Our vibration analyst discovered a 2X line frequency side band around the # of rotor bars times the running speed peak on a cast aluminum rotor motor. MCSA shows a very low rotor bar db peak at pole pass frequency at 1X line frequency. MCSA does show pole pass swirls at the 5th and 7th harmonics of line frequency. I've seen this in copper rotor motors and assume this in normal for them but not cast aluminum rotor motors. I'm thinking casting voids or porosity might cause this in an aluminum rotor, anybody else run into this problem before?

Jim,

2xLF sidebands around the rotor bar frequency is a fairly common source of vibration modulation in a motor of good electrical condition.

Walt

Jim:

Yes, with significant casting voids. You should be able to check for casting voids using your ALL-TEST PRO 31 or PdMA (sorry, I know you have both).

You should also check your voltage harmonics. I think that you will find that the 'swirl' type readings that you are looking at are power harmonics. Without being able to see voltage signatures and compare them with current signature spectra, with the MCSA device that you have, you would find that you have a high fifth and seventh voltage signature (300 Hz and 420Hz) that show as a lower value in current. In this case, it would be the power harmonics showing in current. Often occurs when you have VFD's and other power electronic (3-phase) devices in your facility.

I have often found that casting voids will show as multiple low dB 2SF harmonic sidebands around line frequency.

If you care to share the signature on this forum or via email (howard@motordoc.net), I would be happy to evaluate it a bit more. Also, if you include the vibration data, I may be able to help corrolate the data.

Sincerely,
Howard

Walt, thank you for the tip, which co-incides with the information I've found published by Technical Associates Level 3 of charlotte, P.C., copyrighted in 1997 which says "concern are amplitudes exceeding approximately .06 in/sec at either rotor bar pass frequency or higher harmonics".

quote:

Originally posted by Walt Strong:
Jim,

2xLF sidebands around the rotor bar frequency is a fairly common source of vibration modulation in a motor of good electrical condition.

Walt

No need to be sorry Howard, this is one of those motors that run 24/7 and its hard to get a crack at it with the off line equipment.

Until I get an online Framatome test on this motor, I won't be able to say whether it's voltge induced or not for sure, however in the 19 copper rotor motors that I've been surveying, which have the multiple swirl of Fpp around the 5th and 7th harmonic of line frequency, I can say that they do not have voltage induced spikes with them. I'll let you know what I find with the other tester when I get a chance to get back to it.

I'll make it a point to check for two times slip frequency harmonics around line frequency tomorrow and let you know what I see, but as I recall, the spectrum was pretty clear.

I do have a copy of the vibration data and will share all the data with you here on the forum, for all to see, and help corrolate the data as soon as I figure out how to do that. If I can't figure out how to do it, I'll email it directly to you. For your information, our new vibration analyst and myself are working together to corrolate the data between the two technologies and find it very interesting to do so and I'm sure we'll appreciate any and all help as we hopefully learn the ropes and will some day be able to talk sensibly to one another about a given concern. It's great to have a second opinion (or more) on any suspected problem that your not sure of not only in the medical field but in the predictive maintenance field as well.

Can anyone tell me if this forum will show a PowerPoint attachment?

And thank you Dr. Penrose for personally answering my plea for help! Great to hear from you again!

quote:

Originally posted by MotorDoc:
Jim:

Yes, with significant casting voids. You should be able to check for casting voids using your ALL-TEST PRO 31 or PdMA (sorry, I know you have both).

You should also check your voltage harmonics. I think that you will find that the 'swirl' type readings that you are looking at are power harmonics. Without being able to see voltage signatures and compare them with current signature spectra, with the MCSA device that you have, you would find that you have a high fifth and seventh voltage signature (300 Hz and 420Hz) that show as a lower value in current. In this case, it would be the power harmonics showing in current. Often occurs when you have VFD's and other power electronic (3-phase) devices in your facility.

I have often found that casting voids will show as multiple low dB 2SF harmonic sidebands around line frequency.

If you care to share the signature on this forum or via email (howard@motordoc.net), I would be happy to evaluate it a bit more. Also, if you include the vibration data, I may be able to help corrolate the data.

Sincerely,
Howard

Jim:

Glad to see you on the forum.

I assume that you will be at IMC2005? If so, I have been asked to do a few presentations at the ATPro learning lab. Would enjoy the opportunity to see some of your data.

Yes, the Empath will provide the voltage and current data. I would also be interested in seeing that, too. I have not seen enough data on rotor bars that include the 5th and 7th 'swirl' pattern to conclude that it is an accurate enough way to determine severity of rotor bar fault, and would be interested to see what you have.

Sincerely,
Howard

Howard,

First let me apologize for taking so long to reply.

I'll not be at IMC 05 this year, I've opted for Thermal Solutions in January of next year instead.

