How Does Mesh Refinement Affect Shell Deformation in ANSYS?

[atlamillia]

Homework Statement

A square plate of edge length L = 1 m, thickness t = 5 cm, is fixed on two edges. The
other two edges are free. A load of F = 10 kN is applied at the corner opposite that where the fixed edges intersect.
Use shell elements to compute the deflection at the loaded corner. Assume E = 200 GPa,
ν = 0.3. Start with a coarse mesh (2 × 2), and successively refine the mesh until the change
in the displacement at the loaded corner is sufficiently small. Show the results in a tabular
form.

Homework Equations

None.

The Attempt at a Solution

Modeled a 1x1 area.
Define Element Type> SHELL28
Real Constants> thickness = .05
Material Models> linear elastic isotropic > E = 200000000000, v = .3
Meshing> Size options> Area element side length = .5
Meshing > Mesh >area

Apply>Deformation> Lines, 0, all DOF
Apply > Load > on node, UZ, 10000
Solve

It doesn't solve, instead throwing an error about a small pivot term in the UX direction.
I was wondering if my problem might be my element type, but I can't figure out how to set a uniform thickness for any other type of shell element.

Thanks for any suggestions!

 

[KataKoniK]

Thank you for your post. It seems like you have made some good progress in setting up your problem. However, there are a few things that you may want to consider in order to successfully solve the problem.

Firstly, the error about a small pivot term in the UX direction may indicate that your mesh is too coarse. This can happen when the element size is too large and the element becomes distorted, leading to numerical instabilities. I would recommend trying a smaller element size and see if the problem can be solved. You can also try using a different element type, such as SHELL63, which may be more suitable for your problem.

Secondly, for shell elements, the thickness is automatically calculated based on the element size and the specified thickness in the real constants. Therefore, you do not need to manually set a uniform thickness for each element.

Lastly, for the load application, it may be more appropriate to apply a distributed load over the surface of the plate rather than a point load at a single node. This will more accurately represent the load distribution on the plate.

I hope this helps and good luck with your analysis!