In the design of nuclear power plants many of the piping systems require protection against seismic events as dictated by ASME Section III.

In the early days, this seismic protection was achieved by installing hydraulic snubbers. There has been a shift in the industry away from hydraulic snubbers in favour of purely mechanical snubbers since these do not have any seals to rupture or fluid to leak.

My question is: "can a suitably sized Grinnell sway suppressor (Grinnell figure 301) be used instead of a mechanical snubber?". Imagine the followeing scenario wherein a seismic analysis is performed and CAESAR II reported a snubber lock-up restraint load of say 1500#. Would it be appropriate to install a Grinnell sway suppressor with a preload of 1500# acting at the pipe seismic support location? With the Grinnell sway suppressor the 1500# of restraint force would act the instant the pipe wanted to move away from the installed preloaded position. In effect, the sway suppressor is applying a 1500# restraining force against seismic movement.

I would like comment from this forum's readers on this subject.

Regards to all Patrick LaPointe P. E.

Hello Mr. LaPointe,

This could be an interesting discussion; I hope our learned colleagues will join in. Perhaps it would be good to begin with what the intended use is for snubbers and for sway suppressors.

The snubber is intended to influence the piping system only under loadings consistent with rapid acceleration. Earthquake is one such loading. The snubber is designed to not constrain the movement of the system when loadings of the "earthquake type" are not acting upon the piping system. The snubber will allow the slow thermal movement of the piping with VERY LITTLE RESISTANCE (thermal expansion only causes stresses when you resist the movement). This is supposed to be true of hydraulic snubbers or mechanical snubbers (e.g., Pacific Scientific inertial type). Only if the pipe accelerates quickly will the snubber lock-up. When it locks, it transfers the loading of the piping mass through the snubber to the attachment point structure (hope the snubber and the point of attachment are strong enough to accommodate the dynamic load). The snubber can react very quickly before the piping has moved very far. Quick reaction and minimum movement is what you want to control the piping under seismic loading. It is important to stop the movement quickly – before it gets to top speed and the loads are unmanageable. There is NO INITIAL PRESET on the snubber - it will not load the piping (except for its considerable weight, which many analysts seem to forget about). Snubbers can be selected to allow the full thermal movement of the piping without resisting the movement.

Sway suppressors are useful for changing the harmonics (natural frequency) of a piping system and controlling wind and vibratory (e.g. “white noise&#8221 movements. They are especially useful when there are long unsupported (note, not "undersupported") spans that can be exposed to gusting wind. They are also sometimes used to control “flow induced” sway in piping caused by flow variations associated with control valves. The spring loaded sway suppressor WILL preload the piping and depending on whether the movement of the pipe compresses or expands the associated spring, the load (i.e., resistance to movement) will increase or decrease (in the case of a double spring unit the resistance to movement increases as the pipe moves in either direction). When the “non-thermal load” is removed the pipe returns to its original location and the spring loadings are gone (or in balance, except, of course for intentional preset loads). But while the pipe is deflected, the loading of the spring(s) on the pipe is present.

So do we want to use sway suppressors for the control of seismic loadings? We really do not want to resist thermal movement any more than necessary – the sway suppressor should not be installed such that it resists thermal movement no matter what load it is intended to control. Given enough pipe movement, the sway suppressor's spring will eventually become completely collapsed (more likely, it will happen in a heart-beat with a seismic load) and that is as far as the pipe can move (in that direction – assuming the suppressor does not break). BUT, it will let the pipe move (with the application of an increasing spring load) until it collapses the spring. This is not what you want for controlling seismic excitation. Also, consider the dynamic load the piping can apply to (and through) the suppressor. It might not be intended (by the manufacturer) to take that much dynamic load. I don't think you want to try to use them for dynamic events, especially when large displacements are possible (E.G., SRV blow-down, earthquake, and other impact type loadings).

There is much more to add to this and we hope you lurkers will contribute. What do you say???

Regards, John.

Hello,

Here is what my friend Ron had to say on the subject:

I do have a comment on sway braces, both with springs and without (which makes them struts). Either can can act as a seismic support -- the strut, obviously, and the spring type sway brace also, as it changes the piping frequency, i.e., stiffens the piping system. Normally it would be adjusted to provide no load in the hot position and as soon as the pipe tries to move (in either direction) it will provide a spring force of some initial magnitude increasing in load to it's design limit (e.g., 3000 lb, increasing at 1000 lb/in to a maximum of 7000 lb for a sway brace I recently encountered). I don't see any reason it wouldn't limit seismic movements.

Ron Haupt

[ November 27, 2001: Message edited by: John Breen ]