Re: Re: Vibrations in a pipeline

On Dec 25, 2006, at 2:29 AM, tomcruz55 wrote:

> What ever software you got to analyze pipe stresses wont be of any
> value if youre working to reduce the level of vibration in a piping
> already in-service to an acceptable level without getting the
> vibration reading in several locations.

True as far as it goes. I should add that even vibrometer readings are less than no use if you can't interpret them in terms of the physical response of your system. Dynamic response and excitation are the two halves of the problem--the excitation provides the energy than the response is the objectionable result. Get rid of the excitation and the response goes away. Conversely you can change the response and the excitation won't matter. If you can't identify the excitation and connect it with a response, you're wasting your time because you'll end up fixing something that isn't a problem.

I ran into one once where an expansion joint liner disintegrated and buggered up a turbine. Everyone, including the expansion joint manufacturer looked at the layout and decided that the root cause was flutter caused by turbulence in the steam line downstream of an elbow. The basis for the decision was some rough design guidelines for specifying the liner thickness if the expansion joint was located near a fitting. But when you looked at the collected fragments all the cracks didn't track the stress profile you'd expect from any sort of dynamic response. In fact what had happened was that some sort of surge during the start-up had overextended the expansion joint and jammed the liner hard into a flange and subsequent stress corrosion, probably caused by contaminated feedwater, was the source of the corrosion which eventually broke up the liner.

The moral is that you need to understand both the excitation and the response--you can just measure the vibration amplitude and assume you know that the amplitude is the problem. If you don't know what's causeing the problem, you can only guess at the fix. When you get the amplitude reading get the frequency. Then compare the measured frequency of the span to to its natural frequency. If they match you know that one of your problems is a soft span. Then figure where the excitation is coming from--wind, groundborne vibrations, or flow pulsation, to name several. Then get those fixed. If the span frequency doesn't match the measured frequency, the cause isn't the soft span, so you need to start exploring things that provide energy at the measured frequency.

Christopher Wright P.E. |"They couldn't hit an elephant at chrisw@skypoint.com | this distance" (last words of Gen.

.......................................| John Sedgwick, Spotsylvania