Galled Bearing Cones Puzzle - Shimano FH-M510 Hub ???

[Steve Sr .]

In all of the bearings that I have repacked over my many years of cycling I have never seen this. I
posted a previous question about bearing "roar" when the wheel was spun up on the work stand. A few
folks mentioned that this was probably the freehub and not to worry.

I finally got a chance to pull everything apart today. Both sides had plenty of grease on the wheel
bearings. The non-drive side actually looked pretty clean.

Once I started to clean up all the parts I noticed that both bearing cones looked like they had
galled in rather big chunks. The races in the hub and the freehub looked fine. The ball bearings
also looked fine.

This hub is only about 2 years old and has been ridden about 5-6K miles in good weather. When I
first got these wheels some folks in this group suggested re-greasing them as Shimano didn't use
enough grease at the factory. Taking that advice I tore the hubs apart and regreased them using a
good quality synthetic grease.

Can anyone explain why this bearing failure happened and what to do to prevent it from recurring? Is
this a known problem with these hubs? It almost seems that the bearing cones were too brittle and
rather than wear they just fractured.

As far as repair should I also replace the balls as well as the cones?

Thanks,

Steve

 

[Jim Beam]

Steve Sr. wrote:

> Can anyone explain why this bearing failure happened and what to do to prevent it from recurring?
> Is this a known problem with these hubs? It almost seems that the bearing cones were too brittle
> and rather than wear they just fractured.

spalling, what you're seeing here, is the usual mode of failure for bike bearings [being as they
don't usually run at extremes of heat or speed]. the material fatigues and comes away in small
chunks. it's also commonly confused with ingress of road grit.

better quality bearings last a lot longer than cheap ones because surface finish and materials are
better and both have a significant effect on fatigue.

>
> As far as repair should I also replace the balls as well as the cones?

yes, but replacing either is a pointless exercise if the outer races are similarly pitted or worn.
suggest you check those first with a magnifier before investing any further money.

jb

 

[Steve Sr .]

On Tue, 09 Dec 2003 01:53:54 GMT, jim beam <[email protected]> wrote:

>
>
>Steve Sr. wrote:
>
>
>> Can anyone explain why this bearing failure happened and what to do to prevent it from recurring?
>> Is this a known problem with these hubs? It almost seems that the bearing cones were too brittle
>> and rather than wear they just fractured.
>
>spalling, what you're seeing here, is the usual mode of failure for bike bearings [being as they
>don't usually run at extremes of heat or speed]. the material fatigues and comes away in small
>chunks. it's also commonly confused with ingress of road grit.
>
>better quality bearings last a lot longer than cheap ones because surface finish and materials are
>better and both have a significant effect on fatigue.

So you are saying that Shimano's initial quality was ****?

>> As far as repair should I also replace the balls as well as the cones?
>
>yes, but replacing either is a pointless exercise if the outer races are similarly pitted or worn.
>suggest you check those first with a magnifier before investing any further money.

You mean the inner races? These are absolutely clean. I fact, I can't even tell the contact patch
from surrounding raw material. I.e. there is no telltale bearing "wear line" at all.

Steve

>
>jb

 

[Jim Beam]

Steve Sr. wrote:
> On Tue, 09 Dec 2003 01:53:54 GMT, jim beam <[email protected]> wrote:
>
>
>>
>>Steve Sr. wrote:
>>
>>
>>
>>>Can anyone explain why this bearing failure happened and what to do to prevent it from recurring?
>>>Is this a known problem with these hubs? It almost seems that the bearing cones were too brittle
>>>and rather than wear they just fractured.
>>
>>spalling, what you're seeing here, is the usual mode of failure for bike bearings [being as they
>>don't usually run at extremes of heat or speed]. the material fatigues and comes away in small
>>chunks. it's also commonly confused with ingress of road grit.
>>
>>better quality bearings last a lot longer than cheap ones because surface finish and materials are
>>better and both have a significant effect on fatigue.
>
>
> So you are saying that Shimano's initial quality was ****?

for a cheap hub, yes.

