Is anybody using IR to trend motor bearing and winding temperatures?
I'd be interested in how you are going about doing it, what kind of temperature ranges you consider "hot", and how you handle the data once its collected.

We have a motor monitoring program and an infrared thermography program, and we have made a concious decision to exclude most of our motors from the routine thermography program because:
1 - Large critical motors should have continuous computer monitoring of winding temperature and (for sleeve bearing motors) bearing tempearture anyway.
2 - Smaller motors tend not to be as critical.
3 - Most bearing problems in rolling bearings on these smaller machines would be identified first by vibration anyway (lubrication of course the noteable exception).

We have made an exception for four families of small motors from 10 hp to 100hp which are very critical and have no continuous monitoring. In that case we used a sharpe marker to mark a spot on the bearing housing with a T next to it (for temperautre). When we're doing our vibration rounds, we carry a Raytech thermogun and take a single reading at that single point for each bearing. We entered this as an analogue process parameter within the vib route collection program, so the data collector prompts us to input that temperature manually when we are doing vib rounds on that equipment (we also collect certain key process pressures and temperatures this way). By luck, it just so happens that all four families of motros are in the same building on the same route, so we know to take the temperature gun along for that route. We go back later and take a full thermography camera image if a machine catches our attention, possibly due to a sudden change in that one number which is not explained by recent regreasing.

One other factor that might be considered in selecting which to monitor would of course be history... and which motors run hot. TEFC motor inboard bearings run hot due to the ventilation pattern. 3600rpm TEFC in higher horsepower ranges with large bearings (high D*N) are the ones that I consistently see hottest bearing housings.

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

Electricpete,
Thanks for the reply. Could you expand on yor motor monitoring program?
We have begun a pdm program for our motors that are in service, but not really sure if we are on the right track.
Over the last year we have probably had to change out 50-80 motors verying in size from 5 hp to 600hp with the vast majority of failures being bearing related.
Any thoughts from others are also welcome.

Paul:

As mentioned, it usually takes more than one technology to evaluate the condition of an electric motor. With infrared, you will want to consider that the surface temperature of the motor will normally be vastly different than the core temperature. Bearing housing temperatures are normally very close to the actual bearing temperature, with contact temp measurements.

I have seen a number of successful programs that use infrared as a key component. For instance, one of the large package-moving companies uses infrared on its conveyor motors (all are critical) in order to detect changes that may be due to loading, etc. If any of the motors change, they schedule additional testing during their daily one hour shutdown. Part of the program requires that they place a mark on the floor so that the analyst takes a reading from the same spot each time.

In any case, selecting the correct technologies and critical equipment will be of greatest importance in the development of your motors program. At one automotive company, we have set a best practice to evaluate critical machines and any equipment where the cost to repair or replace exceeds $25k. We also have a similar program at a large steel manufacturer. Each are successful programs. The testing is coupled with motor storage and commissioning of new/repaired equipment and installed equipment.

In this post I am attaching a motor management document (February Uptime Magazine) followed by a second document on the 'Multi-Technology Approach to Motor Diagnostics.'

Sincerely,
Howard

BackToFuture.pdf (1,218 Kb, 106 downloads)

This one is the multi-technology approach article.

Howard

The_Multi_technology_approach_to_motor_diagnostics_(3).pdf (164 Kb, 106 downloads)

Finally, the best methods for the detection of bearing related problems are:

1. Vibration;
2. Ultrasonics

Temperature methods and technologies such as electrical signature analysis can detect bearing problems, as well, but usually at much later stages.

You may also want to check your records to verify whether any of those repairs also included rewinds and general overhauls. It is common motor repair practice to replace the bearings in a motor regardless of condition. We reviewed such information at a large steel facility that thought their primary issue was bearings and mechanical. It actually turned out that they had a higher rate of electrical/winding problems as they had a very successful mechanical program which required the addition of winding analysis.

I have a few infrared standards that cover such programs (military and one civilian) if you are interested, but, as you may hear from John Snell, who takes part in this site, such standards are lacking a bit.

Happy Holidays.

Sincerely,
Howard

Paul

Wow 50 - 80 motors changed this last year.
That sounds like fun.
Have you saved any bearings from your motors that you changed out ?
Just by looking at these bearings you can learn why they are failing and then how to monitor and correct these issues.
alertanalytical.com

Mark:
Yes that is a lot of motors. I am in the process of gathering all the information I can about the motor failures from our records (which are very sparse) and from the motor repair shop. Unfortantly the motor repairs are contracted out so I may have a hard time getting this information. I have started keeping records since I became envolved with the motor pdm program, but it's like starting out at "ground zero".

