Dear Farhad,

If I have to do a Harmonic analysis,my opinion is it should be 225N as that is the mean of the range ie the amplitude.The curve for y= sinx has a range 2 and amplitude 1.A Force=Fsinwt should be worked with amplitiude and not range.

The govening eqn. for Harmonic analysis is:

MX( DOUBLE DOT) +KX +CX(DOT)=Fo sinwt where Fo is the amplitude and not range.450 is the range and 225 is the amplitude.

As I have mentioned before that I am sceptical about doing Harmonic analysis when the frequency content is broadband. However, based on Fourier analysis theory if the exciting frequencies are low, a harmonic analysis can be justified.

The way I would approach a problem if I get the time history of loading is like this ( typically I get it this from our hydraulics group):

1) Use the force vs time data at each ( significant) change in direction to generate the DLF spectrum and use a response spectrum analysis ( or find the max. pressure differential b/w two elbow -elbow pairs and do a static analysis with DLF 2)

2) Do a frequency content check and avoid harmful frequencies from the structural freqencies of the piping system.


It is best to leave the analysis and supporting of compressor piping systems ( upto significant distance)to the vendor as the complexities are too high.It has wlays to be considered that resonance will happen ( there is an engineering dynamics corporation paper on this, the subject matter of the paper is which philosophy of API618 should one adopt).Coupled acoustic-structural analysis which basically is an Eulearian-Lagrangian approach is beyond capacity of most stress engineers ( at least me) and I would prefer not to venture into areas which are outside my area of expertise.

Regards