>
>
>>>As far as repair should I also replace the balls as well as the cones?
>>
>>yes, but replacing either is a pointless exercise if the outer races are similarly pitted or worn.
>>suggest you check those first with a magnifier before investing any further money.
>
>
> You mean the inner races? These are absolutely clean. I fact, I can't even tell the contact patch
> from surrounding raw material. I.e. there is no telltale bearing "wear line" at all.

sounds like it's repairable.

>
> Steve
>
>
>>jb

 

[Qui si parla Ca]

steve-<< Once I started to clean up all the parts I noticed that both bearing cones looked like they
had galled in rather big chunks. The races in the hub and the freehub looked fine. The ball bearings
also looked fine. >><BR><BR>

New grade 25 bearing balls, new cones, clean and regrease and adjust well. OVH more often. These
cones are not a real, hard metal, but it is good that the races are OK...

Peter Chisholm Vecchio's Bicicletteria 1833 Pearl St. Boulder, CO, 80302
(303)440-3535 http://www.vecchios.com "Ruote convenzionali costruite eccezionalmente bene"

 

[Werehatrack]

On Mon, 08 Dec 2003 21:48:26 -0500, Steve Sr. <[email protected]> may have said:

>So you are saying that Shimano's initial quality was ****?

For what it's worth, I've seen similar failures in well-greased Shimano hubs, some of them not very
old or hard-ridden, several times this past year, both on the home fleet and on visitor bikes. I
can't say that I've found a pattern that relates to any specific Shimano product line, but I don't
see the high-end stuff unless it's old enough to get to the thrift stores. As such, all I can say is
that sometimes the cones seen given to failures that do not appear related to any specific cause
that I can see. And, like yours, the cups in the hubs most often show no signs of wear, even under a
strong magnifier.

>>> As far as repair should I also replace the balls as well as the cones?
>>
>>yes, but replacing either is a pointless exercise if the outer races are similarly pitted or worn.
>>suggest you check those first with a magnifier before investing any further money.
>
>You mean the inner races? These are absolutely clean. I fact, I can't even tell the contact patch
>from surrounding raw material. I.e. there is no telltale bearing "wear line" at all.

I'd still replace the balls. In each failure I've seen, at least some of the balls had surface
damage. Ball bearings are cheap and easy to replace. Hubs are a right royal pain in the neck.

--
My email address is antispammed; pull WEEDS if replying via e-mail. Yes, I have a killfile. If I
don't respond to something, it's also possible that I'm busy. Words processed in a facility that
contains nuts.

 

[Steve Sr .]

Peter,

On 09 Dec 2003 13:46:33 GMT, [email protected] (Qui si parla Campagnolo) wrote:

>steve-<< Once I started to clean up all the parts I noticed that both bearing cones looked like
>they had galled in rather big chunks. The races in the hub and the freehub looked fine. The ball
>bearings also looked fine. >><BR><BR>
>
>New grade 25 bearing balls, new cones, clean and regrease and adjust well. OVH more often. These
>cones are not a real, hard metal, but it is good that the races are OK...

Thanks for the advice which I will heed. I have a question concerning "overhaul more often" These
bearings showed plenty of grease and no contamination on disassembly therefore I don't see that more
frequent maintenance would have prevented the problem.

Is it possible that too much bearing preload would cause such a failure? What is the best method of
adjusting bearing preload? I think I hear another thread approaching.

Steve

>
>Peter Chisholm Vecchio's Bicicletteria 1833 Pearl St. Boulder, CO, 80302
>(303)440-3535 http://www.vecchios.com "Ruote convenzionali costruite eccezionalmente bene"

 

[Jobst Brandt]

Steve Sr? writes:

> In all of the bearings that I have repacked over my many years of cycling I have never seen this.
> I posted a previous question about bearing "roar" when the wheel was spun up on the work stand. A
> few folks mentioned that this was probably the freehub and not to worry.

> I finally got a chance to pull everything apart today. Both sides had plenty of grease on the
> wheel bearings. The non-drive side actually looked pretty clean.

> Once I started to clean up all the parts I noticed that both bearing cones looked like they had
> galled in rather big chunks. The races in the hub and the freehub looked fine. The ball bearings
> also looked fine.

