Root cause analysis of ball bearings 2

[Nivrah]

Hi all,
I have an NSK deep groove ball bearing designation num 6008.This has 12 ball elements. Can anyone point out another bearing manufacturer equivalent that has the same number of balls? or more!

Secondly: We have had numerous failure of this 6008 NSK and I suspect that it may be mounting problem from our fitters imparting shock loads to it since there are dents at ball pitch of the stationary inner raceway.
I am looking at different scenarios and one of the question I have :
if I utilise an equivalent bearing with more ball elements, this may help reduce the load distribution on the ball elements, right?

 

[9880501843]

Hi Nivrah,

If you increase no. of balls load distribution b/n balls will reduce naturally.
But i would suggest you to use case carburize bearing rather than Through harden,
This may be one of the reason for dent on ball.Numerous failures means not only because of mountings but I also suggest you recheck your bearing selection calculation.
9108K This is the NSK equivalent TImken bearing

 

[Fugeguy]

I don't agree with the approach of trying to accomodate an incorrect mounting method.

Also, please be advised that case hardening is not likely to solve the problem. It "might" reduce the damage but if you are impacting the bearings hard enough to brinell the raceway of a through hardened bearing case hardened material is not a magic bullet despite decades of marketing by Timken.

The best fix would be a process improvement either to use appropriate sleeves to spread the load out over the entire bearing or better yet invest in an induction heater (if it fits your process) or other thermal process to eliminate the impacts all together.

Attached is some general information on correct mounting procedures.

Additional information here:

http://www.koyousa.com/brochures/pdfs/koyotrainingmanual(web)nov2011.pdf

 

[Tmoose]

what happens to one of these bearings when it "fails" ?

 

[Nivrah]

9880501843,
This bearing is from a paper sheeter and it has been seleced from the OEM. When it fails, the sheeter goes offline and then our fitters oull the roll , dismount the bearing and fit new ones in.

it is a rotating outer ring with a stationary outer ring load. So the inner ring is a tight fit and the outer ring is a loose fit
The reason why i suspect an incorrect mounting method is because of the following:
during mounting, two bearings need to be fitted (one fixed and one floating).. the fitters fir the floating end first using appropriate fitting tool. they then proceed to fit the fixed end bearing again using appropriate fitting tools.Howeve, when they are fitting the fixed end brg, there is nothing supporting the float end brg. My suspicios is that whils fitting the fixed end brg, the load is being transferred to the float end and since it has nothig supporting the inner and outer ring together, this may be causing brinelling.

 

[Tmoose]

Hi Nivrah,

I wondered what condition constitutes a "failure" for you.

Noisy operation?
Smoking destruction?
Excessive clearance develops?
Something else altogether?

Dan T

 

[Nivrah]

Tmoose,
the rolls stop turning when the bearing fails. I have not been on the shop floor so I am not 100% sure what constituted the failure. I was called after and asked to come up with a reliability improvement technique.

 

[9880501843]

Nivrah,

I don't agree the explanation given by you regarding the load transfer due to mounting of bearings.

Usually float bearing will not take much of the load as that of fixed bearing.

one of the reason may be mounting procedure,please check the mounting of float bearings check that the float bearing is mounted properly to allow to float on the shaft.
Also please check the fitting practices.
Also please check is there any excessive vibrations in the system.

If you have photographs of damaged bearings please do share,it will helps lot to trace mode of failure.

Pradeep

 

[Tmoose]

The shop is going to have to provide some failed bearings to figure out what happened. Completely destroyed bearings don't provide as many clues as bearings near the beginning of failure.

Vibration analysis can often detect developing problems. One of those $500 vibration "pens" is a lot better than nothing.

 

[MikeHalloran]

Make a tool that attaches to the end of the roll and pushes the bearing into place without contacting the outer ring. Or whatever. There is no defensible reason for transmitting the mounting force through the bearing.



