How to model analytical rigid surfaces

[Odforu]

Hi,

I want to run a simulation of a composite material consisting of fibers and Matrix. The top surface of the model has to be rigid in order to apply load. I would like to model a rigid plate.I am planning to use analytical rigid surface approach. I have never used analytical rigid bodies before, do I have to create a plate with 3D extrusion shell under analytical rigid profile in the part module and then later on assemble it with the model in the assembly module? When do I apply tie constraints? What about meshing of the rigid body?

Thank you.

 

[btrueblood]

Why not just apply fixed boundary conditions along that surface?

 

[Odforu]

@btrueblood: The fixed boundary conditions are applied on the other i.e. bottom surface

 

[xerf]

You cannot mesh analytical rigid surfaces, only discrete rigid surfaces can be meshed.

For each rigid surface you need to define a single reference point which can be used for applying loads/displacements.

Typically the rigid parts interacts with other parts via contact interactions.

 

[Odforu]

@xerf: I don't know where to start from, I want to model a rigid plate on the top of my model. So how do I begin, in the part module, should I go to the analytical rigid module and then select a rectangular profile and extrude it as a shell. Then in the assembly module assemble the 3D rectangular geometry with my model and then select a reference point on the rectangular model. Is this the procedure?

Correct me if I am wrong.

Thank you.

 

[btrueblood]

"@btrueblood: The fixed boundary conditions are applied on the other i.e. bottom surface"

So, apply a fixed displacement b.c. in the direction of the force, with other displacements set to zero.

One run will tell you the resultant force for an aribtrary displacement. Scale the displacement appropriately to get the force you want. Two runs, done.

Alternatively, go through all the trouble of meshing a seperate body, and then set the stiffness for those materials to an artificially high value. Force the solver to give you all of the stress and strain info. for the rigid body, which you can then ignore. But your CPU will hate you...

 

[corus]

An alternative to btrueblood's method would be to use *eqn to link the freedoms in the direction of the force and then apply a force to some point on the surface. The results under this 'rigid surface' wouldn't mean much though.

Tata