Howard:
Thanks for the valuable information you have provided. I have read the articles you included before but they are certainly worth rereading. My company has a PdMA MCE/Emax tester but as of yet we havn't used it near enough. I would be intersted in your opinion on how the online testing can be used to check bearings conditions. Any thoughts?

Unless they are inaccessible machines, why wouldn't you monitor the bearings with vibration instead?

Electricpete:
In my plant there are 2 reasons.
1) while we do have vibration testing it belongs to the mechanical group and they only monitor a few select pieces of equipment.
2)I'm with the electrical group and we don't have any vibration equipment.

I didn't say they were good reasons, but unfortuntaly thats how it is here.

Dear Paul:

Yes. I would only really count on MCSA/ESA to detect bearings if the machine was inaccessible. I have detected bearings about the same time that vibration would see a fault as a 'second to third stage' failure. In other instances, I have not seen any signature, but the bearings were obviously bad. It partly depends on the load on the machine coupled with the flexibility of the rotor and shaft. Additionally, the bearing signature may not show as a first harmonic of the bearing fault signature, but may show up as the 4th through 6th harmonic of the signature (most MCSA/ESA signatures show up the same as vibration, but as sidebands of line frequency versus the bearing fault peak). Finally, I would call a bearing fault at any level above the noise floor of current, in dB.

Basically, I would use MCSA/ESA as a choice after vibration or ultrasonics, but before infrared.

Howard

Paul

You asked if you are on the right track?

Yes you are - when you start asking questions.
I can see your problem - infared is usally an electrical department PDM tool and vibration is in your mechanical department.

I have seen many electrical departments run vibration programs and do very well.
I say this because vibration is your best tool for catching your motor failures and aiding on why the motors are failing.
1.I would try to get complete mill coverage of all your motors with vibration analysis.
The task is not as hard as most people tend to believe.
2. Save all motor bearings for root cause failure analysis - this is also easy and rewarding .
3. Learn to document your findngs - this will help in convincing management that your task are helping and worthwhile.

Paul
if you do not currently have your equipment being monitored with vibration - i would push that direction first - and try to get involed.
Infared would come down the road after vibration analysis is established.
Mark

Paul
It sounds like you may have alignment or lubrication problems.

quote:

1) while we do have vibration testing it belongs to the mechanical group and they only monitor a few select pieces of equipment.
2)I'm with the electrical group and we don't have any vibration equipment.

Where I used to work as a full time analyst, the mechanical dept. had all the vibration equipment. Why? That is just the way we started out years ago and it worked out great for all concerned. In fact I am a millwright by trade, so all my experience with motors and their monitoring has come from over 20 years experience with the analysis equipment and help from the E and I guys on questions I would have. So for a mechanical dept. to have the vibration analysis equipment is not really all that bad. In fact there was a "problem" once where the E and I dept. wanted to check the motors with the vibration equipment and let the mechanical dept. check the driven parts. In my opinion there is no way this would have worked. A pump or motor can not be given a fairly shake on what might be wrong with it if the mechanic or electrical guy, which ever has the equipment at the time, can not check the entire system. So this "change" never came about, thank goodness.

To lose some many motors to bearing failures, as you stated, in a year's time might mean something is wrong with the way things are installed. The vibration equipment can check the bearings and find when they are failing and report these in time for scheduled shutdowns, most of the time, but there has to some responsiblity and root cause failure into why so many are failing.

For a monitoring program to work, every piece that would or would not cause unscheduled downtime and would or would not effect an entire plant needs to be checked if resources (people) are available. If there is a shortage of "time" then the critical equipment needs to be flagged as "this will be monitored", both driver and driven.

The use of the IR equipment is only another tool in an analyst's toolbox. The IR has its place in a program as does all equipment, but to reply solely on it for checking "heat" in bearings and winding and making a "call" from this and this only, would not be a way to go.

If your mechanical people are not monitoring your motors with their equipment, I would push for them to do so.

Of course this is only my opinion and I could be a little wrong.

We use IR as part of our QC testing when bearing temps are higher than normal. Here's one to look at - warm upper motor bearing. Needless to say the machine did not pass.

Upper_Bearing_No_Go_QC.doc (494 Kb, 45 downloads)

Same motor had a resoance issue at the upper motor bearing. If anyone is interested in the "entire" report give me shout. danvegaii@msn.com

Regards

Dan, Dan, the Vibration Man