We have a 800 hp 1800 rpm 4kv vertical motor built in early 1980’s by Continental Electric and has been in service (except for plant outages) since 1988. It drives a centrifugal pump and the motor is not subject to frequent starts (probably averages 3-4 starts per year).

In it’s lifetime, it has had 3 current signatures recorded in 1998, 2002, and 2007 (last one attached). All three had the same result: the pole pass sidebands are about 44 – 45 dB below the fundamental 60hz. The five sister motors all have pole pass sidebands in the neighborhood of 60dB below fundamental.

I’m not too concerned, since the condition is stable (should I be?). Also the function of the motor within the plant is not critical. I checked speed with strobe and this motor is not running at a detectably lower speed than its sisters. Also no rotor position oscillation evident by strobe.

The motor will be removed in March this year for refurbishment associated with an unrelated condition (we have drawn oil from the upper bearing through the motor, and we want to steam-clean and bake the stator and of course wipe down the standpipe to stop further leakage).

It will be a very quick turnaround in the repair shop. The motor has to be shipped (locally), refurbished and back on-site in a week.

I’d like to take the opportunity to further evaluate the rotor condition which may serve as the basis to decide if we need further long-term actions (we won’t have time for any major repairs during this refurbishment) , but I’m also mindful not to create a huge scope of unnecessary testing that will slow the project down without giving a lot of meaningful information.

For sure I will ask for a rotor visual inspection and T.I.R and to record whether this is fabricated copper or cast aluminum rotor (I’m pretty sure it’s fab copper).

What other tests would you recommend?
Single phase test before motor disassembly?
Growler test?
NDT tests (dye penetrant, mag particle,UT)?
Rotor core loop test with infrared? Green paper?
Rotor current test (current injected through endrings) with infrared?
Other?

This message has been edited. Last edited by: electricpete,

LPHD21.ppt (87 Kb, 54 downloads)

No responses? I guess I made the question too difficult with too much irrelevant background information.

Let's forget about having a short availability of the motor.

The question is: what would be a reasonable shop tests to evaluate the condition of a rotor which tests marginal (45db) during on-line tests?

I'm thinking:
Single-phase test
rotor core loop test with infrared
visual inspect.


Does that sound like a good plan or would you add any other tests? Thanks.

I have often detected the defective bars using a growler as well as by current injection + infrared. However, sometimes a growler shows more than half the bars to be defective, so cannot always rely on it.

What is the point of the single-phase test, as the issue is to identify the defect location, isn't it?

With atleast two cases (these were 735 KW motors), all of the above & ultrasonic & dye penetrant did not indicate any flaw. We then had the repair shop de-braze the bars from the short-circuiting rings. It was then seen that many bars had just been inserted in the rings with some superfluous brazing on the outside. However, there was no material contact inside and the bars were effectively open.

Baker has a version of a surge tester (RT-1)that is designed for rotor bar testing. I found it too expensive to buy but some repair shops might have it over there.

Regards,

Aditya


P. S. What is a green paper test?

Thanks for your response.

You're right. The single phase test probably is not logical in this situation considering the motor would be disassembled anyway. Better to take advantage and get what data we can while disassembled.

So then I have
rotor core loop test with infrared
visual inspect

Maybe I should add growler? Any other comments? (if not I guess I'll ask the shop for their recommendations/preferences)

Baker off-line rotor bar tester... I have never heard of that. I don't see it here:
http://www.bakerinst.com/BakerWeb/Products/Products_Offline_Online.html


Green paper is flux paper. You can put it over portions of the core during the loop test (if not too hot) to show the flux pattern. Like this:
http://www.kjmagnetics.com/products.asp?cat=154

You can see the Baker instrument at http://www.baker-instrument.de/rotore.htm.

Reliance Motors, USA has a standard practice of injecting DC current between the two short-circuiting rings & viewng with an IR camera, better than a rotor flux loop.

Does this green paper work? They are mentioning it for permanent magnets only.

Regards,

Aditya

Thanks for the info on the Baker tester. Looks interesting.

On page 2 below you will see application of green magnetic flux paper to detecting faults in an induction motor.
http://www.lexseco.com/Files/rotorarticle.pdf

Our repair shop sent us a similar image once for a cast aluminum rotor.

quote:

"Reliance Motors, USA has a standard practice of injecting DC current between the two short-circuiting rings & viewng with an IR camera, better than a rotor flux loop."