> This hub is only about 2 years old and has been ridden about 5-6K miles in good weather. When I
> first got these wheels some folks in this group suggested re-greasing them as Shimano didn't use
> enough grease at the factory. Taking that advice I tore the hubs apart and re-greased them using a
> good quality synthetic grease.

> Can anyone explain why this bearing failure happened and what to do to prevent it from recurring?
> Is this a known problem with these hubs? It almost seems that the bearing cones were too brittle
> and rather than wear they just fractured.

It doesn't take long to destroy a bearing if it has excessive preload from QR closure. This occurs
often because the bearing is adjusted "perfectly" out of the bicycle, after which, for safety
sake, the QR is adjusted for a tight closure. This is like putting the hub in a vise and
compressing the axle. Next time check how freely the wheel swings to a stop with the QR just
touching (partial closure) and then again when it is fully closed. For rear wheels, do this
without the chain on the FW.

> As far as repair should I also replace the balls as well as the cones?

That wouldn't hurt. Those balls are probably scuffed although not visibly.

Jobst Brandt [email protected]

 

[Steve Sr .]

On Wed, 10 Dec 2003 18:51:50 GMT, [email protected]
wrote:

>Steve Sr? writes:
>
>> In all of the bearings that I have repacked over my many years of cycling I have never seen this.
>> I posted a previous question about bearing "roar" when the wheel was spun up on the work stand. A
>> few folks mentioned that this was probably the freehub and not to worry.
>
>> I finally got a chance to pull everything apart today. Both sides had plenty of grease on the
>> wheel bearings. The non-drive side actually looked pretty clean.
>
>> Once I started to clean up all the parts I noticed that both bearing cones looked like they had
>> galled in rather big chunks. The races in the hub and the freehub looked fine. The ball bearings
>> also looked fine.
>
>> This hub is only about 2 years old and has been ridden about 5-6K miles in good weather. When I
>> first got these wheels some folks in this group suggested re-greasing them as Shimano didn't use
>> enough grease at the factory. Taking that advice I tore the hubs apart and re-greased them using
>> a good quality synthetic grease.
>
>> Can anyone explain why this bearing failure happened and what to do to prevent it from recurring?
>> Is this a known problem with these hubs? It almost seems that the bearing cones were too brittle
>> and rather than wear they just fractured.
>
>It doesn't take long to destroy a bearing if it has excessive preload from QR closure. This occurs
>often because the bearing is adjusted "perfectly" out of the bicycle, after which, for safety
>sake, the QR is adjusted for a tight closure. This is like putting the hub in a vise and
>compressing the axle. Next time check how freely the wheel swings to a stop with the QR just
>touching (partial closure) and then again when it is fully closed. For rear wheels, do this
>without the chain on the FW.
>
>> As far as repair should I also replace the balls as well as the cones?
>
>That wouldn't hurt. Those balls are probably scuffed although not visibly.
>
>Jobst Brandt [email protected]

Thanks everyone for the response. Based on you comments I am going to chock this one up to too much
preload. Parts are on order.

I just reassembled my rear MTB hub (different bike) and used the adjusting method on the Park Tool
site where you basically clamp the wheel into one side of the frame with the QR skewer and a
spacer/washer. This compresses the axle against one of the cones but does not compress cone-cone as
in real life. Is this a good method to use and should I somehow compensate for the fact that the
other bearing cone is not in compression?

I guess another way to put this is which parts do most of the deforming when compressed by a QR
skewer? Is it the axle mostly or does compressing the other cone and lock nut add significantly more
bearing compression?

Checking wheel rotation is a good idea, however, this should be done with the axle seals removed to
avoid masking the result. On some wheels this is more problematic than others.

Steve

 

[Jobst Brandt]

Steve Sr? writes:

> Thanks everyone for the response. Based on your comments I am going to chock this one up to too
> much preload. Parts are on order.

> I just reassembled my rear MTB hub (different bike) and used the adjusting method on the Park Tool
> site where you basically clamp the wheel into one side of the frame with the QR skewer and a
> spacer/washer. This compresses the axle against one of the cones but does not compress cone-cone
> as in real life. Is this a good method to use and should I somehow compensate for the fact that
> the other bearing cone is not in compression?

