Hydro Dam Penstock Evaluation 2

[woodeispare]

I had originally posted this question in the Mechanical forum but I thought I would try posing here as well. Thanks for Reading.

I am working to determine the minimum thickness of a set existing steel penstocks (14' diameter steel pipes). The penstocks are supported every 15' by concrete saddles and there is some noticeable buckling around the horns of the saddles. I have seen the stress at this point called Zick's stress in some papers on penstocks. Would it be possible to use this method to do determine the stress from the support, and if so where could I find the formulas and methods used.

Another thing I will need to consider is water-hammer in the penstocks caused by closing the gates. There are 5 gates per penstock however so there will not be an immediate stop to the flow however the gates will close quickly and cause an immediate 20% reduction to flow. I would appreciate any good sources/methods for calculating water hammer pressure increase.

 

[LittleInch]

It's generally considered poor etiquette to double post so you may get red flagged.

You missed out that the penstock in the other post you say is over 100 years old!! How you will calculate anything form something that old is beyond me. You will have no real idea of any physical data about the pipe so cannot really calculate anything. Just work on adding more supports or providing more axial support by use of a ladder rack or something similar.

A photo or two or a drawing would help

Surge - that's probably why there are 5 valves. Basically the impact of shutting flow in stages like this is to make surge very low. I doubt you will see any significant increase in pressure, but you can model this in a transient analysis program or just attach a pressure transmitter and a recorder and then analyse the results.

My motto: Learn something new every day

Also: There's usually a good reason why everyone does it that way

 

[racookpe1978]

Isolate the penstocks, dry 'em out, and go measure thicknesses.

remember, old penstocks failed catastrophically in Niagara Falls power plants after many years service because of outside-the-wall concrete, rock, building wall water penetration and subsequent penstock support failure, not of pipe wall wear-through.

But that support and wall failure killed people, destroyed the whole power plant.

 

[rconner]

Back in the 1600's, a scientist named Edme Marriott did a lot of work involving various material beams in bending and ultimately pipe on supports, and most of a 35 km flanged cast iron pipeline subsequently built with commission of the "Sun King" Louis XIV to serve the Chateau de Versailles is still in service today, more than 350 years later. Raymond J. Roark, R.S. Hartenberg and others at the University of Wisconsin subsequently conducted instrumented research that was reflected in multiple papers in the 1930’s and 1940’s (after the installation of your steel penstock). Applying measured concentrated loads in various means to large, rather thin, cylindrical steel shells or pipes, these researchers measured metal strains and movements in the area of the supports etc. Rather ingenious early strain and movement indicating devices called “Huggenberger tensometers” (I think resembling popsicle sticks glued to the pipe) were used to mechanically amplify the metal fiber strains and deformations that were studied, so they could be visually observed and recorded as load was applied. Roark indicated that when the pipe fit the saddle well both circumferential and longitudinal local stresses reached a high value at a point roughly 15 degrees above the saddle tip, primarily due to circumferential bending of the pipe. This and other work formed the basis of the following equation developed for the maximum localized stress at common steel pipe saddle supports now contained in AWWA Manual M11 Steel Pipe – A Guide for Design and Installation (AWWA 2004) and other references:

Scs = ksteel (P/t^2)loge(R/t) where,

Scs = local bending stress at saddle (psi)
ksteel (or “contact angle factor”) = 0.02 - 0.00012 (A-90)
A = subtended “contact angle” (in degrees) of the saddle support
P = total saddle reaction (lb)
R = pipe outside radius (in.) and t = pipe wall thickness (in.)

As Roark and Hartenberg's work involved rather thin steel shells with high D/t ratios, researchers at AMERICAN (American Cast iron Pipe Company) along with the M.E. Department of the University of Alabama conducted more modern, strain gauge-instrumented, localized support reaction testing with lower D/t ration ductile iron pipes a half century later. It was found that with some minor tweaking of coefficients etc. the same basic form of equation for maximum localized stress near saddle supports also fit the results of the testing with that pipe as well, though along with the same location the maximum stress could either be circumferential some distance off the tips of the saddles (at the same, basic location as identified by Roark) or instead longitudinal in nature off the edge of some saddles. That work is now referenced in AWWA M41 Manual of Water Supply Practices for Ductile-Iron Pipe and Fittings.

 

[woodeispare]

Thank you for that info, that's what I was looking for.