Good point. I have seen that test done also: Instead of injecting current in wires looped around the core, you inject current into clamps placed on the end rings.

I hadn't given it much thought before, but now that you mention it, it makes sense that this test would be more sensitive to rotor bar problems than the loop-type core test. The reason is that it seems you will get a lot of bar current with direct injection. With the loop-type core test, I think the majority of flux is deep below the bars and doesn't induce much current in the bars.

I see this direct injection of current into endrings is the test they describe in the Lexseco link:

quote:

Lexseco:
"Squirrel cage rotor testing is a two part process. First we'll excite the cage winding by directly applying voltage through the shaft
clamps. If all is well each path or circuit should be equally excited. We use a magnetic paper to make this determination. Many
of you use iron filings, which is fine. The magnetic paper is just a little neater. Placing the paper on top of the rotor produces
a clear representation of the rotor circuit while excited by the Core Loss Tester's voltage. Our results with this method have been
excellent. Failed cage windings have consistently evidenced a void, gap, in the particle alignment of the paper.
Sometimes, the second phase of this qualification also helps. We'll warm the rotor by increasing the applied voltage. This tends
to demonstrate hot spots indicative of problems. Remember, since you're not attempting to determine the condition of the laminations, you need to look for uneven heating of the bars. We find the majority of problems occur at the connection of the bar and the endring. Depending on the specific nature of the problem, the fault area may also spark as the voltage attempts to cross the gap"

This message has been edited. Last edited by: electricpete,

EP,
A picture is worth a thousand words. I am still a firm believer that the data we can collect while the motor is more than 75% loaded is the best to evaluate the rotor, but while apart, you can't beat current induction and infrared( see attached pic!)

Pole Pass sidebands in the range of 42-48 db indicate rotor bar crack may be developing or possible problems with high resistance joints. Best is to monitor and see if the value is going below than 44 db you see now. We use PdMA MCEmax, besides looking at the pole pass sideband we also check for swirl effect in the current spectrum at 300Hz, or 5th harmonic. The presence of swirl effect is a further indication of broken rotor bars. As you are planning a shutdown on this motor in March, I would suggest having someone with PdMA MCEmax to carry out Rotor Influence Check (RIC), which will provide a graphical representation of the magnetic coupling between the rotor/stator. Analysis of the waveform would help in confirming if there are broken rotor bars.

Download Article of Rotor Bar problems:

http://www.pdma.com/RotorFaultZone.pdf

http://www.pdma.com/PDF/CS0402.pdf

Thanks for the comments Aditya, Ron and UETS.

I decided to go with growler, visual inspecton, and induced current test with thermography cold and hot.

We should have some results in April.

[QUOTE] I would suggest having someone with PdMA MCEmax to carry out Rotor Influence Check (RIC)
[QUOTE]
It was interesting to see the results of the RIC test on a motor with 22 bad bars out of 51 from the PdMA site. As the PdMA states, there is some kind of anomaly on the RIC test.
I would like to ask you your opinion: In case that you see the results of the RIC test only on a motor similar to the one from the PdMA site, would you make a call?
I can see 3 waves that are virtually exact sine waves with total harmonic distortion probably lower than 5%. It seems to me to be a pretty poor resolution considering that 43% bars are broken or cracked. One has to think what would the RIC test look like if only 2 or 3 bars were broken?
Maybe I am missing some important feature of the RIC test because I do not use it. But if the departure from the perfect sine wave variation of the inductance reflects the gravity of the rotor bar problem, I would not bother to perform the RIC test at all.
jank

Jank, you seem to have a lot of negative to say about the PDMA test equipment (different threads), especially for someone who self-admittedly, has little to no experience with it. I personally have used this system with great success… as have many maintenance dept’s of some of the biggest names in Industry… maybe they’re on to something

As for EP's original question... your finial decisions were great choices. Too often rotor defects are missed at the repair shop level only because it was never inspected. Many people in industry have the opinion the "Rotor's Never Fail" or "In my 25 yrs I've never seen one fail"... because of this many people have become lax in inspection.

The Growler or green sheet (same principal), should find any breaks. The IR will give a great second opinion. However, If there is any amount of damage present, the shop personnel will probably see it during the visual inspection.

