Water hammer

Posted by: M Blackman

Water hammer - 02/16/01 12:35 AM

We have a consultant's report on an ESD type valve in a 10" condensate export pipe which closes in the period of 10 seconds. They are predicting column separation, a pressure rise of around 2000kPa and the attendant huge water hammer forces.

I have read the caesar II manual section on water hammer. I am prepared to believe the rise in pressure, but have trouble with the hammer force prediction. The speed of sound in the fluid is, lets say, 1000m/s, the max. length between adjacent elbows is about 20m, thus the pressure wave passes between elbows in about 20ms. Because the valve closes so slowly, the pressure wave on the upstream side has such a long ramp-up in comparison to the 20ms that the net pressure difference between adjacent elbow pairs is always going to be very small, and so too the net force. As pointed out in the manual the closing time should be smaller than 2L/c for significant hammer. The only exception to this I can see is if the pressure wave in fact carries a large negative peak behind it which could cause vapourisation /column separation. Does anyone have comments?

On the downstream side of the valve the pressure variation is immediately negative as described in the manual so I would expect that vapourisation/column separation could occur and large forces could result. Are these forces more difficult to predict than those due to pressure rise ?

regards
Martin Blackman
Pipe stress engineer
Posted by: Dave Diehl

Re: Water hammer - 03/09/01 10:59 AM

Ten seconds is a slow closing time but keep in mind that the majority of the actual closure occurs in a small final percentage of that total time.

In that 2L/c calculation, L is the distance to a major reflection point such as the upstream source or major change in flow area. L is not the length of the run between an elbow-elbow pair.

On hammer analyses, start with the longest run between elbows first. These runs have the longest imbalance and the longer imbalances excite the lower modes of system vibration which should have the larger system participation in the event. The event in that 20 meter run is probably over too quick to get your system excited.

------------------
Posted by: Anonymous

Re: Water hammer - 04/17/01 05:02 PM

The use of Tau (2*Lc) is extremely simplified. Your 10 seconds valves hydraulic characteristic may actually in fact be extremely non-linear.

In other word consider a simple butterfly valve. No matter how slowly its closed 75% of the Q is controlled by the last 20% or less of closure.

Also the maximum surge wave pressure based on the joukowski equation can be tossed out the window if you experience column separation. These separations magnify greatly the effect of the first upsurge wave.

The bottom line is this... CAESAR II is a wonderful structural analysis tool, but unsteady hydraulics should be studied as a separate effort. Do'nt mix up structural and fluid mechanics.

Best of Luck,

------------------
Best Regards,

John C. Luf
Posted by: Anonymous

Re: Water hammer - 04/18/01 08:30 AM

After having slept on my previous response I would like to add the following added thought(s)

1) Rather than fighting the effects of water hammer by restraining the system mitigate the water hammer. Have your fluids consultant put together a system remediation plan.

2) Take a look at the local sources of information available to you on the subject of water hammer. COADES treatment is a first pass but certainly the more you look into it the more you will understand where and how things are occurring in a system.



------------------
Best Regards,

John C. Luf