If the wheel gets tight upon QR closure, the entire axle (and cones) will be pressed together. The
skewer pull on the far side of the axle, whether it is on the far side of the axle or another
dropout. Therefore, this is akin to installing the wheel and closing the QR. I forgot that most
MTB's require QR adjustment every time a wheel is installed because the dropouts have retention
lips. With road bicycles, this is often not the case, so that QR adjustment remains unchanged over
many wheel changes. That's why I don't bother with the Park Tool method. Besides, I know about how
loose the cones need to be to be before closing my QR.

> I guess another way to put this is which parts do most of the deforming when compressed by a QR
> skewer? Is it the axle mostly or does compressing the other cone and lock nut add significantly
> more bearing compression?

The ratio of what changes is the ratio of cross sectional area between axle and skewer. Nothing else
is stressed (besides the QR mechanism).

> Checking wheel rotation is a good idea, however, this should be done with the axle seals removed
> to avoid masking the result. On some wheels this is more problematic than others.

Well, if these aren't brand new seals, they probably won't drag enough to make an oscillating stop
as an over tight bearing does.

Jobst Brandt [email protected]

 

[Mike Jacoubowsk]

> It doesn't take long to destroy a bearing if it has excessive preload from QR closure. This occurs
> often because the bearing is adjusted "perfectly" out of the bicycle, after which, for safety
> sake, the QR is adjusted for a tight closure. This is like putting the hub in a vise and
> compressing the axle.

Agreed. But later on-

> > As far as repair should I also replace the balls as well as the cones?
>
> That wouldn't hurt. Those balls are probably scuffed although not visibly.

If you surmised that the bearings might have been destroyed due to excessive preload, wouldn't
replacement of bearings be a requirement, not an option? Bearings cost so little compared to the
time involved in overhauling a hub, I just can't find much rationale for not replacing them, even if
everything seems OK.

But I'm also curious about the characteristics of a destroyed bearing. Do they get "scuffed" from
overtightening (due to lubrication failure)? Can bearings be dented?

In a shop environment, the only time we think of bearings being "destroyed" is when they come out in
pieces. But, since it's impractical to visibly inspect individual bearings for damage (scuff-
marks?), we simply replace them all.

--Mike-- Chain Reaction Bicycles www.ChainReaction.com

 

[A Muzi]

Steve Sr. wrote: -snip analysis of spalled cone-
> I just reassembled my rear MTB hub (different bike) and used the adjusting method on the Park Tool
> site where you basically clamp the wheel into one side of the frame with the QR skewer and a
> spacer/washer. This compresses the axle against one of the cones but does not compress cone-cone
> as in real life. Is this a good method to use and should I somehow compensate for the fact that
> the other bearing cone is not in compression?

If you close the skewer, the load is across both cones no matter if there's one or two frame ends in
the stack.

I have heard this before, but I do not see why one would do that instead of just clamping the wheel
in the bike.

What is the point? Anybody?
--
Andrew Muzi www.yellowjersey.org Open every day since 1 April, 1971

 

[Jobst Brandt]

Mike Jacoubowsky writes:

>> It doesn't take long to destroy a bearing if it has excessive preload from QR closure. This
>> occurs often because the bearing is adjusted "perfectly" out of the bicycle, after which, for
>> safety sake, the QR is adjusted for a tight closure. This is like putting the hub in a vise and
>> compressing the axle.

> Agreed. But later on-

>>> As far as repair should I also replace the balls as well as the cones?

>> That wouldn't hurt. Those balls are probably scuffed although not visibly.

> If you surmised that the bearings might have been destroyed due to excessive preload, wouldn't
> replacement of bearings be a requirement, not an option? Bearings cost so little compared to the
> time involved in overhauling a hub, I just can't find much rationale for not replacing them, even
> if everything seems OK.

Bearing balls are far tougher than bearing races and if there are none available, they could be
reused if they are still shiny. That could be because the races crack and spall, like an asphalt
street that has been cracked from heavy traffic.

> But I'm also curious about the characteristics of a destroyed bearing. Do they get "scuffed" from
> overtightening (due to lubrication failure)? Can bearings be dented?

