Raceway surface and subsurface failure in ball brgs.

[Gerry45]

Generally it is accepted that if one maintains an adequate ehl film in a rolling brg, the eventual failure will be one of subsurface fatigue due to crack propagation from partical contamination in the steel. Conventional thinking also says that if the ehl is reduced, or contaminents introduced into the lubricant, rolling surface interaction takes place and the failure becomes one of 'surface distress'. Most brgs would be removed from service when either of these two events occur.

It it is logical to assume that even though a brg may be removed because of unacceptable surface damage, some subsurface stressing must have also taken place. However, I have seen many 'abused' rolling brgs in the past,(ie damaged by vibration or small movements whilst idle etc), but not once can I recall seeing one that exhibited both types of failure mode together. Are there good technical reasons for this, or is it simply that the brgs weren't left in service long enough to allow time for both types of damage to develop ?

The reason I ask is this. I have a slewing ring application where 'surface distress' (due to the slow and occasional nature of the rotation) is normal and tolerated. (It is not noise or smoothness sensitive, but because of the operating conditions there is no alternative but to accept gradual with wear and deterioration of the raceways.) However, in the future, the frequency of slewing movements is likely to increase dramatically on my machines so I am wondering if I can expect to see not only surface damage, but some subsurface fatigue developing too after a period of time. My concern is that if this happened this could take the deterioration beyond the norm and up to unacceptable levels !


Does anyone have any thoughts or experience in this field ? My own feelings are this. Although increasing frequency of angular oscillation will contribute to more sub-surface stress cycling, it will also increases the surface wear rate. This may have the effect of wiping out the fatigue. For example, I can imagine that if I had the potential to create a 'false brinell' pits of say 10% of the ball diameter, this would be far deeper than the initial maximum shear stress point from where fatigue would normally originate. So the progression of the wear would mean that the max shear stress cycling point is constantly being moved lower down to fresh material and the onset of fatigue is delayed. I look forward to your comments.

Gerry

 

[diamondjim]

Gary,
Generally failure is often defined
as .0001 to .0003 times the ball
or roller diameter. And this definition
is to race surfaces.
What is you application specifically?
Ball or Roller Design?

The failure in slewing applications can
be defined in the amount of clearance
progression that has taken place in the
bearing and also what type of slewing
application. Does it take moment, radial
and thrust simultaneously? I would submit
your load data and duty cycles to the
bearing manufacturer for life data calculations.
These clearance measurements greatly exceed
the general failure guidelines as first stated
and I would be surprised to hear of 10 percent
values times the ball or roller diameter.
From the information that you gave initially it
should be a straight line value base on duty
cycles. Actually more cycles may end up in
more life if indeed the slewing redistributes
the grease to the surfaces. Slow slewing is
such a poor definition. Surface speed might
be more helpful for analysis purposes.

 

[electricpete]

Gerry - I didn't understand most of your comments.. above my head.

But your last paragraph on false brinneling during angular oscillation... it looks like you are considering the fatigue aspects but not the lubrication aspects. I think lack of lub is what really causes the damage during false brinneling, not fatigue. I think the angular travel needs to be a little greater than space between adjacent rolling elements to ensure that lub is maintained over time.

 

[Hippo41]

Some people repair slewing rings bu gringing out the raceways and fitting oversize rollers/balls. I have always been under the impression that this is the incorrect way and I wonder if this is because of the sub=surface stresses that you are talking about.

Slewing rings generally withdrawn from service once the defelction under load increases by a given amount from that when it was new.

If you want to fax me the original and proposed load/speed conditions etc. and current bearing details I can have the life calculations done for you. At least you'l have some idea of the change in expected life. But whoever you got your current bearing from should be able to do this for you.


Lester Milton
Telford, Shropshire, UK
technical@nbcgroup.co.uk

 

[Gerry45]

Chaps,
Thanks for your comments on this one. The application is a wind turbine blade root bearing. The loading regime is extremely complex, the pitching highly active and unpredictable and the structures flex like Spanish galleons. No brg manufacturer seems to be able to cope with all these variables and make a fatigue life prediction.

The brg makers say 'wear' is the likely failure mode of these brgs. However I've stripped a few slew rings and its obvious to me that the main reason for this is that their seals arent very waterproof ! If I can sort the sealing out, then use a very clever grease, what failure mode will I see ? It may not be wear but I hope its not fatigue. Either way it looks like I am just going to have to find out the hard way !

Gerry

 

[electricpete]

What type of inspection can be easily done to identify the subsurface distress in a post-mortem inspection?

 

[Gerry45]

As you correctly assume, sub-surface fatigue isn't visually detectable till the cracks reach the surface and bits of raceway fall out. The only non-destructive way I'm aware of is x-ray diffraction. All sorts of strange metalurgical changes take place below the surface resulting in residual stresses. X-ray analysis can measure this and make a guestimate of residual life. This kind of service is only available from the bigger/better brg makers.

