Stress extrapolation for tubular joints - FE models 1

[loowheezey]

Hi!

I'm researching into stress concentration factors (SCFs) for tubular joints by running FE models in ANSYS. My models have been divided up in order to develop as symmetrical a mesh as possible and I've defined a region for the upper and lower extrapolation points. The extrapolation region has been defined so I can get a stress distribution plot going around the perimeter of the brace. Results reporting on the upper and lower bounds are then linearly extrapolated back to the weld toe to give the stress at that point.

With the case of a K joint, overlapping of extrapolation regions on the chord is possible when the gap between braces at crown locations reduces. How do you go about reporting stresses in this situation for extrapolation? My models include the weld detail however reporting results directly at the weld toe gives higher stresses due to discontinuities at this location.

For reference, I'm using DNV-OS-J101 2010 Design of Offshore Wind Turbine Structures which gives the extrapolation points in Appendix D Cl 102 Figure 2.

Any help would be much appreciated! The papers I've reviewed cover such K joints but there seems to be no indication of how the authors have dealt with this situation.

Many thanks,
Louise

PS - If you need more info, please PM me or indicate so in your reply. I'll endeavour to help explain further!

 

[flash3780]

It seems pretty well laid out in the document you provided (DNV-OS-2010). What specifically are you having trouble with?

I did notice that they got the colors wrong in the Figure 4: gray=green, light gray=yellow, and dark gray=red.

From the looks of it, they're having you plot stresses along a path leading up to the weld and performing a straight line extrapolation from two points at a distance defined by the weld radius and the thickness of the welded parts.

It looks like figure 2 would be most applicable to your situation. I would consider modeling the weld geometry as in figure 4a, rather than omitting it.

 

[loowheezey]

Hi flash3780!

Thanks for your reply. I'm happy with the procedures for conducting the stress extrapolation - I'm just stuck as to what to do when it's not possible to plot the regions due to the geometry of the joint.

Attached is a picture which hopefully should clarify things. As you can see, I've included the upper and lower bounds of the extrapolation region in the model. In this case, the gap between planar braces is only 50mm so it's not possible to fully plot the extrapolation region as shown in the image.

What approach should therefore be taken for such a case? The extrapolation points are positioned close to the weld toe so stresses are likely to be much higher due the model's discontinuities. I'm a bit stumped by this and don't want to exclude certain geometries because of an FE modelling issue!

Cheers!
Louise

 

[flash3780]

Louise,
You're right. The method outlined in the referenced document does not apply to that situation. I would do a hand calculation to determine the adequacy in that location. Bolodgett's Design of Welded Structures is a good reference for that sort of calculation.

For what it's worth, if you overlapped the smaller pipes, you'd likely have a much stronger joint.