Hi guys,

I just need some help in the calculation of defect length in an antifriction bearing!! I got some equations for calculating the defect length from Taylor's Vibration Handbook. But some explanation part was lost due to some damage in the book.

It says,

Lp = 1*Cir / (Nb/2)(1+Bdcos Φ / Pd)

Llz = 2*(Cir / Nb)

[(Np-1)*Lp – Llz] <= Ld < [Np (Lp) – Llz]

Where,

Lp = Length between pass.
Cir = Circumferance of the inner race.
Nb = Number of balls.
Bd = Ball diameter.
Φ = Contact angle.
Pd = Pitch diameter.
Llz = Length of the load zone.
Ld = Length of the defect.
Np = ?

I want to know what "Np" means… It was lost from my book!!!Can anyone help me?

Have a nice day!!

Dear Jenish,

I guess this is an academic argument only or else what does it matter to the guy on the shop floor whether the damage is one cm long or two cm.s long?

Regards
Ibrahim

Jenish,

Page 198 of Taylor's book: "distinctive number of pulses (Np) is generated in the time signal". "The number of pulses equals the number of rollers that hit the defect while in the load zone."

There may be some academic value of trying to calculate fault length, but what is the practical value?

Walt

I would guess his reason might be like others. Who like myself wants to know if our training and learning in the feild has paid off. Knowing what you are looking at physically as well as what the waveforms and spectrums are telling us.
It does seem redundant but if you can tell the mechanic the defect is this deep and this wide it puts a realistic value to the problem. I know of quite a few times saying we are at 7G's impacting in the waveform and the spectrum has a high floor and have the what does that mean to me look but if I say the inner race has a hole the size of a pea they listen a little harder. This is the hook that can get a lot more people involved they have a better concept of what you are telling them.So the next time you say you have high impacting and a wonky spectrum they have a much better idea of what it is you are telling them.

I agree it is useful to be able to understand as much as we can about the bearing condition.

But one the danger to bear in mind is that any predictions communicated to your plant customer may be checked when the bearing is removed. If you have predicted the defect 1" and it's only 0.05", your customer might view that as a failure. (butit shouldn't be - you correctly identified that the bearing needed to be removed).

I have doubts on how accurate that method is. It relies on identifying the beginning of the defect and the end of the defect within the time waveform. I have never seen a time waveform where I could clearly identify the beginning and end of the defect (usually there is some ring-down which complicates the picture). Of course it is often easy to identify the length of time between impacts, but not the duration of the impact.

Hi guys,

I agree with you all. This may not have that much importance in practical situations.But i think it will definitly increase your confidence in yourself. What i think is, there should always be some challenge in our work so that you will not be bored by just predicting an inner or outer race defect by looking at the fault frequencies only. Just imagine a situation that you predict a defect length of more than 1" and you got a defect of 1.5".Is it thrilling for you? At least you can bet with your collegue that this defect would be this much length and lose your bet.Ha ha ha..

The practical value is the confidence it may build on you.You would become capable of predicting not just a defect, but itz length also. Now people are calling my analyzer a "Magic Box". Then how would they call if it can predict even more.Is it too dangerous?

Pete,
I agree with you. It may be difficult most of the time to identify the "Np". Thatz why most of the guys says it may be a waste of time.ryt?

Walt,
Thanks for the help!!

Have a nice day!!

From a practical point of view, trying to be too specific in diagnostic results can turn success into failure. To call out a particular defect and length doesn't really serve a purpose.

A good example of this was a report my colleagues did on a ship's main propulsion gearbox. These are very complex, with two input shafts turning a single propeller shaft through many stages of reduction. A problem was noted with one pair of gears on one side of the gear box. Rather than stop at that point, a specific fault was detailed with one of the quill shafts in that gear set.

Removing the cover on a main propulaion gearbox isn't a trival task, requiring serveral days of round the clock effort. The powers that be were convinced of the seriousness of the problem, so the ship was taken out of service, the cover pulled and the reported quill shaft inspected. No problems, no loosenes, nothing out of the ordinary was found. The gearbox was closed.

The next time the ship sailed, the gearbox failed. Guess what? A different area on that same pair of gears had failed! A more general recommendation would have resulted in a boarder look at the gearset, possibly requiring several additional hours, but the problem would have been found and corrected.

Instead of looking like heros for identifying the problem and preventing failure, my colleagues ended up looking somewhat incompetent and instead of getting a problem resolved they wasted a huge amount of time and labor.

Jon
Spintelligent Labs

Jon, you couldn't be more right. I have analyzed many failed bearings and correlated the observed physical characteristics back to historical vibration data. I've learned a great deal, but I hope I'm never "smart enough" (I'm being sarcastic) to make THAT specific of a call. I'd only be setting myself up for failure. Now, off the record, that's a different story!

