Circular partial penetration weld subjected to moment

[SiggiN]

Hi!

Suppose you have two circular segments welded to a pipe as shown below (pad eye plate prevents me to use a full circle)

These segments act as a "tempoary hang-off flange" and will see see a load like this:

I am curious about how to verify the weld size.

Initially I imagined using a force pair, but that will not take the effect of curvature into account?

I am assuming no bending stress in the top weld (only shear)

Would it be (more) correct to use the moment of inertia in the center of the bottom weld to calculate the bending stress in the weld:

sigma_bending = (M/I)*r

("r" being the outer radius of the pipe)

Any thoughts would be appreciated

Thank you!

Regards

 

[JStephen]

My thinking on it-
There's not going to be an exactly right way to do this.
But it's hard for me to imagine that the weld would ever fail first out at the ends, due to twisting in the attached segment.
With that in mind, for weld checking, I'd assume a force distribution as below, and check that little length of weld like it was two straight lines.
If that gives you a reasonable result, then great, if not, maybe ponder some more.

If the arc is long enough, you could figure the section properties in the plane of the weld and get some significant strength.