I was looking for literature on how to calculate resulatant forces from waterhammer traces and am not having much luck. Perhaps somebody could fill me in on the subject?

I am using a transient analysis software package which generates the pressure transients along a pipe at certain instances in time. The way I understand it is that you will obtain unbalanced pressures in in the bend pairs on the ends of a pipe. The unbalanced pressure being equal to the difference of the pressure transients at the two bends at any instance in time. If I multiply this unbalanced pressure by the pipe internal cross-sectional area I then get a resultant force. A force trace is produced and the maximum of the absolute value of this trace is used. This force is then multiplied by a dynamic load factor (typically a maximum value of 2 although I realise in certain instances it could be greater) or else input into the Caesar dynamic analysis and the dynamic load factors computed more accurately.

If not input into Caesar, would this then be the force a pipe support should be designed for - obviously ignoring momentum loads at this stage? I realise that this is perhaps maybe a conservative approach, but is perhaps the best way of identifying where critical sections of piping are, that might need a more indepth look using Caesar.

I have spoken to some people about this and they are of the opinion that the force trace if plotted, is positive and negative in time (owing to the changes in direction) and that the actual force is taken as the difference between the positive and negative peaks along this time trace and that the maximum difference is what should be used for a maximum force. Typically this would almost double the magnitude of the force I would have used - is this perhaps already accounting for the dynamic load factor, although this is still added on afterwards?

I am now confused as to which is the correct approach as the one yields loads almost double in magnitude. Perhaps someone could shed some light on this?

Regards
Dacre

Your water hammer calculations should yield a pressure and Q data vs time at given data points.

If you look at two data points which make up an elbow pair the unbalanced force is determined by subtracting the two pressures from each other multiplying the delta P amount by the cross sectional area of the pipe.

Finally that force would then have the momentum force added to it by considering the delta Q (Q=Va where V gets you momentum) vs time at the two elbow pts (usually this term is small and could probably be ignored in most cases)

DLF is really based upon the participation of the system and how it will react to the sudden load. Stiffly restrained systems have greater participations and may even exceed the standard DLF of 2.

If possible I would use your water hammer data to do a time history dynamic analysis of the system. You can find more about this in the on-line help of CAESAR II as well as in a COADE newsletter... I can't remeber which one so you'll have to look thru them.

Thanks John, that confirms my original thinking then.

Regards
Dacre