Actually, I have written an article related to this topic that appears in the August issue of Uptime Magazine (Soul Mates: Vibration Analysis and Bearing Analysis Were Made for Each Other). There are a lot of errors in the editing (captions, etc.) but it should at least be interesting to those of you trying to gain a deeper understanding of your vibration data as it relates to bearing condition monitoring. (I would appreciate any comments related to article, if anyone is so inclined.)
andrew.calcagno@gm.com

I once called a newly installed bearing in a calendar stack as having a cracked inner race. The bearing had been sitting mounted on a roll for about 20 years or so and had been previously used. The roll was installed as a unit.

The bearing was removed and spray penetrant was used to check for a crack in the inner race. It was reported to the Maintenance Manager that no crack had been found and that there was nothing wrong with the bearing.

I brought the bearing back to the shop, cut it open and found serious brinelling in the inner race. I took it back to the plant to show them my findings and they were unimpressed despite the fact that the damage was measurably deep around the entire perifery of the inner race. Unbelievably, this "bad call" was used by the Maintenance Manager as evidence to discontinue my services at that plant.

Lessons learned:

1. Don't be too specific.
2. Don't question Ralph Stewart who had told me the race was not cracked.
3. Don't give an ineffective Maintenance Manager any help in covering his butt at your own expense.

I agree it's ridiculous they would look at the bearing inner ring bore without cutting open the bearing and draw any conclusions... then ignored the evidence on the inner race.

A small comment on terminology that might help your cause if your call was "cracked inner race"

The inner race refers to the surface of the inner ring on which the rolling elements ride. The "ring" is the entire piece and the "race" is a surface of that piece.

Calling cracked inner race would mean a crack on the working surface and you have to cut open the bearing to see it. If you intended to mean a crack of the whole piece which extended all the way through, you would have said "cracked inner ring", but not cracked inner race.

I doubt this subtle distinction will mean much with the mentality you're up against, but I just wanted to mention it.

Pete,

Maybe if I had used the proper terminology I would still be working there.

Danny

Danny,

I think you DID use the correct terminology ("cracked inner race", as opposed to "cracked inner ring"). As Pete points out, you can't really see the inner race without dismantling the bearing.

About the only fault I've seen that is vaguely predictable is a cracked inner ring on a drying cylinder on a paper machine; these are relatively common, and, if you catch it early enough, the time waveform (especially in peakvue) shows very sharp, localised impacting at BPFI. I've attached an example. But even these are a dubious call - a few months ago, I made a similar call and it turned out the inner ring was not cracked, but there were the beginnings of fatigue cracking on the raceway.

Ian

cracked_inner_ring.doc (31 Kb, 26 downloads)

Hi guys,

I agree with you all..Sometimes it will hit back if we go for more specific.Actually i started my career as a vibartion specialist with a very intelligent and nice boss.During my early learning period with him,i noticed that even though he was finding the faults to the exact, he was not mentioning it in the reports. It irritated me alot at that time.But now i'm familiar with the way, we; Condition Monitoring guys; have to talk.

Have anice day!!

What an interesting thread this turned out to be! An opportunity to bring out our most embarrassing moments of our careers.

In the early days of my career, I one time predicted that "The electric motor has no faults!"...Lo and behold, that @#&* motor burnt down to ashes the day after my report was received. Culprit was very low insulation resistance but how could anyone detect insulation resistance using vibration analysis but explain that to the customer. Ever since that day a couple of decades away, my reports claim "There are no faults, detectable by vibration analysis, in this machine". All ends of one's body are covered that way.

Regards
Ibrahim

Hi guys,
I also came into a similar situation like what vibraTek told.
The villain was an Overhung Dedusting Fan.It was running at a high level for sometime and we were monitoring it.The fault was Unbalance due to dust deposition in the impeller..We were waiting for a proper time for replacing the impeller coz as per the plant maintenance it was difficult to clean the impeller on site.But suddenly the vibration levels increased to around 50mm/s and we decided to have an emergency shutdown.New Fan with new bearings were installed and we checked the vibration levels and it was under acceptable limits. But we notice some high frequency peaks in the new fan bearings.I report it but they consider it as a lubrication problem and grease the bearing.Since the levels were very low, this problem was presented as a minor lubrication problem and not properly attended.

But the Fan failed the next day.The bearings were burned and melted and it was a total disaster.Normally the blame was on me!!I checked the Fan bearings. The interesting thing was, the bearings were interchanged. Fixed bearing was on the Non-Drive End side.Later it was revealed that it was assembled not by the workshop,but by the plant maintenance people.

Moral of the Storyon't trust anybody..Everybody can make mistakes. Also be careful with newly installed equipments. They may be far more dangerous than a machine running in danger for long time.

Sorry for such a bad story!!
Have a nice day!!