As for UETS's suggestion to do a RIC test... also a great test, but in my opinion if it is in the shop and dismantled anyway, better to go with the growler, etc. The RIC however earns its merit in that it is able to be preformed in the field, WITHOUT DISASSEMBLY. Meaning that with you would have confirmation of a possible rotor issue using 2 separate technologies (C.S.A & the RIC)... This would give you the confidence in making your decision to pull the motor for tear-down/inspection.

Vibration data would also help as it would give 3rd opinion.

[QUOTE]Originally posted by Druncle:
Jank, you seem to have a lot of negative to say about the PDMA test equipment (different threads), especially for someone who self-admittedly, has little to no experience with it. I personally have used this system with great success… as have many maintenance dept’s QUOTE]

I have actually worked with PdMA for few months and I have done the RIC test. I must have tested at least 100 motors including some DC ones. I have been interested in this technology since something like 1996. So when I said that I do not use RIC test, it does not automatically mean that I don’t know what I am talking about.
When you go back to those threads where I am critical of PdMA, I have never criticized their equipment (as you have suggested), but always talk about a very specific aspect of their testing.
But let’s rather go to the technical aspect of the question: In case of the RIC test the explanations have been modified over time (no problem with that, we all change our minds). But one does not to have to have a ton of experience to know, that the residual magnetism does not induce anything into the stator at the standstill. Read in the PdMA Fault Zone Analysis “AIR GAP”: “Evaluate the influence of the rotor’s residual magnetic field on the stator’s phase-to-phase inductance as the rotor is positioned”.
It is statements like the one above that raise the suspicion about the rest. The RIC test is a very plain single phase test. There is no need to talk about residual magnetism.
jank

This message has been edited. Last edited by: jank,

quote:

Origionally posted by Jank: Maybe I am missing some important feature of the RIC test because I do not use it. But if the departure from the perfect sine wave variation of the inductance reflects the gravity of the rotor bar problem, I would not bother to perform the RIC test at all.

Jank, not to turn this into a pissing match, but if you do have access to an MCEmax, I would suggest refering to pages 10-20 of the "Data Interpretation Guide." This is the section which outlines and expalins the RIC test (Theory behind the test, results etc).

Your assumption that anything aside from a perfect sinewave means an issue with the rotor is clearly debunked ... In fact in manufactured, copper bar rotors (as found in most larger/higher quality motors) are not likely to exhibit any sort of sinewave when the 3ph inductance vaules (measured in 18 increments over 1 pole face) are plotted. Most copper bar rotors are infact "Low Influence," meaning that the position of the rotor has little or no influence on the stators inductance.

The following table is from pg 13:

condition waveform


Normal: Smooth 3-phase sinusoidal waveforms or non-sinusoidal waveforms in Low-Influence Rotors

Rotor Defect: Erratic inductance throughout the peaks of the waveform or the development of sinusoidal activity on a Low Influence Rotor

Air Gap Eccentricity: Inconsistent variations in the amplitude of the waveforms. Static eccentricity sometimes causes a consistent separation in the three sinewaves, coupled with a low inductive imbalance.


I have easily tested >>>thousand motors using the PDMA system (not to mention other Brands) and have seen time and time again just how successful the RIC test can be. I will however, be the 1st to admit that it has some short comings:

1) Most effective when a baseline was taken (as with most technologies)
2)Difficult to preform in a plant setting when coupled to the load as the shaft needs to be turned. However, as this test is intended mostly as a follow-up/2nd opinion to other tests that the MCEmax offers, which may question the rotor/air gap/stator condition (high inductance imbalance reading in the Standard test or high pole-pass sidebands), I personnaly feel that decoupling is much easier then sending a motor to a shop for teardown.
3) In Cast rotors, casting voids may affect the shape (smoothness) on the sinewave... however, if you have baseline data as stated in #1, side-by-side comparions will allow you to see any changes from the healthy condition

If anyone has more interest, I can provide case studies of just how effective the RIC can be.

Druncle,

This is getting to sound way too much like a PdMA pitch. I am sure you have tons of data with MCEMax.

I too have data on a few thousand motors with Framatome & BJM products (& Baker & Tettex & others) and can make a pitch such that any of them sound like the best thing since sliced bread. Plenty of great case studies.

