Are principal strains fully functional yet in Karamba?
I want to analyze principal strains of a simple steel shell, that is being pulled. The results seem a bit messy and incorrect. Here is a comparison between Karamba and RFEM:
Karamba
RFEM (expected)
Furthermore, why do I get a rigid body mode in Karamba, when rotation around z on the left linear support is not blocked?
strainsTest.gh (39.9 KB)
Dear @rudolf.neumerkel ,
The Principal Strains Approximation component you are using is designed to estimate principal strains for gridshells (see reference here), where you can also find a detailed description of its functionality. The output is based on three points located near the input point for which the approximate strains are calculated.
This method also works for shell elements; however, in that case, you have the option to output exact principal stress lines and vectors aligned with the principal strain directions.
In your situation, the mesh is fairly regular, and you are using mesh vertices to generate the approximate strains. Occasionally, the approximation algorithm identifies neighboring vertices that are nearly collinear, which can lead to numerical issues due to the limited precision of computer calculations.
To resolve this, you can either:
Use the exact principal stress directions, or
Provide seed points to the Principal Strains Approximation component that are located at the centers of your mesh faces (see here: strainsTest_cp.gh (57.4 KB)).
– Clemens
I am specifically looking for principal strains… which of course differ from principal stress directions. Which is why this sentence is confusing me, as for shell the 1D Hookes law cannot be applied:
The principal strains are tangents to the principal stress lines of a structure.
They technically probably are always parallel, but definitely not proportional in magnitude, in many cases even with a parallel “length” of 0, thus in another axis
Seems to work perfectly now. Sorry, I glanced over the first paragraph of the documentation, thinking I had already got what it said. The documentation mentiones, that the principal strains approximation can be used for gridshell. Is it however also applicable for shell continua?
One example: red = principal stresses, cyan = principal strains
Dear @rudolf.neumerkel,
ad 1.) For isotropic, linear elastic materials the principal stress- and strain directions are aligned; for anisotropic maerials this is not the case. The magnitude of principal strains and stresses are of course different.
ad 2.) It is possible to apply the approximate strain calculation to shell elements, but it’s important to keep in mind the limitations inherent in the underlying algorithm.
– Clemens
Thanks for clarification and for your quick response!
