Pump Pulsation

[jvian]

Hello All,

I have a customer who wants me to evaluate the pressure pulsation of our new pump design and this I have never done in detail but told them I would attempt to do so and learn in the process. Only in one previous application have I been exposed to such a topic so hopefully someone here may be able to provide me with some assistance. I have a piezoelectric transducer and an oscilloscope and have collected a bit of data so far with the pump. My question is with respect to interpretation of the results, where I have collected time domain data with amplitude and peak-peak data as well as frequency domain spectra with rms data. What is of most value with respect to pump pressure pulsation when integrated into the overall system? My customer is interested in any impacts the pump may have on a system monitor downstream of the pump. My guess is that the peak-peak data is too high as the peaks are not constant and vary over a collection of many pulses. A persistence plot does show a common trend at the driving frequency as expected but the individual pulse peak-peak amplitude varies quite a bit and are on average lower than the absolute peak-peak measure in the persistence plot. On the other hand though the spectral plots of rms amplitude at a given frequency seem too low as there are clearly much higher pulses on the individual pulses.

In general, what is a typical customer asking for when asking for the pressure pulsation of a pump? Can anyone provide some insight as to how this data is normally reduced and presented or provide references to material which may help?

Thank you in advance,

- J -

 

[LittleInch]

I guess one of your many issues is in how you gather the data as the downstream system will affect the results so best to state what your test loop / assumptions are for size, type of pipe, flow or pressure control.

As a pipeline engineer I would be interested in mean pressure fluction around a set pressure (e.g at 20 bar, mean fluction is +/- 1.5 bar), or range of pressures combined with frequency, plus max peak to trough variations. If you have set frequencies you can extract and the mean peak to trough if they are different, e.g. 20 htz, +/- 1.5 bar, 45 htz +/- 0.75 bar). This will be important as an input into pipe vibration studies and also any DCS system or transmitter as the scan frequency needs to be double the pressure / flow variation frequency - see

http://www.driedger.ca/ce2_pdp/CE2_PDP.html

under measurement.

That's my best guess but others may have more info / requirements.



My motto: Learn something new every day

Also: There's usually a good reason why everyone does it that way

 

[Artisi]

It would probably help if you let everyone know what pump style , size and application you are talking about and what / why you are trying to collect this data.

It is a capital mistake to theorise before one has data. Insensibly one begins to twist facts to suit theories, instead of theories to suit facts. (Sherlock Holmes - A Scandal in Bohemia.)