Compressive stress in penstock ASCE-79 1

[DamsInc]

Please inform me if I should have posted this in the pressure vessel forum?

I am a junior structural engineer tasked with creating an alternative conceptual design for an 8’ dia. exposed steel penstock. To complete this task I have been reviewing design literature, focusing mainly on ASCE-79 “Steel Penstocks” as that seems to be the most commonly used standard, based on specifications on other projects.

My question has to do with the allowable compressive stress in the penstock shell. Equations 4-9, 4-10 describe a limit to the allowable compressive stresses in the penstock shell (stress must be less than 1.5*10E6(t/r), or 1.8*10E6(t/r). I understand that this is a limit to prevent local buckling in the shell under pressure, similar to the limits in the CISC handbook to prevent local buckling of a thin flange in a beam.

When reviewing the design example in Chapter 17, this limit is not checked when evaluating the compressive stress due to beam action at mid-span. It only appears to be used on page 372 when evaluating the bending stress over the supporting saddle.

Adding to my confusion, the design example references ASME UG-23(b), which is not previously mentioned in the guide.
For a ½” thk shell, this results in an allowable compressive stress of 15.6 ksi, much less than the basic allowable stress intensity of 0.3*70=23.3 ksi for ASTM A516 Gr. 70.

I would assume that this limit should be applied to all areas of the penstock shell, however this lower limit significantly reduces the span I can achieve between supports for a given thickness, much less than that listed in the typical span tables.

Although AWWA M11, page 49 describes a different formula for the limit on the compressive stresses, I still do not see this limit checked in their design examples for beam bending stress.

Can somebody shed some light on why I am not seeing this limit checked in these design examples? Is it as simple as the examples are not complete? This would not make sense to me because the numbers I get using the limit don’t jive with typical span values, suggesting that it does not need to be checked mid-span.

 

[JStephen]

The equation on page 49 of M11 is for external pressure, not bending moment (referring to page numbers in the 4th edition.)

See the example starting on page 79.

This is not my normal field, but it looks like what is going on is that they are combining pressure stress and bending stress via the Hencky-Mises theory, which with higher pressure, results in a relatively low net allowable bending stress, so they don't bother checking buckling. If you had a similar span with low pressure, you would need to consider buckling.

There are a number of different compressive stress criteria available for cylinders. In some cases, allowable stress is taken quite a bit higher when due to bending than when due to uniform compression.

 

[DamsInc]

Ah yes, I see now that formula is for external pressure (in the case of a penstock, this would be vacuum conditions and should be relieved by adequate venting).

I reviewed my spreadsheet and realized I had a critical error in my calculation for bending stresses, resulting in values that were much too high. After fixing my error I can see that the compressive stress due to bending is much lower than I originally thought, and that the compressive stress limit would rarely govern in this case. It would make sense then that they didn't bother checking it at this location.

Thank you.