The biggest damage to bearing balls would probably arise from running over break-outs lying in the
track of the passing balls. Most failures that I have seen, cones and balls, were spalling... part
of the hard crust breaking out. This is an obvious blemish that can be weeded out if there are no
replacement balls around.

> In a shop environment, the only time we think of bearings being "destroyed" is when they come out
> in pieces. But, since it's impractical to visibly inspect individual bearings for damage (scuff-
> marks?), we simply replace them all.

When I say "scuff marks" I mean clearly visible surface irregularities. In a repair shop
environment, there is no excuse to not use new bearing balls when there has been any sort of failure
in a bearing.

Jobst Brandt [email protected]

 

[Jim Beam]

A Muzi wrote:
> Steve Sr. wrote: -snip analysis of spalled cone-
>
>> I just reassembled my rear MTB hub (different bike) and used the adjusting method on the Park
>> Tool site where you basically clamp the wheel into one side of the frame with the QR skewer and a
>> spacer/washer. This compresses the axle against one of the cones but does not compress cone-cone
>> as in real life. Is this a good method to use and should I somehow compensate for the fact that
>> the other bearing cone is not in compression?
>
>
>
> If you close the skewer, the load is across both cones no matter if there's one or two frame ends
> in the stack.

sort of. you get compression of the spindle. but compressing the cone ends also exercises the small
amount of play that exists in the threads. it may not be a lot, but neither is spindle compression.

>
> I have heard this before, but I do not see why one would do that instead of just clamping the
> wheel in the bike.
>
> What is the point? Anybody?

 

[Jobst Brandt]

Jim Beam writes:

>>> I just reassembled my rear MTB hub (different bike) and used the adjusting method on the Park
>>> Tool site where you basically clamp the wheel into one side of the frame with the QR skewer and
>>> a spacer/washer. This compresses the axle against one of the cones but does not compress cone-
>>> cone as in real life. Is this a good method to use and should I somehow compensate for the fact
>>> that the other bearing cone is not in compression?

>> If you close the skewer, the load is across both cones no matter if there's one or two frame ends
>> in the stack.

> Sort of. You get compression of the spindle. But compressing the cone ends also exercises the
> small amount of play that exists in the threads. It may not be a lot, but neither is spindle
> compression.

The cone is already loaded in that direction by the jam nut. There is no additional compression with
two dropouts in the stack. Cone and jam nut compression is irrelevant, they having more than five
times the cross section of the axle alone.

>> I have heard this before, but I do not see why one would do that instead of just clamping the
>> wheel in the bike.

>> What is the point? Anybody?

I think the idea is that the cone and lock nut are accessible, not being covered by a dropout. That
is meaningless however, because you can't adjust it without access to the other cone that IS under a
dropout. It's an esoteric stunt as I see it.

Jobst Brandt [email protected]

 

[Jim Beam]

[email protected] wrote:
> Jim Beam writes:
>
>
>>>>I just reassembled my rear MTB hub (different bike) and used the adjusting method on the Park
>>>>Tool site where you basically clamp the wheel into one side of the frame with the QR skewer and
>>>>a spacer/washer. This compresses the axle against one of the cones but does not compress cone-
>>>>cone as in real life. Is this a good method to use and should I somehow compensate for the fact
>>>>that the other bearing cone is not in compression?
>
>
>>>If you close the skewer, the load is across both cones no matter if there's one or two frame ends
>>>in the stack.
>
>
>>Sort of. You get compression of the spindle. But compressing the cone ends also exercises the
>>small amount of play that exists in the threads. It may not be a lot, but neither is spindle
>>compression.
>
>
> The cone is already loaded in that direction by the jam nut. There is no additional compression
> with two dropouts in the stack. Cone and jam nut compression is irrelevant, they having more than
> five times the cross section of the axle alone.
>
>
>>>I have heard this before, but I do not see why one would do that instead of just clamping the
>>>wheel in the bike.
>
>
>>>What is the point? Anybody?
>
>
> I think the idea is that the cone and lock nut are accessible, not being covered by a dropout.
> That is meaningless however, because you can't adjust it without access to the other cone that IS
> under a dropout. It's an esoteric stunt as I see it.

agreed.

