Harmonic Analysis on Pumps

Posted by: mul211

Harmonic Analysis on Pumps - 10/06/09 12:11 PM

Once the load and frequency are determined how far from the pump discharge nozzle should the piping be included in the Harmonic modeler? I'm assuming X load, say 50#, for X direction piping and 50# Z for Z direction piping.

Also how is the stress computed compared to a B31.3 stress allowable? I don't see a place in the load case combinations to do this. Is it a manual calc where you add Sustained and OCC dynamic?
Posted by: Farhad

Re: Harmonic Analysis on Pumps - 10/09/09 04:53 PM

Hi mul211,

Quote:
Once the load and frequency are determined how far from the pump discharge nozzle should the piping be included in the Harmonic modeler?

It depends on the magnitude of forces in pipe segments far from source or origin. Note that, low range shaking force may cause high cycle fatigue.

Quote:
Also how is the stress computed compared to a B31.3 stress allowable? I don't see a place in the load case combinations to do this. Is it a manual calc where you add Sustained and OCC dynamic?

See page 5-41 of Technical Reference Manual.

dynamic equation of motion: (K - Mw^2) A = Fo

This is exactly the same form of the equation as is solved for all linear (static) piping problems. The appealing thing about this is that the solution time for each excitation frequency takes only as long as a single static solution, and, when there is no phase relationship to the loading, the results give the maximum dynamic responses directly.
Posted by: Dave Diehl

Re: Harmonic Analysis on Pumps - 10/12/09 11:24 AM

If you include damping in your harmonic analysis (and, by default, it is), the maximum response may not be automatically displayed in the results. Damping essentially "delays" the system response and this delay is expressed by the phase angle of the response. C2 will display results for the phase angle that produces the maximum nodal displacement of the system.
Posted by: Dave Diehl

Re: Harmonic Analysis on Pumps - 10/12/09 11:33 AM

C2 can calculate an occasional stress or a fatigue stress for harmonic loads.

I would suggest a fatigue approach for harmonic loads. To calculate a general von Mises (or equivalent) stress for fatigue evaluation, specify a number of cycles in the harmonic input. You will then see a stress type of (FAT) rather than (OCC). If you entered a fatigue curve in your (static model) piping input, C2 will compare the calculated stress to stress amplitude associated with the cycles provided. You may, instead, use a fixed limit to this stress based on API 618 or ASME OM-3