I've not taken the time to go over, collect and email the the data yet, but going back to my original question and assumption about this problem being related to casting voids in an aluminum rotor is based on what I've seen in testing copper rotor bar motors and because they have voids (that's the way they are built-with lots of holes and voids), I would expect to see that in them if that is the case. Then, if that is true, can one expect to see this in aluminum rotors? If true in the copper rotor with a known good rotor bar signature, then would it not also be applicable to voids or porosity in cast aluminum rotors with a good rotor bar signature at line frequency?

In general, after testing 19 copper and 21 aluminum rotors with emax, I saw the 5th and 7th harmonic swirl of Fpp in all of the copper ones (no significant Fpp at line frequency), but did not see this on any of the 21 aluminum ones. If the swirls were voltage induced in the copper ones, why wouldn't they be induced in the aluminum ones as well?

There was only one of the 40 300 HP motors that showed a significant rotor bar problem, and it was aluminum, around 45db at line frequency pole pass. The 40 motors were on 5 different power supplies.

I have a problem with getting the resolution fine enough in empath to enable seeing the above mentioned frequencies in either voltage or current. I'll have to work that out with the vendor to see what I'm doing wrong.

I'll keep you posted as to what I find.


Jim

Jim:

Give Don Ferree a call at Framatome (Avera) about the resolution. I have not had a problem with the newer system.

I assume that you are talking about cast copper and not copper bars. What is interesting is that the cast copper bars are supposed to have fewer voids - one of the sales points that the copper association has been pushing. The process for extruding and rolling copper bars, however, does not allow for much in the way of voids. Instead, the issues tend to be poor connections between the bars and endrings.

Anyways, yes, keep me posted on what you find.

Howard

Howard,

As a matter of fact, I'm waiting for an answer from Mr. Ferree (AREVA/Framatome-ANP) on the resolution question now!

Actually the motors under scrutiny are 35 year old 509US frame motors, have been running very well with only one recorded stator winding failure! This is compared to their replacement 449TZ with cast aluminum rotors that are not fairing nearly as well. The old Louis Allis's do have copper bars in them, not poured or cast like the cast aluminum.

The amazing thing in this study is the fact that the old L-A's pull 1/2 the current of the replacement in slow speed and 10 to 30 amps less in high speed! We are now trying to find more old salvagable 509US ODP frame 460 VAC, 3 phase, 300/75 HP, 1800/900 RPM motors to replace the 449TZ's with, because of their outstanding performance, as of to date anyway.

Having said this, please note that the real solution to this problem is replacing all of the above with a simple low cost single speed, 250 HP, off the shelf throw away motor and put it on a variable speed drive. This is the expensive answer and is being pursued, however the budget and the process doesn't allow us to do all of them at once, so we must maintain a supply of the old motors as spares until they are all phased out over time.

Stay tuned, more coming later,

Jim

Howard,

I'm going to agree with you and say that the Fpp swirls that I've seen around the 5th and 7th harmonic in the Winvis software are most likely caused by voltage related spikes. I've not been able to achieve a good enough resolution in Empath software around those frequencies of concern to make that call, however the Explorer does have the ability to do this with much better resololution and it doesn't show any swirls at the 5th and 7th line frequency harmonic in the data of the motors in question.

Explorer does show a very low level rotor bar problem as do the other two MESA devices. Relating this to 2 sideband peaks of line frequecy around #RB X RS peak in the vibration analysis spectrum of .08 in/sec db, is a trendable concern at this point and should deem decreasing the wait interval in between sampling. (The reading was a little under .04 in/sec db one year ago).

Thank you Dr. Penrose for your input!


Jim

Jim,

If you're trying to find some older motors, a good web site is (http://www.surplusrecord.com). The company I work for, can help you in other ways too, but then I'd have to put on my vendor hat. I hope you find what you're looking for.

Don

Don,

Thanks for the URL, I don't believe your web site came up the last time I did a search engine using the frame number "509US" and I see your listing is by HP and RPM which most likely explains why it didn't pop up. FYI, we did find a source and have two on order now.

Pease, see material on this item "COMPARATIVE ANALYSIS OF ELECTRICAL AND MECHANICAL
FAULT SIGNATURES IN INDUCTION MOTORS" at
http://txspace.tamu.edu/handle/1969.1/1618

Victor

The link doesn't seem to work.

Yes, indeed now it does not work, but last week everything was OK. I can send this material for you (dowloaded from this link), but I need you e-mail.
Victor

The Link works now.

quote:

Originally posted by VictorP:
Yes, indeed now it does not work, but last week everything was OK. I can send this material for you (dowloaded from this link), but I need you e-mail.
Victor

dear Victor, i cannot download the file...
can you plase send it to me
my email is Daniel.bracho@dolphinenergy.com

thanks