[monnda]

Hello everyone!

I am currently working on a fsi problem with solids4Foam and I have a small question about boundary conditions.

I want to apply load on the solid patch but I want the solid to move towards only one direction. I thought I can do this with some sort of direction mixed boundary conditions and I found solidDirectionMixed and fixedNormalDisplacement.

Problem here is that I do not know how to use these boundary conditions and the difference between them.

I would be appreciate if someone could help me.

[bigphil]

Hi,

For each Cartesian direction, it is possible to apply a force or displacement (not both).

So what exactly do you wish to apply? e.g. force in normal direction, zero displacement in tangential direction , or zero displacement in normal direction, zero force in tangential directions.

Philip

[monnda]

Hello,

I want to apply traction in y-direction but I want to constrain the movement of the solid body in x-direction as well.

[bigphil]

Originally Posted by monnda

Hello,

I want to apply traction in y-direction but I want to constrain the movement of the solid body in x-direction as well.

What is the unit normal of your boundary?

Philip

[monnda]

Hello Philip,

My case looks basically a journal bearing so normal vectors are pointing towards outside of the shaft. Force is applied on the end on the shaft. Is it possible to constrain the displacement or the rotation of the shaft?

[bigphil]

What is the unit normal of the boundary where you want to apply the boundary condition? I can't advise without this information.

[monnda]

Hello

I'm not quite sure if I understand your question correctly but here is the picture of the shaft inside the journal bearing.
The blue part is where I wanna apply the force and the white part is the fluid-solid interface. Other parts are just solidTraction so that they deform freely.

shaft.jpg

[bigphil]

So just so I am clear, on the blue patch you want to apply a traction in the y direction and set the displacement to zero in the x direction; is that correct?

This can be done with a modified version of the fixedDisplacementZeroShear boundary condition: the valueFraction controls the "fixed" (displacement) direction and then you can set the displacement with refValue and traction iteratively via refGradient.

[monnda]

Thanks Philip! That is exactly what I wanted to know!

But what about solidDirectionMixed boundary condition and fixedNormalDisplacement?
How are they different?

[bigphil]

solidDirectionMixed is the same as directionMixed except that solidDirectionMixed implements non-orthogonal correction.

fixedDisplacementZeroShear derives from solidDirectionMixed. It applies a user-specified vector displacement in the patch normal direction, where only the normal component of this vector is used; and it then applies zero force/traction in the patch tangential directions.

fixedNormalDisplacement is the same as fixedDisplacementZeroShear, just a different implementation/implementor; I plan to remove it.

I have just created a new boundary condition called displacementOrTraction, which I have pushed to the development branch. This boundary condition can act as displacement or traction or displacement in a direction or traction in a direction for each face, depending on how the valueFraction and specifyNormalDirection fields are set.

You can achieve the behaviour you want with the following settings:

patchName
{
    type    displacementOrTraction;
    constantTraction uniform (0 1e6 0); // or whatever you like
    constantDisplacement uniform (0 0 0);
    valueFraction uniform (1 0 0 0 0 0); // X direction is displacement, Y and Z are traction
}

valueFraction means the same as for directionMixed, and optionally a field called specifyNormalDirection allows the normal direction to be set as displacement or traction.

It is available now on the development branch, which I will merge with the master once I have performed the usual checks.

Philip

[monnda]

Hello Philip!

Sorry for being late to reply but I just checked your boundary condition worked just like what I wanted!

Thank you very much!

[frantov]

Hi Phillip

I have a probably related question.

I would like to restrict the movement of a boundary to only +y (assuming at t=0, y = 0)


This is because at y- in real life, there is another surface (contact). But I dont want to model the contact.

Maybe I dont have another option but modelling the contact, and hence include friction, right?

Cheers
Fran