Waterhammer

[Chris73]

Hi Eng's,
Is there an older engineer out there who can still handle a waterhammer analysis as described in John Parmakian 1963? I have tried to do the graphical solution starting with using the charts in paragraph 48, page 87-92. I calculated the two necessary parameter for the charts (2*rho; k*2L/a) but the second (k*2L/a) came up to 4.15 and is therefore way off the charts. I checked all the numbers and there is no error. It's just that the time for one wave passage through the pipe (L/a) happens to be already 3.1 sec. So I am not sure if I can extrapolate the charts for the up-and downsurge curves or what else I could do. I am wondering also why the max downsurge (either at pump or mid-pipe) in the charts reach only 100% of the normal head. That would mean that no negative pressure can occur, right? But I think that is exacty the case with my project. I suspect pressure below -14.3 PSI - meaning cavitation and column separation. But I need to quantify that (without buying modelling software or spending weeks to get familiar with the software). Anybody has some ideas?

 

[hydrae]

I do not have the book you reference but I have the following information on water hammer in long pipes.

For the Peak pressure:

The simple equation for water hammer which assumes a rigid pipe and instant closing valve is:
P= roh*Cs*V/Gc
P = pressure rise by the wave
roh = fluid density
Cs = velocity of sound in the fluid
V = Velocity

ref ASHREA fundamentals 1985 pg 34.2

This results in a change if pressure of 63 psi per 1 ft/sec of initial water velocity, if this change results in a static pressure below the vapor pressure of the water (around -10 psig or 5 psia for 70 deg water) then you will get collum separation. When you have collum separation the equation really changes. Hence the charts do not go above 100%. You will get will also get damage by the colapsing vapor bubbles.

This is worst case, the actual peak pressure value is reduced by thin wall pipes, friction losses, pump momentum and other things. A good number to use instead of 63 psi per ft/sec is about 55 for ductile iron, 50 for steel and plastic and 40 for thin wall plastic.

In discussion about the time function:

If your actual time is shorter than your calculated time then you may have a shorter pipe than what you think you have. Check for a break in the pipe 14,600 feet from the source (3.1 sec * 4720 ft/sec) a small break may allow air or other material it enter the pipe and act like the end of the pipe.

Also some surge tank manufacturers offer free surge analysis.