The more important point though is that none of them can be trusted blindly. With every instrument that I've come across, there have been many false results till I figured out the pitfalls. And no manuals give those. I have met many PdMA users whose instruments are gathering dust right now.

So, RIC may be good but nothing out of the world for sure.

Regards,

Aditya

Aditya,

I totally agree with you that none can be trusted blindly and as i have said in other posts, I have used many other instruments (Baker, BJM etc) and I like many of the features of each. These are all great products and the choice of one in perticular usually come down to what exact goals the end user has for his testing.

Anyone who swears that any of these Brands are the "Be-all, end-all" is sadly mistaken.

However, what I don't agree with is someone making negative posts about a perticular test or product without first doing a proper investagation... which, according to the staements of ______, they have not. That would be compareable to me saying, "You should never do surge testing b/c it can be destructive," when a surge test, when used properly is invaluable.... its just meant for new or reconditioned (clean) motors.

I have read many of your posts and can easily see that you have a personnal preference, however, these same posts show you have a great understanding of the topics... I'd be willing to bet that someone such as yourself could have many success stories with either of these testers (even your least favorite).

We all need to remember that Running down anything based on a personnel bias is not productive to anyone ....

Druncle,

I have no personal issues with you & apologize if I came across that way.

Jank has taken apart many manufacturers on this board, I don't think he is personally against any one. He does make us question ourselves & not accept things at face value and his contributions on this board are most valued. Guess I went off the handle a bit to see him being run down.

Regards,

Aditya

I agree with Aditya. Certainly anyone should be encouraged to express their opinions and the reasons behind them. Then we can all question and judge for ourselves which opinions to believe and which to not. All of jan's comments on this particular thread made good sense to me. The case study appeared to show RIC missing a call, and the presence of residual magnetism seems irrelevant to the NORMAL functioning of the RIC test to detect defective rotor bars (residual magnetism is only relevant as a possible source of interference in the test, as far as I know).

Also I was interested to hear Druncle's comments. I would be interested to see your case histories on RIC.

This message has been edited. Last edited by: electricpete,

Aditya, no apoligies needed... no offence was taken. I certainly enjoy reading your postings (as well as those of Electricpete & Jank). I should apoligize if my comments seemed to be attacking. We have all seen previously how quickly comments get out of hand when we become unprofessional.

I will galdly post a couple case studies shortly,(away on assignment for a few more days)

In the meantime, the following link is to an article on the PDMA website which gives a better explaination of the theory behind their RIC test(http://www.pdma.com/Rotortest.html)

Not that we should take any manufacturer at face value without a question, but I believe some earlier comments regarding the RIC test("Somehow it seems to me, that performing a test that is not based on sound science is not an efficient use of time." in a different thread ) may be based on a misunderstanding of the test itself.

Not only should we not accept without question, but we should also remember to not reject an idea without giving it a chance.

Electricpete, I'm Looking forward to seeing your finial results... please keep us posted.

I would like to look into little more detail to explain my skepticism for the RIC test. Enclosed is a curve of reactance of the motor measured on all 3 phases of a 60 Hp motor, 480 V, aluminum cast rotor, closed rotor slots. The reactance was measured with variable voltage and I measured from as low as 0.118 Volts (phase to phase) to almost 30 Volts. Note that the curve for each phase is quite different. If measured for example with 5 Volts, there would be a huge inductive imbalance. Note the growth of the impedance as the voltage is lowered, then there is a maximum and the impedance goes down again.
The increase of the impedance is a very well known phenomenon and is pointed out in the standard IEEE 112. The growth appears on motors with closed slots on the rotor. As the voltage is lowered the flux that crosses the airgap doesn’t bother to go around the bars of the rotor, but simply flows over the surface of the rotor. There is enough iron above the closed slots to support this flux. However if the voltage is large, particularly close to the nameplate voltage, the narrow bridges over the bars will saturate and the impedance will go down. You can see the decrease as the voltage approaches 30 Volts. Note also, that the 3 lines for each phase became one. In other words, the inductive imbalance disappears.
The decrease of the impedance, as the voltage approaches zero, is caused by decreasing the permeability from the maximum to the initial permeability.

