AnsweredAssumed Answered

Weld analysis results

Question asked by Umberto Zanola on Nov 10, 2015

Using SW simulation professional 2015 here.

Linear static analysis.


I tried to simulate a tube (shell elements) straight (no gussets) welded to a thick flange (solid elements), I use a the edge weld connector between the flange and the tube,

then I applied some forces perpendicular to the tube axis that generate a bending moment.

I have setup weld calculation according to European code and added some parameter like the expected size, one side edge welding (account for bending option on).

Edge Weld results window give me an ideal size over the welded edge with min max mean value and its graph along the edge and this is OK.

What I don't understand (and the help file doesn't state it) in the same result window is "stress value at critical location" (not available in American code analysis by the way):

Does this stress value  refer to the calculated weld size or the expected size of weld  I entered before running the simulation?


Incidentally I have another question not directly related to the weld connector as I get pretty much the same behavior with the bonded connector.


My model is basically:

- Flange (t 20mm) dia 550mm

- Counter flange  (same as flange)

- Pipe diam. 406 t 9.5 mm 6m long welded to the counter flange (weld size is not important)

- Flange and counter flange are bolted together, but we can ignore this excluding the first flange

- 6000kg Load over the pipe length (apply it like it was self weight over half of the pipe surface)


What I see on the tube is the expected "triangular" pattern of stress rising from the edge to the attachment at the flange.

Then I see a huge stress concentration (Von Mises) over the pipe at welding expected position, and I can somehow understand (and eventually ignore) it.

What it seem not so intuitive is a sort of gap between the flange and the max stress area (the real one), basically it is not at the beginning of the pipe,

but a bit shifted, try the simulation and you will see what I mean.

Is it an effect of the circular section shape and the composed stress from shear and bending?


PS cannot upload files from here, but the shapes are very simple to reproduce

This is a core part of a more complex structure attached to the pipe so no beams please.

Excel calculation show that the pipe should take up to 8000kg while FEM is around yield (240MPa) at 6000kg (ignoring the weld issue)