[xena]

Hello Foamers,

I simulate a partially filled liquid tank subjected to external exitation using an FSI solver based on icoFsiFoam by replacing the fluid solver with InterDyMFoam.

In order to make a motion in the liquid, an acceleration term is added in the equations of the two physics (fluid & solid).


1- for the fluid-strcture interface and the outter boundaries, I'm using this condition

{
type tractionDisplacement;
traction uniform (0 0 0);
pressure uniform 0;
value uniform (0 0 0);
}

2- The bottom of the tank walls are fixed.

{
type fixedValue;
value uniform (0 0 0);
}


I looked in the solid equation and there is no gravity, hwo can I add it ?
"code :"

""do
{
volTensorField gradU = fvc::grad(Usolid);

fvVectorMatrix UEqn
(
fvm::d2dt2(Usolid)
//- acceleration_s //new 'source term'
==
fvm::laplacian(2*mu + lambda, Usolid, "laplacian(DU,U)")
+ fvc::div
(
mu*gradU.T() + lambda*(I*tr(gradU)) - (mu + lambda)*gradU,
"div(sigma)"
)
);

initialResidual = UEqn.solve().initialResidual();

} ""

question1 : the big problem I'm facing is that the model work for a very very rigid material, but dosn't for a less rigid one (E=210.10^9). does anyone face that ????


question 2 : Hwo to apply a displacement or acceleration (a*sin(w*t)) in the bottom of the tank (not as a source terme but a BC in the solid bottom) ??

question 3 : in the setPressure.H file, we have this {// Setting pressure on solid patch
Info << "Setting pressure" << endl;

scalarField solidPatchPressure =
interpolatorFluidSolid.faceInterpolate
(
p.boundaryField()[fluidPatchID]
);

solidPatchPressure *= rho;

tForce.pressure() = solidPatchPressure;
}
Why is the pressure multiplieed by rho ?? can I desable it ?


Thanks in advance for answering.

Best regards

[hxaxtma]

Why is the pressure multiplieed by rho ?? can I desable it ?

...cause the pressure p is divided by density rho for incompressible solvers in OpenFOAM. Have a look at the dimensions.

[xena]

Originally Posted by hxaxtma

...cause the pressure p is divided by density rho for incompressible solvers in OpenFOAM. Have a look at the dimensions.

Thanks for answering !

Okay I see. but in my case (a multiphase solver) I checked and the pressure is not divided by the density, so I don't have to multiply by rho.

btw, the 'rho' here is the density of the mixture (liquid+gas), isn't it ?

[bigphil]

Hi xena,

I am not sure if you solved your problem but you may like to try out solids4foam, where interDyMFluid is already implemented as a fluid model. PM me if you would like access.

Philip

[xena]

Hi Philip,

I solved some problems but still have others !
Yes I'm interested to use FSI under solid4foam (interDyMFoam for the fluid).

Please can you tell me if there is problem of damping in strctural response with solid4foam ?

In my case, the structural displacements are damped despite that no damping term is added in the equations. (maybe numerical ?)

I can give more details (configuration, numerical schemes ...) of my case if needed.

regards,

[bigphil]

Hi Xena,

Originally Posted by xena

I solved some problems but still have others !
Yes I'm interested to use FSI under solid4foam (interDyMFoam for the fluid).

Great.

Originally Posted by xena

Please can you tell me if there is problem of damping in strctural response with solid4foam ?

In my case, the structural displacements are damped despite that no damping term is added in the equations. (maybe numerical ?)

Like many numerical methods, numerical dissipation/damping can be an issue. For solids4foam, the default time scheme is 1st order Euler which is known to be highly dissipative (adds a lot of numerical damping): you can find a number of threads on this issue if you search the forum. It can be solved using one of two methods:

1. Use a smaller time-step: this will reduce the damping. In fact, I suggest you perform a time-step sensitivity analysis
2. Use a higher-order time scheme: you can try "backwards" which is second order, however, this introduces quite a bit of dispersion errors, so it would be better to implement another 2nd order scheme, like the trapezoidal rule or the Newmark-beta scheme

Philip

[xena]

Thank you Philip,
I PM you about interDyMFoam under solids4foam