Measuring the impedance with, let’s say, 24 Volt AC RMS at 600 Hz is an equivalent of measuring with 2.4 Volts @ 60 Hz, from the flux densities point of view. The results are nowhere near the impedances the motor “sees” in normal operation. Large inductive “imbalances” may appear.
Let’s now look at the LIR (low influence rotor), just an abbreviation that explains nothing. In my view, the majority of the LIR motors are LIR, because they have open rotor slots (mostly manufactured cages). When testing those motors with low voltages, the flux does not have a chance to travel over the surface of the rotor iron; it has to go around the bars. If a rotor is “LIR” it obviously has no inductive unbalance [unless there is a broken bar(s)].
Then there are motors with closed rotor slots. I have been testing rotors long enough to know that the iron on the surface of those rotors is notoriously non-symmetrical. On some slower motors some slots are even open (above the spider) while some are closed. The amount of iron over the rotor bars varies widely. It introduces asymmetry that can be seen as a variable inductance on the RIC test. Yet those asymmetries are totally irrelevant during the normal 60 Hz operation. (This paragraph may need some revisiting, and I will get back to it if somebody wants me to).

The attachment shows the impact of the low voltage on impedance measurement. Everybody can repeat such a test for himself. It requires only a variable transformer, voltmeter and ammeter and the will to do it. But there is another proof. You can go back to your MCE data. From the measurements of the inductive unbalance, take the average inductance and from that you can calculate the ratio of the locked rotor current to the nameplate current. The ratio should be something like 5 to 6 to 7 for majority of the motors. But from the MCE inductances you will find numbers much lower (2x, 3x). The explanation is, that the test does not actually see the rotor bars, it just sees the surface of the rotor and its irregularities. The difference between the open slots and the closed slots is striking. I have done that years ago on about 50 motors, unfortunately do not have the raw data any more.
One more thing: The results of the RIC test from the PdMA site
www.pdma.com/PDF/CS0402.pdf
The data from the link for the motor are: 3500hp, 4160 Volts, 3590 rpm, FLA= 425 Amps. We can also read the average inductance as measured by PdMA: 17.441mH. Assuming that the inductance was measured single-phase, line-to-line, the reactance of the motor per phase at 60 Hz is:
X= ½ w*L =1/2 *2*pi*f = ½*2 *pi* 60* 17.441/1000= 3.285 Ohm per phase. (The factor ½ reflects the fact that the measurement was made single-phase). So we can calculate the locked rotor current I:
I = (V/1.73)/ X= 2400/3.286= 730 Amps.
So the locked rotor current for this motor is 100*730/425= 171% of the full load current. NOT! Obviously something is wrong. The locked rotor current should be 500 or 600 %! The Reliance catalogue gives me data for 3500 hp, 2-pole, 4000 Volts, FLA= 429 Amps, Locked rotor current = 2516 Amps.
The answer is in the shape of the attached curves. The flux density during the inductance measurement was so miniscule (note that the frequency was 1200 Hz), that the measured impedance was totally wrong. The RIC test did not have a clue that there are bars (broken or not) on the rotor. The tiny magnetic flux bypassed the bars completely, flowing over the surface of the closed slots of the rotor. However it obediently created the RIC pattern (with some irrelevant “anomalies”). Considering that the waveforms are created from only 18 points while spinning the 3500 hp rotor in babbit bearings by hand (!), the shape is almost perfect.
It is quite obvious that a proper single-phase test with reasonable current would show high variations - 22 out of 51 bars were broken! I don’t think that 425 Amps would be necessary; 200 Amps would be plenty good. A test like that would require much more than battery-powered instrument, but at least it would see the bars.
I am glad that the article http://www.pdma.com/Rotortest.html was brought to attention. Read for example on page 4 in the paragraph Testing with the Motor Disassembled on Growler Testing:”….If a rotor bar is broken, the alternating voltage at the location of the break will cause the thin piece of metal to vibrate. …”
This article must have been on the net for good 12 years. Nobody seems to have noticed that it is the exact opposite. There are other pearls in that article that rival the: “influence of the rotor’s residual magnetism on the stator’s phase-to-phase inductance...” as pointed out before.
Another controversial topic, but this time with a big difference. The interested parties want to keep discussing.
The support I am receiving is encouraging.
jank

reactance_line_to_line.pdf (39 Kb, 40 downloads) dependence of the motor impedance on voltage