Gerry

 

[rnd2]

Gerry,
What is your observation of existing longevity ? (moisture damaged units.)
Love the imagery of wind-turbines flexing like spanish galleons!

 

[Carburize]

You are in unknown territory - one of my contacts in the lubricant industry has spent nearly ten years working on developing materials for wind turbine applications which are known in the lubricant business as - "tribology experiments on sticks"

 

[Gerry45]

rnd2 - we hope these brgs will last 20 years but evidence Ive collected suggests that some wont get anywhere near this. But the limit for 'acceptable deterioration' isn't defined because we havent been there yet. One brg I looked at recently ran fine for 3yrs. On stripping, the raceways showed virtually no wear, in fact it was one of the best examples Ive seen. However within 1 month of non-use it seized solid due to rusting. I found 3% water in the grease. Obviously in this case it was only a matter of time before the grease reached a saturation level and let go of the situation completely!

Carburise - 'tribology experiments on sticks' - good one and that sums it up precisely! There seems to be no experts in this field at all. We are all just stumbling around in the dark and making advances by pure serendipidy.
Gerry

 

[diamondjim]

Gerry,
I worked for Rotek Inc. in the states
and we had many contracts for blade bearings.
Rotek was a subsidiary of Hoesch Rothe Erde
who is now owned by Krupp Steel.
We sold bearings to European
manufacturers but all of these bearings were
grease applications with very extensive seal
arrangements and I assume are not the
bearings you are speaking of. Early designs
were not preloaded and many of these failed
because of false brinelling. I am certain
that you are aware of the high vibrations
causing somewhat of an indeterminate load
situation in these applications. Are yours
two blade or three blade machines?
One of the problems is that these do have
little rotation and thus the grease does not
get distributed properly to maintain a permanent
film under the rolling elements. When the rolling
elements track in the raceways the rolling elements
essentially squeeze away the grease or film and the
only way to get the grease back under the rolling
elements would be complete rotation or enough rotation
to replenish the grease in this area.
Recently was down in the Palm Spring area of California
and saw many machines there. Travelled in Spain and
was amazed at the number of machines there. Happened
to be in Wisconsin last year when a tornado came thru
and the machines shut down in winds of excess
of 60 mph. I left more than 24 hours later and almost
all of the machines were still not operating. I do not
know if they have some procedure where each machine
must be inspected before retarting. It was really
eerie to see all of those machines idle.
Keeps us posted with your progress. And thanks for
posting the application.

 

[rnd2]

Gerry,
In your Nov 29 post you stated:
"One brg I looked at recently ran fine for 3yrs. On stripping, the raceways showed virtually no wear, in fact it was one of the best examples Ive seen. However within 1 month of non-use it seized solid due to rusting. I found 3% water in the grease."
Can we assume the old grease was washed out, new grease(any change here?) and seals(any change here?), applied and the unit returned to service at its original location?
Whatever was the prime cause of the failure looks suspiciously like it was introduced after the unit was first de-commissioned.

 

[diamondjim]

Gerry,
I think rnd2 is on the money and seems
that you have some installation ploblems
in that a bearing could last for 3yrs in
its original application and then sieze
up in such a short period of time.
Did you install new seals? The fact that
you had so much water indicates that the
seals were defective. Was the grease
cavity fully filled when you reassembled
the bearing? Did you change the type of
grease? I too would be worried if that
much water was getting into a bearing.
Are the seals double lip seals ie acting
both radially and axially?

 

[Gerry45]

Sorry but I shouldn't be too specific on detail here but
sealing does seem to be a big weakness on these large brgs when fitted to w-turbines. The problems I've seen range from blatently poor manufacture, to totally insufficient design/development for the duty. The slewring industry does very little testing and assumes that 'one seal suits all'. And since in general other customers dont complain then there isnt a problem with sealing. Its probably fair to say that in other types of application, (cranes etc) users probably aren't so fussy. They probably expect to replace a brg after 5 or 10 years or so anyway, so when grease appears around the seals after 2 years they dont worry too much.

The corrosion issue is a difficult one to accept till you experience it for yourself. But Ive had several examples where the evidence indicates that the grease can actually tolerate quite a bit of water mixed in with it, and resist damage, providing the brg continues to work. Its only when you stop that the corrosion begins.

Water that gets in to one of these brgs during service is usually broken up into small globules that are well dispersed around the interior cavities. When you decommission the brg the rusting begins on every surface. This is very different to a brg that suffers water penetration while idle. The rusting localised.

In service there is a constant flow of fresh grease to each brg which helps wash out the normal wear debris. It also helps combat the effects of leaky seals. However, when faced with evidence of poor sealing, the manufacurers response is just increase the flow rate. But I want longer service intervals and lower grease consumption and I dont just want to solve their problems by buying more grease! I sometimes with Toyota or Honda would make slew rings !

