Local Pipe Stress on Saddles 6

[john1525]

We have 10 thru 20 diameter (GA 10 = 0.135") pipes filled with water and supported at 8 ft OC. I am concerned with the local and bearing stresses at the supports (Clevis Hangers & Saddles)

Roark's Formula for Stress and Strain (page 642) has a formula to check these local strsses but I dont feel comfortable with it. Is there something that I should be aware or check besdies the local stresses?

S = k(P/t2) ln (R/t)
P = Load
t = thickness
R = Radius
k = 0.02-0.00012(beta-90) for R/t < 50
k = 0.03-0.00017(beta-90) for R/t > 28

 

[TERIO]

I don't have a good answer so I will ask some questions. What is it that is making you uncomfortable with the equation you are using from Roark:

- is there some failure mechanism that is not reflected by the stresses you are calculating?
- do you think the equation is incorrect or unconservative?
- unusual loading or geometry?

 

[Canadieng]

You could compare Roark to:

Design of Ductile Iron Pipe on Supports

http://www.dipra.org/pdf/pipeOnSupports.pdf

Zick

http://www.codeware.com/support/papers/zick.pdf

 

[Canadieng]

and the "other" thing that you should check besides local stresses at the saddles/supports is the stress due to bending between & over supports (basic beam theory - watch for "ovalization" when near supports)

 

[rconner]

While there may be nothing necessarily wrong with "comparing" a design involving support of light gauge steel pipe with the DIPRA pipe-on-supports procedure, I would caution that the DIPRA method and particularly some coefficients therein was developed after testing of actual ductile iron pipe cylinders exposed to support reactions and beam loads etc. Ductile iron has some unique properties, and particularly in compression and bending. A description of this testing program and results therefrom was incidentally published by researchers of the University of Alabama and ACIPCO many years ago in the AWWA Journal.

 

[DSB123]

john1525,
You could do what a number of piping engineers/designers do and ignore it. Remember pipe support makers sell you the supporting components, which are load rated, but do not guarantee that when using these components that you are not going to cause overstressing in the pipe - most pipe designers I have come across do not grasp this fact. If you can get hold of the very good book by van Laan covering pipe supports there are sections in there which promote formulae (actually from Roark) for calculating "local" stresses around supports. From what I remember in the example there is a typo so be careful. Anyway very thin wall as you have is vulnerable to buckling and ovalisation as a poster above has indicated. If you are that concerned why not "model" a section in FE. The time to build a simple model is not exhaustive if you have the software.

 

[BarryEng]

Have a look at AWWA M11 (steel pipe design & installation) for a detailed analysis of 'local' support stresses (generally large water utility pipes).

Also ASCE 79 (power penstocks) for similar (but more comprehensive) analysis & design (large power penstocks).

Bednar (pressure vessels) also has a good section.

For your small pipes, I would just use AWWA M11.

For very thin pipes (ducting), the ducting manual gives a conservative view of these stresses.

 

[MJCronin]

John..

Whomever designed the piping system had no sense of piping sizes or codes. (however, he may have been an HVAC designer...)

Specifically, to which code was this system designed to ?

If you have 20"OD pipe with 0.135 wall thickness, you have, essentially, ductwork filled with water.

The thinnest piping schedule in common usage is schedule 5S.... and for 20 inch OD, Schedule 5S is 0.188 inches.

Thank god there is no pressure in this system !! ( or....gulp... is there ??/?)

-MJC

 

[john1525]

Thank you

Process pipe for a water treatment plant. Therefore, B31.3