Vibration in a Motor Solo Run Test

[JHAESLE]

This will be a two part thread. I’m working in a new refinery project witnessing about 3,600 motor solo run tests. The majority of which start up and behave as one would expect - normal vibration and temperature profile in line with API-541. However, we have two recurring issues vendors are telling us are normal but are outside of API-541 & comparable trend. My overall question - is this normal behavior for new motors during the initial plant solo run test?

1. 3600 rpm, Medium Voltage, Ball Bearing, 500hp. Makes an intermittent high pitch medium volume noise on run down. At running speed vibration and temperature (bearing & stator) are normal. Temperature rise in stator is smooth/ gradual and settles at a constant temp. Vendor indicates the bearings are skipping on rundown causing the noise. We made the vendors replace the bearings. Noise persists.

2. 1800 rpm, Medium Voltage, Jounal Bearing, 650 hp. High displacement vibration. During Solo test DE vibration increases to 65micrometers then decreases gradually over a 30 to 45 min period as motor temp increases and stabilizes. It seems like a thermal growth related issue. Vibration stabilizes at 35micrometers after motor running temp is reached. Stator temperature is stable and normal. Vendor doesn’t give a reason just states vibration is normal behavior and will be dampened when coupled. Will be coupled to a pump with anti friction bearings - double overhung.

Both tests are 4 hrs in length. And both examples can be applied to similar situations in the plant.

 

[edison123]

Don't see any problems with both the motors. Note the baseline values of motor bearing vibrations (both overall and signatures), bearing and winding temperatures and note down the conditions (open shaft, brand new equipments commissioning etc.) for future reference.

Muthu

http://www.edison.co.in

 

[Strong]

Item-1 possibly incorrect bearing preload or poor lubrication
Item-2 possibly short-tern rotor thermal bow. Are you measuring displacement because motor has journal bearings with proximity probes? If yes, then you should look at spectrum for instability (near 45% SS) as well and 1xSS for unbalance from thermal bow.

You say you are witnessing the solo tests. Does this mean you are making your own/independent vibration measurements, and you are capable of interpreting the data?

Walt

 

[ScottyUK]

Possible cause of #2 - the oil wedge characteristics change as the bearing warms up and the oil becomes less viscous. If you have bearing metal or bearing drain temperatures available then you should look at these temperatures alongside the vibration trends.

 

[JHAESLE]

Strong - I hadn’t thought of thermal bow. In answer, yes we take our own independent casing vibrations and pull the proximity data off the DCS. However, most of our analytical tools such as the ADRE or BN System 1 are unavailable at this stage. Your right- if we could see the spectrum it would tell us a lot about the behavior.

This motor is a part of a 4 motor set A B C D. All three other motors behave as expected immediate vibration increase to 30 micrometers and stable.

 

[Karuppu. Bala]

Point 1.

From your observation, nothing so serious in this machine.
Many of the 2 pole machines, normally have these kinds of Noise.

Point 2.

There is two things to highlight.
First is, journal bearing & Shaft clearance on both sides may be not equal or DE side is more clearance.
Second is, the air-gap may be not even. This makes vibration either DE or NDE.

 

[Tmoose]

Along the lines of W Strong's #1 comment, as far as I am concerned, expecting a single row ball bearing outer race to slide axially gracefully is wishful thinking. The Length to diameter ratio of a typical single row ball bearing is simply too short to preclude jamming.

https://images-na.ssl-images-amazon.com/images/I/41JtHFK4CNL.jpg

Tweaking fits and finishes, polishing outer race feature transitions, and even using coatings occasionally helps, but really is the skimpiest of band-aids.

I don't think the normally quite thorough API in Standard 541 requires particularly detailed vibration measurements.

1 - I'd be interested in the details of the preload springs, if any. Spring style, installed force, and location.
2 - During testing start to finish I'd be measuring the high frequency vibration (see attached image) in acceleration (gs) and with one of the proprietary demodulated etc methods. Like Spike Energy, PeakVue, etc. Even UltraSound might offer some valuable insight.
I'd expect to see an increase in vibration as the motor warms up, and the shaft expands axially, causing the balls to run off center or the bearing outer race to tip slightly, causing the balls and cage to annoy each other to the point of squealing. After a while, and further axial expansion, the outer race might jump to a more neutral location and position and the irritation might even stop.

When fully warmed up (which would likely require running as in service), if the high frequency vibration remains high (see attached image), even though inaudible I think there will be a long term negative effect on bearing life.