Gerry

 

[diamondjim]

Gerry,
You seem to have the proof that the seals are the
major mode or cause of failure. I would go back
to the manufacturer with the data you are presenting
here on the forum. I have seen at least 6 seal
change designs from different durometer hardnesses
and different materials happen on one single
bearing design. From single lip to double lip.
From a single seal design at each side to two.
Because the seal profile is complex, getting a good
line up when the seal is joined to make a complete
circle is difficult. The seal must be flexable to
allow it to be stretched to fit into the seal groove
or channel and still be stiff enough to seal the
cavity. I wonder if a pressure check could be made
to make certain that there is some uniform pressure
from one bearing to the other. You are right in that
most manufacturer's do not have expertise in seal
design for this unusual application of a large bearing.
Crane design bearings have axial seals which are usually
nothing more than dust covers and to keep rain out and
also lay flat like a record and the only tilt that they
may see is the amount of tilt that the terrain would
impose.

I do not know if you are talking about a 4 foot diameter
bearing or something larger.
You are the expert in the requirements for these bearings
and it is very important to track the service conditions
and life that these bearings are seeing.
Depending on the color of the blades, I have seen white
grease and black grease used to match the color of the
blades. The greases were essentially the same other
than color. I would assume these greases are very
tacky and difficult to wipe off.

I wonder if you could ensure that the idle blades be
turned at least once every 8 hours or are they locked
in one position for a purpose?

We all must be committed to the great service being
rendered by the wind industry in supplying clean
energy and an alternative to using oil. Any improvement
in seal design, bearing design, proper grease developed
would help us all. Thanks for sharing your results.

 

[EdDanzer]

Are these bearings used on the blade pitch? How many degrees of rotation is there?

 

[Gerry45]

Ed, yes, and it varies constantly. Minimum movements are 0.5 degs. Less frequently it could be 5 degs. Every few days it may be 90 degs. All depends on the winds !

Diamondjim, you mentioned preload in an earlier posting. This is now thankfully standard design feature of pitch brgs .... but unfortunately the makers of such brgs aren't scientific enough to establish the optimum values and build to them. The result is that 'preload' could mean anything between 1 micron and say 200 microns radial tightness. In terms of rigidity, these limits are a million miles apart! In terms of resisting false brinelling, 1 micron is about as useful to me as an ejector seat in a helicopter! Perhaps I just deal with the wrong suppliers. Did Rotek do any better when you worked for them ?

Gerry

 

[diamondjim]

Gerry,
How tight is tight? That has been an issue
for practical manufacturing using standard
size balls. Grinding accuracies of raceways
for large bearings is very difficult.
Often they build with undersize balls
then measure the inners to match the outers
and then use the standard ball for final assembly.
Or any combination of the above depending on the
balls or rollers in stock. If you are grinding
many bearings, you can match inner and outers
to the preload range that is specified more easily.
The preload values are mathematical derived values
by knowing the raceways diameters and the true ball
diameter that were assembled in the bearing. It is
very expensive to have these various sizes in stock
to match the grinding operations. It can be done.
diameters

 

[Hippo41]

What an interesting topic... may I add a few more comments.

1. Lubrication for bearings that are static. I used to work on bearings in the sugar mills in South africa. The large sphericals on the crushers run constantly, then go into an "off crop" period when they are idle. Because fo teh environment (hot, humid, outdoors) and 'less than recommended attention' to the seals they used to injest a lot of water and therefore rust - creating hugh problems on start-up (as they were sometimes overlooked.

We changed to a TOTAL FINA ELF CERAN grease, an overbased calcium sulphonate grease which had superior performance in the presence of large percentages of water (i.e. did not emulsify easily) and also retained film strength over the rolling elements and raceways during the shutdowns.

Result - we saved the need to replace many bearings annually.

Moral of this story? Involve specialist lube manufacturers as well as the bearing companies re. lubrication - someone will be able to extend your grease life under the conditions you experience. Which leads me onto point (2) below

Slewing ring seal life

I have seen some slewing rings last in excess of 10 years in mining equipment - which is probably as arduous as wind turbine machinery. However, they did have the luxury of constant grease injection.

So, and you can send me a cheque for this stroke of brilliance, why not ask your bearing manufacturers to install 2 single seals either side of your rolling elements - thus creating a thin circumferential grease chamber into which you can inject a small but constant supply of fresh grease.

Some may weep to the environment (so look into the newer bio-degradeable lubes) and some into the bearing (depending on the seal lip direction) thus refreshing the bearing grease. Add a suitable escape valve for the excess raceway grease through the bearing into some chamber for later collection - keeps your environmental issues to a minimum and should minimize grease consumption.

I know some manufacturers in Europe who may be interested in doing this if you like the idea. Let me know.

Lester Milton
Telford, Shropshire, UK

 

[EdDanzer]

You should try replacing the grease with gear lube. We have found that grease done not fill into the voids of bearings and bushing under going small movements at high loads. We have made second seals for smaller slewing ring bearing to contain gear lube on for a product that is built 1 at a time every few years. Another option is a rubber bellow seal, but 90° is probably too much movement. If you can keep oil in the water will probably stay out.