FEM - maximum tensions

[Pfeife]

Hello,

I am currently using Marc / Mentat as an FEM program as part of a specialist work and am investigating the pressure stability of a triaxial ellipsoidal shell (or an eighth of it, the symmetry makes it representative - picture in the appendix). For this I drive a path-controlled load, that is, a non-
deformable plane which moves from the outside along one of the three radii to the inside. The geometry is getting squeezed.

This produces tensions that are expected to be highest at the point of contact. Maximum von Mises stresses of approx. 2500 N/mm² now occur.
The material parameters I used are an Emodule of 27500 N / mm² and a Poisson's number of 0.3.

The question now would be: How do I know when a failure of the material can be expected? I would like to be able to make a statement about the increment or degree of compression from which the ellipsoid takes irreversible damage.

Much Greetings,

the Pfeife

[NewJoker]

What kind of material is that ? This determines the failure criteria to be used.

Do you use solid elements for this model ? Shells might be better in such a case.

[Pfeife]

It is kind of an unknown material. All informations i have is the youngs modulus and the poisson-ratio.
And yes i am using solid elements. Is it just because of the computing time or why should i use faces?

Thanks so far for the answer!

[NewJoker]

Then it will be hard to interpret the results of the simulation. You could at least compare them to yield or tensile strength but if you don’t have these values then you can’t predict failure and only deflections might be useful.

Shell elements are much more efficient in most cases involving thin-walled structures. Their use decreases solution time since you would need quite a lot of solid elements (at least 4-5 in the thickness direction) to obtain accurate results. You can always compare the performance of different element types in this particular case.

[paul18]

Well I do not understand the link between the pressure stability, and the use of a rigid surface; furthermore you've been using the sentense "The geometry is getting squeezed", so:

• if you're working on the pressure stability i.e. the pressure is applied on the outside of the ellipsoid, you've to take care on the structure buckling; on the other word, collapse occurs pior to yielding
• the second item involves contact modelling: for areas in tension, yielding and necking will lead to the damage, but for areas in compression, that's the buckling (what will appear first ?); a non-linear solver is needed and the Riks method must be activated

For thin structures, shell elements must be used (quadratic quad elements) on the complete structure.

An example here of buckling : https://www.youtube.com/watch?v=jAkNuQAvK3g

[thschrader]

...
An example here of buckling : https://www.youtube.com/watch?v=jAkNuQAvK3g

Thanks for the example. I ran a linear buckling analysis with 2D-elements in FC.
For the nonlinear analysis the tutorial uses predefined imperfections in Abaqus.
To do that in FC I see two ways:
1. Construct a cylinder with imperfections
2. Use the deformed mesh from linear analysis as a basis for nonlinear statics.
How can I do 2.?

imperfections.JPG (104.21 KiB) Viewed 3002 times

[paul18]

Try using quad elements and compare results to tria ones. Did you used linear or quadratic elements?

[thschrader]

Try using quad elements and compare results to tria ones. Did you used linear or quadratic elements?

Quadratic triangles. See here:

beulenBeispiel.FCStd

(15.12 KiB) Downloaded 39 times

EDIT:
I used E=73100 N/mm^2, nu=0,35 acording to the tutorial.
You must edit the inp after case-writing

Emodulus.JPG (11.59 KiB) Viewed 2902 times

Done with
OS: Windows 10 Version 2009
Word size of FreeCAD: 64-bit
Version: 0.20.26155 (Git)
Build type: Release
Branch: master
Hash: 0926a4148bcff11249fd4f56bc30256102ffe105
Python version: 3.8.6+
Qt version: 5.15.2
Coin version: 4.0.1
OCC version: 7.5.0
Locale: German/Germany (de_DE)