Sei sulla pagina 1di 3

Advanced Rotorbar Analysis False Positives:

Spiders / Axial Cooling Vents


Ernesto Wiedenbrug, Ph.D., SM IEEE,

Virtually all rotors of motors above 100kW are designed with axial
cooling vents, potentially causing false positives for both prevalent
diagnostic technologies, vibration and MCSA. Fig 1 shows the rotor
with 8 spider-legs of a 2.4MW 6.6kV motor that was pulled out of
service due to such a vibration and MCSA false positive.
This paper explains how to recognize problematic motors, the
reason for these false positives, how these motors react, and how
to deal with them.

Figure 1: Rotor with 8 spider legs

Spiders / Axial Cooling Vents:


What are Spiders / Axial Cooling Vents?
Starting around 100kW in size, induction motors rotors are
designed with axial cooling vents on the rotor. These vents serve
multiple purposes, aiding in cooling the rotor and also lowering
the rotors weight and inertia. These axial cooling vents can
either be manufactured by punching holes on the rotor
laminations (also called kidney-holes), or by mounting a thinner
ring composed of laminations and the squirrel cage onto legs
that are fixed onto the shaft (called spider). Fig 2 (upper) shows
two rotors of the same dimension, with a spider design to the
left and kidney holes to the right, with their rotor laminations
being depicted in fig. 2 (lower) [1].
Potentially Problematic Motor Designs
In 1992 Bill Thompson from Aberdeen, Scotland, was the first to
report the problem of motors being pulled for maintenance
under the suspicion of cage damage, only to encounter
completely healthy rotor cages [2]. He tracked the issue back to
motor designs that had the same number of poles and axial
cooling vents (for either spider or kidney hole designs).
Sangbin Lee, from the University of Korea, interviewed several
major motor manufacturers in this regard. They commented
Figure 2: 4-leeged spider and kidney-hole.
that such designs arent favored any longer because they are
known to cause higher vibration levels. Dr Lee conducted an exhaustive study of this issue [1] and reports that
nearly 37% of the motors used today have this potentially problematic combination of axial cooling ducts and
number of poles.

Symptoms:
Why do these designs show with false positive cage faults?
Lets revisit how a motor reacts if it has a broken bar by imagining that we are sitting on the rotor and see how the
magnetic field of the stator is slowly spinning around us. The poles move by a rotorbar (north south north
Copyright 2012

Wiedenbrug, LLC

south ). If the bar is healthy, the moving magnetic


field causes that bar to carry a current (plus
minus plus minus ). That current interacts
with the stator field and generates maximum
torque whenever the current through the bar is at
its peak. When a bar is broken it will not pull the
rotor forward at the times when it should be
carrying current, letting each magnetic pole slip by
easier. The effect is a slight ratchet-effect, where
the average torque dips every time a pole passes
the broken bar. Vibration, current and torque
Figure 3: Flux distribution depends on rotor position.
spectra used to diagnose cage damage are the
result of this ratchet-effect.
Now lets look at what happens with axial cooling
vents. Figure 3 shows the magnetic field of a 4-pole
motor with four kidney holes. The lines show how
the flux paths in the stator and rotor. The areas with
the highest amount of flux are orange, lesser flux is
yellow, then green, and areas that have no flux are
blue. Both pictures show the same instant for the
stator (the maximal flux is at 12, 3, 6 and 9 oclock).
The difference is that the magnetic poles on the left
Figure 4: False positive. Kidney-hole Design.
picture are facing the leg of the rotor, whereas
they are facing the axial cooling hole on the right picture. When facing a cooling hole, the magnetic pole can go
deeper into the rotor creating a better magnet and higher torque. When the pole faces a leg it allows the pole to
slip by easier.
The net result is the same type of a ratchet-effect in the torque for motors having the same number of poles and
axial cooling vents, as for motors with broken cages. Both cases will react with the same frequency components in
vibration, current and torque spectra. Fig. 4 shows the false positive caused by the kidney-hole motor of fig. 2.
Healthy Cage Load and Voltage Dependencies
The good news is that it is possible to identify problematic motors in the field. The more flux is pushed through the
legs and around the cooling hole, the bigger the ratchet-effect. This means that conditions that push the flux
deeper into the rotor will cause higher signatures.
Voltage Level: Higher voltage in the stator will create more stator flux, pushing deeper into the rotor. A few
percent higher of stator voltage will increase the amount of signatures. It is possible that motors show no brokenbar signatures on days where the buss voltage is a bit lower, and start showing signatures on days with higher
voltage levels.
Load Level: The easiest way to verify whether were dealing with a potential false positive may be to change the
load point. Lower load level cause higher signatures for
motors with healthy cages (due to the lower counterEMF). This effect is opposite to what happens with real
rotorbar damage, where low load points show a lesser
signature. But be careful: Motors with spider-signature
may also have broken bars!
Damaged Cage Location and Load Dependency
The bad news is that once knowing that it is a
problematic motor, it remains almost impossible to
diagnose using vibration and MCSA. We can try careful
trending, ensuring consistent voltage and load levels.
But even this is prone to fail because the same motor
Copyright 2012

