I want to use shell elements in Abaqus in order to formulate a cylindrical shell under compression. However, I am stuck a little bit in small and finite strain part. Does this mean that elements with small strains are unsuitable when nonlinearities come to light? And how could i know when small strains are occur in a complicated structure?
Dear Spyros Arvanitis
The selection between finite and small strain shell elements depends on the way in which the deformation of the cylindrical shell is accomodated in the structure. The membrane strains of the elements is the critical factor that determines if the finite or small strain formulation should be used.
For example, if the shell structure does not buckle during the analysis and the analysis is geometrically linear (NLGEOM=OFF) then I think that selecting small strain elements should give relatively accurate results.
The same analysis with NLGEOM=ON (and no buckling response) allows for large membrane strains in the elements, so in that case finite strain elements should be used.
Finally, if the cylindrical shell undergoes large scale buckling with small amounts of membrane stretching and compression, then I think that small strain elements should be enough.
You can check the strains that develop in the shell elements directly by selecting suitable outout variable identifiers under the *ELEMENT OUTPUT option (e.g. E).
George Papazafeiropoulos thank you for response. However you mentioned the case of buckling. Assume the next two cases:
1) Buckling of a thin cylindrical shell under axial compression. In this case can the Abaqus thin shells ( with large rotations and small strains) to predict and give accurate results of post-buckling behaviour when geometric-nonlinear analysis is applied?
2) Buckling of a thin cylindrical shell under pure bending. In this case where we have a combination of local buckling and nonlinear prebuckling behavior( ovalization) can small strain shell give accurate resulates with such kind of deformations?
- Badges
- Science topic