Pump Guy Mailbag: A New Method for Cavitation Detection?

Hi Larry,

I attended your pump seminar last year in South Africa. As an electrical engineer, I really appreciated your ability to explain pumps in an easy manner. Your book has become my “go-to” resource on pumps.

I work for a manufacturer of electronic motor protection relays and controls. We hope to design and develop a relay using a Fast Fourier Transform (FFT) Spectrum Analyzer to detect cracked and broken rotor bars, as well as predict pump cavitation and bearing failure.

The only cable going to the motor that definitely stays connected is the three-phase supply cable. We are going to try to detect and isolate the vibrations associated with cavitation and bearing damage using the motor’s power cable.

I notice that all current motor relay designs purport to detect cavitation by associating cavitation to under-current or inadequate power. I would appreciate your insight.

Kindest regards,

Stuart L.

Hello Stuart,

Your plan is plausible if you can tie vibration analysis to the motor power cable. Everything vibrates. Pumps, motors, flowers and the passing breeze vibrate. The FFT spectrum analyzer reports most pump cavitation as a broadband vibration at 2 to 5 kHz.

I always thought that vibrations were transmitted to the analyzer through a contact or probe. I guess the motor’s hardwired power cable can be the vibration probe, but this is outside my area of expertise.

You are very observant. Most existing motor relay designs propose to detect cavitation by associating cavitation with a drop in power or amps to the motor. This is misleading. Motor load is a property or condition of the motor. Just because the electric motor is coupled to a pump, you cannot project motor load to a specific problem with the pump.

Using the fluctuating load (amps) on the motor to predict cavitation is like looking at the clouds in an effort to predict rain. It is true that it cannot rain if there are no clouds in the sky. But clouds don’t necessarily mean rain. There can be clouds without rain. The clouds only indicate moisture in the air.

Using motor amperage draw to predict cavitation is like using a quick rise in body temperature (fever) to predict appendicitis. It is true that you will develop a fever if your appendix is inflamed. But you can suffer a fever from other causes besides appendicitis. Your body temperature will also rise from food poisoning, prolonged exposure to the sun, a drug reaction, an allergy, diarrhea, cancer, Ebola, an ingrown toenail, a bee sting, and many other causes besides appendicitis. A fever only indicates that something is out of control in your body.

Regarding pumps, the load on the motor can fluctuate or drop due to any number of factors:

The pump’s specific location on the performance curve (duty head and flow) causes load fluctuations. And the pump is constantly moving on its curve, reacting to changes in the system.

  •  Pump Cavitation
  •  Internal recirculation
  •  Turbulence
  •  Fluctuating liquid specific gravity
  •  Fluctuating liquid viscosity
  •  Entrained air or gas
  •  Incorrect internal tolerances
  •  Slippage at the coupling between the motor shaft and the pump shaft
  •  Slippage at the impeller to the pump shaft.
  •  Dry-running the pump
  •  Pumping a hot liquid
  •  Pumping a liquid close to its vapor pressure

You can see from the list above “cavitation” is only one of many problems indicated by a drop in motor load. If you can marry the FFT technology with the motor relay, the world will beat a path to your door.


The Pump Guy