harmonics are multiples of the basic (lowest vibration frequency).
modes are the different ways a system can oscillate... e.g. a rope held by the ends.
this is the 1st mode (2 nodes):
)
My take has been that you want to stay away from operating at a harmonic frequency (i.e RPM). Oscillations will continue to increase until system damage is done. Operation with a VFD can lead you into this problem. It is hard to find information from pump manufacturers on what RPMs to avoid.
I am not actually positive that I am right in all this, but could be some good food for thought, or lead you to the answer that you really want.
In their broadest context, I view harmonics as being higher orders of fundamental vibration sources or vibration responders. For vibration sources, higher harmonics may be either integer (eg. impeller blade passing) or non-integer (eg. fluid standing wave) multiples of the fundamental vibration frequency. For vibration responders they are mostly noninteger multiples of the fundamental frequency. For structural responders, the plate, beam, shell, etc. deflection modeshape complexities dictate the energy required to produce the shape of the particular harmonic frequency which sets the frequency separation between modal harmonics. For hydroacoustic and aeroacoustic responders in complex flow systems with restrictions , expansions, terminations,etc, it may be the relative spatial locations of these flowpath discontinuity features that determine the the spacing between fluid modal harmonic frequency spacings. Very little has been published about hydro/aero-acoustic resonances in complex flow systems so it is rather difficult to predict such frequencies except via fluid network analysis methods, which are mostly one-dimensional at the present time.
The terms you have mentioned are quite broad. Please provide some context for your question.
"Harmonics" means "multiples" (of a given frequency).
You can have an impact once per revolution which will give time domain impacts at once per revolution and frequency spectrum with components at 1x running speed, 2x running speed, 3x running speed etc... which are harmonics of running speed.
You could have a bearing defect which will give rise to harmonics of the bearing fault frequency.
Nothing in the term harmonics has anything to do with resonance. In general a machine resonance which is excited will not be accompanied by harmonics of that resonance in the vibration spectrum.
Modes - I think these might roughly describe the different shapes and associated frequencies that the system can vibrate at without excitation. (This term is related to resonance).
Any body has some natural frequency. When an external sources frequency is matched with it resonance occurs. This resonance will occur at other frequencies that are multiples of natural frequency and are called Harmonics.
But Harmonics due to VFDs are electrical harmonics resulted because of frequent swithcing of power supply.
Electricpete! the experiment we did during our studies (rising and lowering water column to detect harmonics) is based on resonance only.
I agree that resonant frequencies of a given system may be harmonically related. That will be the case for an air column above water in a tube I think Quark is referring to, and a vibrating beam or string that I think abel is referring to.
If I have given the impression that I disagree with that then I apologize.
My point was that the term "harmonic" has a much broader meaning and application which can be completely unrelated to resonance. In the case of rotating machinery monitoring, harmonics in a spectrum most often arise from repetitive impacts. The fundamental frequency is the impact frequency, not any resonant frequency. I have never heard of a case in rotating machine diagnostics where two different harmonically-related resonances were both important.
I don't know what the original poster had in mind... that's why I asked for clarification.
If you have rubbing contact with a rotating rotor it is possible to have "disconnected" arrays of rotational (1X) harmonics whose focal point is the predominant structural response frequency at the rub site. This is apt to occur with "soft" rubs such as a piece of rubber O-ring squeezed thru a narrow metal-walled gap like a motor magnetic gap. If there is metal to metal contact at such a narrow gap, the rotational harmonics again show predominant amplitudes
at the frequency(s)of the rubbing components but this "hard" rub case may have 1X harmonics running all the way to near zero frequency as well as to very high frequencies. I'm not sure whether rubbing events produce "true" 1X rotationally-ordered harmonics or just spread out from the contact response frequency in which case the rotationally-spaced peaks could actually be be rotationally-unordered if the contact response frequency is
not rotationally ordered.