Mike Halloran
Pembroke Pines, FL, USA

 

[Fugeguy]

"9880501843,
This bearing is from a paper sheeter and it has been seleced from the OEM. When it fails, the sheeter goes offline and then our fitters oull the roll , dismount the bearing and fit new ones in.

it is a rotating outer ring with a stationary outer ring load. So the inner ring is a tight fit and the outer ring is a loose fit
The reason why i suspect an incorrect mounting method is because of the following:
during mounting, two bearings need to be fitted (one fixed and one floating).. the fitters fir the floating end first using appropriate fitting tool. they then proceed to fit the fixed end bearing again using appropriate fitting tools.Howeve, when they are fitting the fixed end brg, there is nothing supporting the float end brg. My suspicios is that whils fitting the fixed end brg, the load is being transferred to the float end and since it has nothig supporting the inner and outer ring together, this may be causing brinelling."

If this was how the bearings were being mounted most likely you would not have a problem.

Look for something else.

 

[Nivrah]

when u mount the float end bearing, you have to have a bearing sleeve supporting the inner and outer races of both the float end bearing and the fixed end bearing. failure to adopt this practice will cause relative movement of raceways. it doesnt take much to brinell a deep groove ball bearing. Machine floor vibration is "Normal" such that it increases actual load by a factor of 1.5.

We have agreed, as "MikeHalloran" that we are making a jig with two end plates and two bearing sleeve. This will be used in conjunction with a hydraulic plunger to press fit the bearings in place.

 

[Tmoose]

"when u mount the float end bearing, you have to have a bearing sleeve supporting the inner and outer races of both the float end bearing and the fixed end bearing."

If you are referring to axial support during installation, and installation is by pressing or driving the bearing onto the roll journal, I'd say "supporting" either outer ring is not a good idea. The force to install the bearing should be applied only to the inner race. The force to resist motion of the roll should only be applied to the roll.

Or, are you referring to the situation shown in Figure 6, on page 13 of the SKF Guide available here?

http://www.ien.com/uploadedfiles/ien/Blog/SKFBearingInstallationAndMaintenance.pdf

The shaft shoulders and abutments and fillet must be VERY nicely made to avoid tilting the bearing even when the cylindrical diameter is an interference fit.

 

[mfgenggear]

fugeguy

nice info

thanks
Mfgenggear

 

[unclesyd]

A posted above some pictures would be nice. Also as posted above do you have the analytical tools to asses the bearing on startup after the change out? You got to catch it before mass destruction. It might pay to check the bearing at idle.

Does the bearing sit idle for any time, with or without proper lubrication.

 

[Nivrah]

Picture of dents at ball pitch attached .
the bearings do not sit idle. it is a 6008 VV.
the only analytical assessment I did is to undertake a performance assessment of the bearing during operation: check whether equivalent load rating, equivalent static load rating, critical speed are all within requirements and they were.

 

[Nivrah]

And this is what I meant by having to bearing sleeves to support both float bearing and fixed bearing during mounting (see attached). this practice ensures that there is no relative movement between inner and outer race for both fixed end bearing and float end bearing.

 

[unclesyd]

That doesn't look brinelling or even false brinelling to me.
A few questions>
Is the damage all around the race?
Is there any color on balls or race?
Can you take another picture with without the cage and with a more oblique view.

 

[Nivrah]

the pattern is not offset to one side of the raceway so i don't suspect excessice axial load.
i would say its showing signs of dents at ball pitch followed by flaking.
no, damage is not all around the race
no colour on balls.
sorry, other pictures did not turn out good.

 

[Skogsgurra]

One thing is for sure, that is not bearing currents/EDM.
The balls are shiny and the damage is very local and on exact ball pitch.
That kind of damage can, in my world, only occur as a result of inappropriate mounting method.

Gunnar Englund

http://www.gke.org

--------------------------------------
Half full - Half empty? I don't mind. It's what in it that counts.