal accumulated displacement of interface points: 1.19232e-21
GAMG: Solving for cellMotionUx, Initial residual = 2.72925e-05, Final residual = 9.15396e-07, No Iterations 10
GAMG: Solving for cellMotionUy, Initial residual = 5.57559e-06, Final residual = 8.20284e-07, No Iterations 4
GAMG: Solving for cellMotionUx, Initial residual = 3.8065e-05, Final residual = 6.18694e-07, No Iterations 12
GAMG: Solving for cellMotionUy, Initial residual = 5.24285e-06, Final residual = 7.8125e-07, No Iterations 4
Evolving flow model: consistentIcoFlow
Courant Number mean: 19.8487 max: 3657.12 velocity magnitude: 1069.42
DILUPBiCG: Solving for Ux, Initial residual = 0.97421, Final residual = 3.53006e-07, No Iterations 874
DILUPBiCG: Solving for Uy, Initial residual = 0.976795, Final residual = 3.25234e-06, No Iterations 1000
Floating point exception (core dumped)


Also I get the same error for when I reduce the width of the beams to half the previous thickness (or when I increase the shear rate (velocity of the top plane)):


Time = 0.74, iteration: 5
Current fsi under-relaxation factor (Aitken): 0.0251294
Maximal accumulated displacement of interface points: 1.42491
Floating point exception (core dumped)

I've attached some images of the initial configuration and the Fluid-Solid interface mesh as the time evolves. There seems to be a separation between the fluid and solid meshes which increases with time and I suspect that causes the error but I don't know how to fix that.



I am pretty new to OpenFOAM and I apologize if I'm not using the right terminology in my post. Any help/hint is appreciated.


P.S. I'm using foam-extend 3.2.

[paulbr]

Hi,

This is an old post but, have you found a solution to your problem ?
I'm facing the same issue, which is I think related to the displacement transmission at the Fluid Structure interface.

The long story short: I'm trying to rotate a beam in a flow. Yet, the fluid mesh at the interface stays really close to the structure mesh except at the fluid boundary. This also seems to be your problem.

Have you manage to make your case worked ? And how ?

Regards,



Paul