Hi Guys,

I have a boiler feed pump that has a very speed specific 3x VPF peak of which approaches 5.5 G's. This peak is out at around 1595 HZ.

I use the medium sized PCB Modal hammer that I think everyone who takes vibe data has (sorry, I don't know the model number off hand) which comes with the 4 tips.... brown, grey, red, and black (Black being the hardest).

This peak certainly behaves like resonance.... very directional and speed specific. I have even considered the possibility of acoustic resonance.

I have attached a force spectrum from this hammer using the black hammer tip..... please take a look... I have attached the same spectrum in both log and linear scale.

I don't believe I can excite that frequency range with the tip that I have (based on the spectrum I attached).

I have a few questions:

- I know there is the 10 or 20 dB drop off practice.... my software shows LBS Force RMS... what is the drop off rule for this?

- Has anyone had to excite a frequency range anywhere around 1595 Hz? What did you use for excitation?

- Does the amount of force input have anything to do with the frequency range/response? Meaning could I put more force into the impact and possibly get out to the required range?

- What would you suggest for this case?

Thanks for your help and please let me know if you have any questions!

Kind Regards,




Original Post

An acoustic natural frequency would be sensitive to fluid temperature.  You could run a test with vibration response; better would be to include dynamic pressure transducers.

What is vibrating?  Surely, the entire pump cannot be responding at this high frequency. 

Did you notice the vibration on permanently install probes?  Check for the simple first, like bracket resonance.  

If this is seen with a magnet mounted transducer, one may also investigate mounting resonance at this frequency.  


Hi Bill,

I am actually gearing up for this job and will be bringing dynamic PT's with me to check/verify the presence of any acoustic modes

This peak is present at the POV position and is dominant there only.

This pump does not have permanent accelerometers so we have installed temporary wireless vibration transmitters. We have gone behind them with a portable analyzer to verify the peaks presence however.



I would use the 1-lb. PCB hammer with hard nylon tip or probably the metal tip. If you only have the 3-lb. mini-sledge hammer, then the black plastic tip is probably not good for your test. The Coherence function can indicate whether the signal from hammer is adequate. You could try making a metal impact tip by finding a bolt (hex or carriage head) that fits the screw threads. Some contour grinding may be needed for a slightly rounded impact surface. I have done this with the smaller hammer with success. Be careful with impact force to prevent overload of the load cell. Use at least 2000 Hz F-max to measure 1500-Hz.

Check bearing housing oil pipes (if present) and cooling plan pipes for resonance.



I do not answers (sorry). I have fundamental questions.

What symptoms make you sure the vibration is related to VPF?

Because this is a boiler feedwater pump, I would assume it is mounted on journal bearings and its speed is >5000 rpm. Right?

The plots you posted are for the force (input). To measure the natural frequency of the structure, don't we look at the vibration (output) or the transfer function (output over input)? 

Regards- Ali M. Al-Shurafa


The vibration in question is a 3rd multiple of vane pass.

You are correct on the bearing arrangement and it runs at 5100 RPM's and lower.

My question was related to how high of a frequency would I be able to excite with a modal hammer. With the force spectrum I have provided, you can see the frequency range I am currently able to excite. Since the vibration in question is above 1500 Hz, I will not be able to excite it with the current hammer tip I am using (need a different tip to be able to go higher in frequency).



So, if I rephrase your posts perhaps I would put it this way:

I measure a high amplitude (value is unknown) at 1590 Hz from the bearing housing at pump outboard vertical position. This frequency coincides with 3X vane pass frequency of one of the impellers inside the pump. Other vibrations do not show high amplitudes at this frequency. I suspect the bearing housing has a resonance condition excited by this 3xVPF and I'm trying to confirm this possibility by doing an instrumented impact test on the bearing housing while the pump is running.


Regards- Ali M. Al-Shurafa 

Add Reply

Link copied to your clipboard.