Figure 5: Signature depends on location of bar and load.


Wiedenbrug, LLC

reacts differently, depending of the particular location of the broken bar with respect to the cooling duct. Fig 5
shows on the amplitude of the signature sideband (y-axis), as a function of the load (x-axis). Different colored lines
show how the signatures will vary, depending on what particular bar is broken. Imagine that we trend a motor
running at 70% load. If it breaks the blue, green or red bar, the signature would go up. If it broke the magenta bar
the signature would go down. But if it broke the cyan-colored bar, we wouldnt even be able to find the problem!
Methods other than Vibration and MCSA
Squirrel Cage motors with rotors that have the same number of axial cooling vents as poles can cause false
positives for cage damage for Vibration and also MCSA. It is also clear that not even careful trending can always
distinguish these false positives from true cage damage. This segment will discuss alternatives:
Trend Startup Times: Broken bars will deteriorate a motors ability to start. Trending of startup times can be used
to track the cage condition, however voltage level and load must be comparable for valid trending.
Single-Phase Rotor Test: The Single-Phase Rotor Test is probably
the more feasible solution to this date. It is a standstill test,
injecting sufficient AC voltage into a single phase of the motor to
magnetize the rotor at line-frequencies, and then slowly spin the
shaft while measuring the amperage on that phase. The concept is
simple. The induction motor is basically a transformer with the
squirrel cage being the shorted secondary winding. Single-phasing
the stator magnetizes two thirds of the primary. By slowly turning
the rotor, any broken bar will enter and exit the magnetized area,
changing the number of turns in the secondary winding. Fig. 6
shows this concept, which is well explained in [3].
Pulsating Field Test: The method proposed in [4] solves the main
disadvantages of the Single-Phase Rotor Test, while applying the
same concept thoroughly magnetizing the rotor at or near lineFigure 6: Single-Phase Rotor Test.
frequencies. But instead of slowly rotating the rotor, no proximity
to the motor is necessary. The test is applied from the MCC by magnetizing different sections of the stator with a
three-phase pulsating field with sufficient current and at an optimal frequency. This concept has received several
IEEE awards, but isnt available on the market yet.

Guidelines:
KNOW THY ROTOR!!!

We must know which motors have the potential for false positives due this issue.
Work with your motor repair shop and document the number of cooling ducts of your rotors.
Attempt trending, but comparing measurements at same load and voltage levels.
The only reliable method remains the cumbersome single phase rotation test. It needs a line-frequency voltage
source capable of pushing several Amps, and the ability to turn the shaft slowly while reading the current.

Further Reading
Advanced Rotorbar Analysis: False Negatives Dual Cage Rotors
Advanced Rotorbar Analysis: Cage Damage vs. Porosity

References
[1] Evaluation of the Influence of Rotor Axial Air Duct Design on Condition Monitoring of Induction Motors,
Sungho Lee, Jongman Hong, Sang Bin Lee, E. Wiedenbrug, M. Teska, Heedong Kim, IEEE ECCE 2012.
[2] On-line current monitoring the influence of mechanical loads or a unique rotor design on the diagnosis of
broken rotor bars in induction motors, W.T. Thomson, Proceedings of ICEM 1992.
[3] Squirrel Cage Rotor Testing, Tom Bishop, EASA Convention 2003.
[4] Automated Detection of Rotor Faults for Inverter-Fed Induction Machines Under Standstill Conditions,
Kwanghwan Lee, Jinkyu Yang, Sang Bin Lee, E. Wiedenbrug, M. Shah, IEEE IAS Transactions, Jan.-Feb 2011.

Copyright 2012

Wiedenbrug, LLC

Potrebbero piacerti anche