Moment for shell

[pato]

Because the field I work on, I am more interested on moment for a shell. However, it seems there is no moment assign function for shell. So I choose still assign force in PreView software, but I change the ‘roles’ of force and moment in the FEBio codes as following:

(in void FESolidSolver::NodalForces(vector<double>& F))
// n = node.m_ID[bc]; // old line for force
n = node.m_ID[bc+3]; //new line for moment

I find if I assign the moment about ‘x’ axis, however, the result is like bending about ‘y’ axis.
I will appreciate if someone could give me the answer or hints.

Thanks in advanced.

Kun

[pato]

here is the example.

Because the field I work on, I am more interested on moment for a shell. However, it seems there is no moment assign function for shell. So I choose still assign force in PreView software, but I change the ‘roles’ of force and moment in the FEBio codes as following:

(in void FESolidSolver::NodalForces(vector<double>& F))
// n = node.m_ID[bc]; // old line for force
n = node.m_ID[bc+3]; //new line for moment

I find if I assign the moment about ‘x’ axis, however, the result is like bending about ‘y’ axis.
I will appreciate if someone could give me the answer or hints.

Thanks in advanced.

Kun

[pato]

Hi ,

Is there something I do not clarified in my question?
Thanks.

Kun

[maas]

Hi Kun,

The shell formulation in FEBio doesn't use rotational degrees of freedom. Instead it uses a so-called director approach where the additional degrees of freedom are the components of the normal at the shell node. (see the Theory manual for details). When you apply a force to the degree of freedom as you defined above you are essentially modifying the x-component of the shell normal. And that this is probably why it seems that you are bending about the y-axis.

To be honest, I don't think there is an easy way to apply a moment directly to a shell. You might be able to do this by attaching the shell to a rigid body and then apply the moment on the rigid body, but I've never tried this, so I'm not sure if this would work.

Cheers,

Steve.

[pato]

Hi Kun,

The shell formulation in FEBio doesn't use rotational degrees of freedom. Instead it uses a so-called director approach where the additional degrees of freedom are the components of the normal at the shell node. (see the Theory manual for details). When you apply a force to the degree of freedom as you defined above you are essentially modifying the x-component of the shell normal. And that this is probably why it seems that you are bending about the y-axis.

To be honest, I don't think there is an easy way to apply a moment directly to a shell. You might be able to do this by attaching the shell to a rigid body and then apply the moment on the rigid body, but I've never tried this, so I'm not sure if this would work.

Cheers,

Steve.

Hi Steve,

Thank you very much for your reply.

In your reply "you are essentially modifying the x-component of the shell normal", do you refer to the x-direction of the shell normal? Because I think in my case the x-component of the normal is zero.

I also check the Betsch's shell paper which you implement, it seems that they also do not say more about the 6 force components. Could you please give me some more details?

Thanks.

Kun

[pato]

Hi Steve,

I would like to add an explanation about my questions. The "x-component" in your reply is based on the local element coordinates or global coordinates?

Thanks again.

Best
Kun

[maas]

Hi Kun,

The element formulation in FEBio is not exactly that of Betsch', at leat not yet. We have to do some more work to get the same formulation. Regardless, the basic idea (using a director as an alternative to rotational degrees of freedom) is the same.

Even if your x-component is zero, applying a "force" on it will make it nonzero. And yes, this is in global coordinate system.

Cheers,

Steve.

[pato]

Hi Steve,

Thanks for your kindly reply.

Kun

Hi Kun,

The element formulation in FEBio is not exactly that of Betsch', at leat not yet. We have to do some more work to get the same formulation. Regardless, the basic idea (using a director as an alternative to rotational degrees of freedom) is the same.

Even if your x-component is zero, applying a "force" on it will make it nonzero. And yes, this is in global coordinate system.

Cheers,

Steve.