how can we identify the cracks of t 2

[brij]

how can we identify the cracks of the rotor while running of the any motor.

 

[electricpete]

The following may be symptoms:

oscillation of the current meter

pulsating vibration which may tend to get worse over the first few hours after start

pole-pass frequency sidebands around running speed and twice line frequency in vibration spectrum

pole pass sidebands around line frequency in the current spectrum.

slip higher than expected (speed slightly lower than expected).

Try searching for "rotor bar" and "RBPF"

 

[purplepete]

If the motor can be taken off-line I'm told to energise one phase (eg R-W) with an ammeter in the circuit. Maybe at a lower voltage (so it doesn't bolt). An ammeter must be in the circuit. Rotate the motor shaft SLOWLY by hand and watch the ammeter. If the ammeter changes significantly at the same pont on the shaft position as it rotates then you may have a broken rotor bar. Any other offerings?

 

[electricpete]

I agree with the "single-phase" test that purplepete mentioned.

If the motor can be taken off-line there are additional options:

- visual inspection.
- ultasonic and dye penetrant inspection of the rotor bar braze areas.
- rotate the motor and hold a flexible wooden stick against the bars, rotate the rotor and listen for a different sounding clicks as the bad bar passes by.
- perform a core currentloop test of the rotor and monitor by infrared. Loose bars should show up as hot spots at the braze joint and colder along rest of the bar. This is the subject of a previous question I have posted on this board but apparently works.
- PDMA rotor influence check. Similar test procedure to single-phase check but different principle of detection.

 

[bosha]

Can this be done using 120 vac,or is it better to use 277 across two legs,I've never heard of this before,but sounds like a pretty handy trick.

 

[electricpete]

EPRI suggests that you shoot for approx 10% of rated line-to-line voltage applied to one phase or approx 8% of rated FLA.

A variable voltage supply is ideal, but of not always available.

If it's a 460v motor and 120v is all that is available then you can probably get by with that. Try to do the test quickly to avoid possible overheating. If motor RTD's are available you may monitor those as well.

Generally less than 5% variation in current is expected for a healthy rotor.

 

[electricpete]

When I said to do the test quickly I hope you don't misunderstand. Get everything set up and practice rotating the rotor BEFORE you apply the voltage.

Then turn at 1/8-1/4rpm while monitoring current.

 

[bosha]

Thanks Exellent site

 

[jbartos]

Suggestion: Using the induction motor electrical equivalent diagram, the rotor current will be decreasing with number of cage cracks increasing. It will decrease to zero, when all bars crack. This means that the stator current will be decreasing and also the speed will be decreasing; especially, when the shaft is loaded with rated load. Notice that
Pmech=(1-s) x Rr' x |Ir'|**2 x m/s
where
Pmech is mechanical power
s is slip
Rr' is rotor resistance transferred to the stator side
Ir' is rotor current transferred to the stator side
m is m-phase stator

 

[electricpete]

jbartos - when all bars have cracked, you don't have to worry about finding it.... it has found you.

I don't think that decreasing current is a noticeable symptom of rotor bar problems. Decreasing speed is. Current will oscillate.

In general increasing rotor resistance will have only a second-order effect on current, but it will have a direct effect on slip. Pmech=(1-s)/s * R *I^2. As R increases s increases roughly in proportion for relatively constant P.

 

[jbartos]

Suggestion to the previous posting marked ///\\I don't think that decreasing current is a noticeable symptom of rotor bar problems. Decreasing speed is.
///The current at a question is the magnitude of current and rms. The oscillating current approach is more dynamic approach to the analysis. However, one cannot separate the speed decrease (or slip increase) from rotor current decrease since the slip increase decreases the rotor current (Irotor,rms). The various aspects identifying the motor rotor cage cracks are acknowledged. However, there appear to be missing specific values for a specific motor, e.g. motor size. So that the postings are qualitatively good; however, quantitative data as examples are missing.\\ Current will oscillate.

 

[electricpete]

jbartos - Pick any general purpose (Nema design B) squirrel cage induction motor that you want. You will be missing the boat if you are looking for current decrease as a symptom of rotor bar problems.

Your math proof is flawed. Increasing R does not directly result in decreasing I. Increasing R results in an directly proportional increase in slip s. If R doubles at constant power than slip doubles. Pmech=(1-s)/s * R2 *I2^2 ~ R2/s * I2^2 for s<<1. For Pmech constant as per your assumption the changes in R2 and s will cancel out and I2 will not change.

As you are aware the decrease in speed will result in some decrease in Pmech due to load characteristics but this is a 2nd order effect.

There simply is no diagnostic value in looking at current magnitude as an indicator of rotor bar problem. How do you know the mechanical load? You have to look at mechanical parameters and consider measurement errors on those parameters and pump efficiency along with motor parameters... simply not practical or useful way to monitor for rotor bar problems. Comparing slip to various indicators of load (including current) also requires careful consideration but is much more effective since slip is varying directly with rotor resistance for constant load.

I have read many articles on rotor bar detection. I have never heard anyone suggest monitoring rotor current. It is an indefensible proposal and it only adds confusion to the discsussion.

 

[jbartos]

Suggestion: The motor rotor current Ir' or Ir are mentioned for initial understanding of the problem. Obviously, the stator current Istatorvector=Imagnetizingvector+Ir'vector.
Actually, Pete's very first posting and very first symptom indicates &quot;oscillations of the current meter.&quot; It seems to sound very strange in the previous posting to separate or decline the Irmsstator=Im+Ir' from oscillation of meter indicated in the very first Pete's posting.
It is evident if all bars crack, Rr=infinity (theoretically) and Irmsstator=Im only.
One does not need any expensive EPRI literature ($25000 per membership) to observe this phenomena.

 

[jbartos]

Small encore: It appears that the visual inspection, as mentioned in Pete's posting #1 item, is most revealing since the crack in a bar must have produced a significant amount of heat that alters the bar and surrounding surfaces color.
Any incipient cracks have to be identified by usual methods searching for defects/flaws in materials, e.g. in propellers, turbines, etc. (xrays, infrared, ultrasound, etc.)