>
> Jobst Brandt [email protected]

 

[Adam Rush]

> It doesn't take long to destroy a bearing if it has excessive preload from QR closure. This occurs
> often because the bearing is adjusted "perfectly" out of the bicycle, after which, for safety
> sake, the QR is adjusted for a tight closure. This is like putting the hub in a vise and
> compressing the axle. Next time check how freely the wheel swings to a stop with the QR just
> touching (partial closure) and then again when it is fully closed. For rear wheels, do this
> without the chain on the FW.

Okay, i've heard the "a little bit of play"-thing a thousand times and I don't doubt that it has
some merit, but why doesn't the same apply to bolted axels? Isn't the exact same length of axel
(from one outer side of a dropout to the other) getting a hefty squeeze, too?

Is the resulting lateral force from a QR tightening just many times greater (if so, why should it
need to be?), or is it a result of a hollow axel just being easier to compress?

On that note, I adjust all hub bearings to the same perfection off the bike and can't recall a QR
wheel which didn't respond to wheel imbalances when properly clamped in. Maybe it's those ShelBroCo
lead valve stem cozies...

 

[David Kerber]

In article <[email protected]>, [email protected] says...
> > It doesn't take long to destroy a bearing if it has excessive preload from QR closure. This
> > occurs often because the bearing is adjusted "perfectly" out of the bicycle, after which, for
> > safety sake, the QR is adjusted for a tight closure. This is like putting the hub in a vise and
> > compressing the axle. Next time check how freely the wheel swings to a stop with the QR just
> > touching (partial closure) and then again when it is fully closed. For rear wheels, do this
> > without the chain on the FW.
>
> Okay, i've heard the "a little bit of play"-thing a thousand times and I don't doubt that it has
> some merit, but why doesn't the same apply to bolted axels? Isn't the exact same length of axel
> (from one outer side of a dropout to the other) getting a hefty squeeze, too?

No, because the nuts are connected to the axle. The only thing under significant compression is the
fork dropout, and the only part under significant tension is the very short section of axle between
the attachment nut and the bearing cone lock nuts.

--
Dave Kerber Fight spam: remove the ns_ from the return address before replying!

REAL programmers write self-modifying code.

 

[Steve Sr .]

On Thu, 11 Dec 2003 21:53:28 -0600, A Muzi <[email protected]>
wrote:

>Steve Sr. wrote: -snip analysis of spalled cone-
>> I just reassembled my rear MTB hub (different bike) and used the adjusting method on the Park
>> Tool site where you basically clamp the wheel into one side of the frame with the QR skewer and a
>> spacer/washer. This compresses the axle against one of the cones but does not compress cone-cone
>> as in real life. Is this a good method to use and should I somehow compensate for the fact that
>> the other bearing cone is not in compression?
>
>
>If you close the skewer, the load is across both cones no matter if there's one or two frame ends
>in the stack.
>
>I have heard this before, but I do not see why one would do that instead of just clamping the wheel
>in the bike.
>
>What is the point? Anybody?

I don't think you have a proper picture of what is happening in this situation. The object is to be
able to adjust the outboard bearing cone while the axle is under compression. To do this you are
adding a washer to the end of the QR shaft so that you are actually compressing the end of the axle
and not the outboard cone and lock nut.

Steve

 

[Ray Heindl]

[email protected] wrote:

> Mike Jacoubowsky writes:
>> If you surmised that the bearings might have been destroyed due to excessive preload, wouldn't
>> replacement of bearings be a requirement, not an option? Bearings cost so little compared to the
>> time involved in overhauling a hub, I just can't find much rationale for not replacing them, even
>> if everything seems OK.
>
> Bearing balls are far tougher than bearing races and if there are none available, they could be
> reused if they are still shiny. That could be because the races crack and spall, like an asphalt
> street that has been cracked from heavy traffic.

I've seen spalled cones, but never the cups. Are the cups made tougher because they're more
expensive to replace, or is it because they're loaded differently, or have I just been lucky?

--
Ray Heindl (remove the Xs to reply)