Stresses in a cylinder single lap fillet weld

[ScottAC]

Hi All, I am trying to work out the stresses in a single lap fillet weld. Looking at formulas for fillet welds there doesn't seem to be any for a cylinder single lap fillet weld. I thought about using the formula for a single lap fillet weld but this is for a flat plate and doesn't account for the stiffness of the cylinder? I've attached a pic/image that shows the weld placement (blue). There is a internal pressure (red). I've worked out the hoop and longitudinal stresses experienced by the cylinder. I think I know the forces on the weld via Ansys. Am I right in thinking of drawing a FBD of the weld but I am not too sure how to do this. Any light or direction would be greatly appreciated.

 

[tbuelna]

Since the design shown in your sketch seems to show a huge difference in radial stiffness (cross section) between the tube and ring right where the fillet weld terminates, it is likely you will need to apply a very substantial Kt factor to your analysis. Since the thick internal ring is far stiffer radially than the thin wall tube right where the fillet weld terminates, this will create a huge local stress concentration.

Also, due to the large relative difference in section thickness between the thin tube and thick ring, the fillet weld will likely result in a substantial amount of undercutting in the tube wall at the point where the fillet weld ends. A good ANSYS FEM will give far better results than any hand calcs will for this situation.

Good luck to you.
Terry

 

[desertfox]

Hi ScottAC

If your designing a pressure vessel then you really should be following a code, its not a good idea to use fillet welds as per your sketch, normally the ends of the vessel and the cylinder would be joined by a butt weld.
The butt weld would then be subjected to the same loads as the cylinder itself at least that's the simplified theory, although in actual practice the weld is also experiencing bending stresses etc do to the nature of the vessel ends, the codes for pressure vessels usually have enough safety margin in them for that effect to be negligible.
The link below shows how to make a butt weld between to differing material thicknesses.

http://patentimages.storage.googleapis.com/pages/US5591360-4.png

 

[ScottAC]

Hello, thanks for the reply.

tbuelna- Thank you Terry. I have done FEA analysis using Ansys but because the the thin ring is bonded by the edge touching the thick ring/cylinder (without the weld) there are huge stress values where the weld would be. Im not too sure but I think the values aren't accurate due to the edge bonding? Would it be a good idea to try and model the weld and add it then run the analysis again?

desertfox- I agree that butt welds would be a far better weld as the theory states that two pieces of material butt welded can be treated as one (same thickness). Unfortunately I am not the designer in this case and I am just looking at the stress in the welds but I will see if a butt weld is an option.

I've added a more accurate cross section pic that better represents the model.

 

[desertfox]

Hi ScottAC

I'm not surprised that your finding high weld stresses, you have longitudinal, hoop and bending stresses to cope with in that area.
For a rough estimate do a hand calculation and apply loads to the weld throat in both the vertical and horizontal direction to simulate the pressure.

 

[gtaw]

You want to pay particular attention to the root area of the single fillet welded lap joint. The tensile stress at the root of he fillet weld can be very high if there is any tendency to rotate about the root of the fillet weld.

Best regards - Al