[sultzan]

Hello there,

I'm working on the problem of some untypical pressure vessels, those are geometries of revolution. Im looking to compare FEM results with an analytical solution based on the membrane stress concept. Therefore I would like to read two resultants of the stress: longitudinal and circumferential. Im taking into consideration quarter of half of the surface model (1/8th) in FEM analysis. Reading the circumferential stress is easy, but does anyone know how to read the longitudinal resultant? It feels like the most adequate way to do so is to read the stress resultant tangent to the path defined on the longitidinal edge. No idea how to perform such action. Membrane, linearized, normal and principal stresses cannot really show longitudinal stress, at least not in all cases.

Extra question concerning described analysis: can someone explain to me why Ansys does not analyse normal-to surface stress component while using first order (linear) finite elements (in this case shell181)? Using second order (quadratic) elements (shell281) always show some thickness-through stress value (which is expected to do so).

[LK_1000]

Can you post a picture of the analysis?

[sultzan]

There you go.