Showing results for 
Search instead for 
Did you mean: 

Different models, but the same results


I want to know the opinion of experienced FEMAP users on the following question.
There are 3 models for the central loaded curved circular tube.
Model 1 - the same solids are used along the length of the pipe, the thickness of the tube wall is given by 1 element


Model 2 - a shredded (reduced, non-equivalent) mesh is used, the tube wall thickness is set by 3 elements, the mesh size along the pipe is variable - a fine mesh in the middle of the tube length


Model 3 -here I used elements of type Plate


In all models, a non-linear static analysis is performed. Material is a bilinear diagram.

I'm interested in the dependence "Load - axial shortening"

For all models, I obtained practically the same results and almost identical values ​​of the maximum force.


Also here I gave a comparison of the stresses (nonlinear VonMisses) in the most stressed element (solid type) in the cross section of the pipe along the middle of the length. Apparently, here, too almost the same result (for models 1 and 2).

Thus, I concluded that it is better to use a model with plates to analyze tubes with other sections.

1. The computation time is reduced, 2. The result will be almost the same as for the model with plate elements like the solids.

? Am I right?

If I'm wrong, tell me where my mistake is.
All three models are attached below.



Re: Different models, but the same results


Dear friends.
Excuse me for worrying and for my post above.
I received recommendations from Blas Molero Hidalgo, where it is written in detail how to correctly generate the Mesh grid.

Therefore, I will create another model and compare now 4 models.
I will add the result for model 4 below.
So do not waste your time.
See you soon.

Re: Different models, but the same results

Just FYI, you would normally use maybe 3 or more solid 8 node brick elements through the thickness if the stress condition produced "through-thickness" bending, ie. local bending within the wall of the tube. The reason why the results are so close is because you are modelling a fairly simple "beam" type stress condition: tension at one extreme fibre and compression at the other extreme fibre - without much local bending through the wall thickness. This means that you could get by with only one solid element through the wall because of the mild variation in the stress field for any individual element.