Strength of Welds on Valve Stem Coupling

[af09]

I work in the Valve & Automation industry and often design couplings to adapt a valve stem to the actuator drive. Long story short, sometimes these couplings are made from two welded pieces and I'm looking for anyone experienced in strength of weldments calculations to help guide me in calculating the strength of some welds. Currently they are partial-penetration groove welds under torsional load. See attached pictures for an example of a two-piece welded coupling design:

Pic 1 -

https://files.engineering.com/getfi...c99-955e-56a3a03e6ad6&file=Coupling_Pic_1.jpg

Pic 2 -

https://files.engineering.com/getfi...321-b349-551e8991f6bc&file=Coupling_Pic_2.jpg

Pic 3 -

https://files.engineering.com/getfi...4bd-9e8d-02a235e41614&file=Coupling_Pic_3.jpg

Pic 4 -

https://files.engineering.com/getfi...cdc-8989-3b718a7b0bd2&file=Coupling_Pic_4.jpg

Pic 5 -

https://files.engineering.com/getfi...bfe-b575-ba3f4e2a35f3&file=Coupling_Pic_5.jpg

Pic 6 -

https://files.engineering.com/getfi...358-addc-c3c2ce0fad7f&file=Coupling_PIc_6.jpg

Pic 7 -

https://files.engineering.com/getfi...8dd-8c42-248686e275aa&file=Coupling_Pic_7.jpg

Pic 8 -

https://files.engineering.com/getfi...899-aabd-5bfbba9eb3ec&file=Coupling_Pic_8.jpg

 

[TugboatEng]

It shouldn't be too hard to test one to failure. I think you'll find the keyed shaft will fail before your welded coupling does.

 

[af09]

I need a way to calculate this without having to physically test it.

 

[hacksaw]

You consult the various standards that might apply, or you test.

In your case there are no defined answers, given the quality of the welds, and materials involved, and the stress concentration factors, and fatigue issues you face.

 

[TugboatEng]

If you have a welder you can trust to make a proper weld that fills the groove completely plus a small crown, you can assume that the weld will behave as a tube with equal wall thickness to the depth of the groove. If it's a 300 series stainless steel the tensile strength will be right around 70 KSI. also run the calculation for the allowable loads on the keyed shaft and. I believe you will find the key shaft will fail long before the weld does.