Elbow Stress Analysis

[MJN2]

Hi All,

First post for me here so please excuse any faux pas.

I was asked to determine if a pipe support is required for a valve that is cantilevered off of an elbow on a vertical riser. This is for NG transmission which is governed by CFR part 192. I turned to ASME B31.8 for guidance and found that section 833 refers to Appendix E for calculating bending stress in fittings.

This is my first time performing this calculation so, I’d like to ask that someone who’s familiar with these calculations check my work in the attached spreadsheet. There is a sketch of the assembly in the spreadsheet as a visual aid.

The assembly is constructed of an 8”, STD wall, GRB vertical pipe coming out of the ground with an 8”, STD, WPB, 3R 90 on top. An 8” WNRF flange is attached to the elbow. The flange is connected to a FXF Nordstrom plug valve. The other side of the valve is blind flanged. To be conservative (read lazy) I calculated the moments of the flanges from the furthest points rather than trying to calculate the centroids. weights and dimensions were pulled from manufacturer websites.

Finally, the spreadsheet doesn’t include the flange factors because I was uncertain if it should be applied. It appears the correction factors reduce the stress intensification factors.

Thanks in advance for the help.

 

[RVAmeche]

Not familiar with this specific calc in B31.8 but this is just a dumb idea. Don't cantilever 850+ lbs and let it sag or break something. Either add a short spool on the blind flange side of the valve and support it or at minimum grab the elbow, which I still wouldn't love.

I don't see how this calc means anything when its looking this section in a vacuum and not considering other support locations or anything else.

 

[GD2]

OP,
I don't review any work in this forum. I can provide you with the following guidance:
1. The portion of the pipeline you are analyzing is 'Unrestrained'.
2. You will need to qualify the bending stress for both elbow and net section (straight pipe) separately. Bending stress is one component of the Longitudinal stress. Equations are given in Para 833 of ASME B31.8. Allowables are also given in the Code.
3. I haven't seen computation of M[sub]t[/sub]. You need to apply the rule of superimposition to do that.
4. Check calculation for H36.
5. You are almost there. You need some refinement.

GDD
Canada

 

[MJN2]

RVAmeche, I agree but this was installed around 1970 so I’m stuck with what’s there. We’re trying not to spend any extra $$ where it’s not necessary. This is one of several locations that need to be evaluated and I think my calculations show a support is not necessary.

 

[MJN2]

GD2, this is very helpful. Thanks for the info. I will check H36. Regarding M[sub]t[/sub] won’t that be 0 in this case? I can’t think of any loads that would cause torsion.

Also for anyone else that runs into this in the future, I found B31J to be a great reference for SIF and flexibility factors.