Indexed-headset pictures
carl...@comcast.net
http://img84.imageshack.us/my.php?image=250headsetdimplesve2.jpg
Here's an enlargement of the dimples:
http://img84.imageshack.us/my.php?image=250gheadsetdimplesrb9.jpg
You may be able to enlarge either picture even further by clicking on
it.
I don't know whether the magnified picture of the dimples supports
erosion by fretting or impact by denting or both.
But some posters may enjoy interpreting details that may not have been
visible before on RBT.
The headset belonged to my faithful Schwinn LeTour '98. It spent about
two miles per day coasting down a 65 mph highway on 120 psi 700x26
tires at 30-50 mph, depending on the wind. Occasionally, it drifted
over the rumble strips and buzzed like a jackhammer. Its other daily
13 miles were at about 20 mph over somewhat cracked and bumpy
pavement. No impacts or accidents were severe enough to bend the fork
back.
After years of cruel neglect, my faithful Schwinn was retired and
replaced by its twin brother (easier than maintenance).
While fiddling with my camera this evening, I wondered about headset
indexing and remembered what a shameful example I had close at hand.
After putting the bike in a vise and disassembling things, I pounded
uselessly on the top of the stem for a while with a 4-lb hammer.
Then I gave up and took more drastic measures that resembled a
partial-birth abortion. By sawing through the exposed but hopelessly
frozen stem, I extracted the stem and the intact lower race without
pounding on it.
Cheers,
Carl Fogel
jobst....@stanfordalumni.org
http://img84.imageshack.us/my.php?image=250headsetdimplesve2.jpg
http://img84.imageshack.us/my.php?image=250gheadsetdimplesrb9.jpg
Unfortunately this is such a poor quality head bearing that the races
are deeply eroded between the ball home positions at the dimples.
You'll find that high quality (hard races and bearing balls) show
essentially no wear between dimples that are primarily in the fore and
aft quadrant of the races, these being the only places that fret
significantly when the fork articulates about the lower bearing. Tall
bicycle have enough space between lower and upper head bearing that
the steer tube flexes enough to cause dimples in the upper bearing.
Even then, I suspect you can see less wear and dimpling in the lateral
quadrants of the races. This is hard to assess from the picture
because it is not a view looking at the whole race on axis.
Damage in the pictures is so gross that it obscures what most people
with good equipment could expect to on old style cup and cone
bearings. Unfortunately, all examples I had seem to have been
scrapped when I changed to Shimano cartridge angular contact bearings.
I think I posted some pictures a few years ago though.
Jobst Brandt
carl...@comcast.net
Dear Jobst,
Here are the original pictures again, showing three dimples in
considerable magnification:
http://img84.imageshack.us/my.php?image=250headsetdimplesve2.jpg
http://img84.imageshack.us/my.php?image=250gheadsetdimplesrb9.jpg
And here's the whole headset, showing all dimples:
http://img170.imageshack.us/my.php?image=253wholeheadset3za6.jpg
I haven't tried to identify any quadrants with less wear, much less
where the 3 dimples in the other picture ended up.
You can measure things with pixel-level precision with a program like
Mouse-o-Meter:
http://www.freedownloadscenter.com/Utilities/Mouse_and_Keyboard_Utilities/Mouse_o_meter.html
If anyone wants to describe the position of the dimples, I expect that
treating the headset as a clock face will be the easiest method.
Again, clicking on the picture may enlarge it.
Cheers,
Carl Fogel
jobst....@stanfordalumni.org
http://img84.imageshack.us/my.php?image=250headsetdimplesve2.jpg
http://img84.imageshack.us/my.php?image=250gheadsetdimplesrb9.jpg
> Here are the original pictures again, showing three dimples in
> considerable magnification:
http://img84.imageshack.us/my.php?image=250headsetdimplesve2.jpg
http://img84.imageshack.us/my.php?image=250gheadsetdimplesrb9.jpg
> And here's the whole headset, showing all dimples:
http://img170.imageshack.us/my.php?image=253wholeheadset3za6.jpg
> I haven't tried to identify any quadrants with less wear, much less
> where the 3 dimples in the other picture ended up.
> You can measure things with pixel-level precision with a program like
> Mouse-o-Meter:
> If anyone wants to describe the position of the dimples, I expect
> that treating the headset as a clock face will be the easiest
> method.
> Again, clicking on the picture may enlarge it.
As I said, the deepest dimples are generally centered about the front
of head set races when the bicycle is steering straight ahead.
Therefore, when photographing it you could put that position at the
top of the picture so that 12:00 is straight up and is also the front
of the bearing.
If this were road shock induced rather than fretting the dimples would
be uniformly distributed (assuming the head tube had square ends). As
should be obvious, the forward rim of the bearing bears most of the
wheel loads and wears accordingly. The cone opposite the cup shown
will have matching marks, leaving no doubt about in use orientation.
This is more evident on a better quality bearing and if the bicycle is
used with heavy braking will have dimples in the front AND rear
quadrant. My dimpled bearings had nearly symmetrical wear in that
manner, while the side quadrants showed almost no visible wear.
Jobst Brandt
Tim McNamara
carl...@comcast.net wrote:
> Here's a badly indexed headset race:
>
> http://img84.imageshack.us/my.php?image=250headsetdimplesve2.jpg
>
> Here's an enlargement of the dimples:
>
> http://img84.imageshack.us/my.php?image=250gheadsetdimplesrb9.jpg
<snip>
> After putting the bike in a vise and disassembling things, I pounded
> uselessly on the top of the stem for a while with a 4-lb hammer.
Beating on the stem would not dimple the lower headset cup under any
circumstances. It'd dimple the upper headset cup and/or cone, if it was
going to dimple anything.
The photos you provide look like an excellent example of bearing wear.
The dimples are not round, as they would be if they were caused by
impact, and the wear trails between the dimples are nearly as bad as the
dimples!
I don't recall from previous threads whether the balls or the cups/cones
tend to have greater surface hardness. Anyone know?
David Griffith
Was the headset indexed when you retired it?
Or did it develop the problem while in storage - i.e. hung by front wheel?
dcg
carl...@comcast.net
Dear Jobst,
So where is the "front" in the picture of the whole bearing, which was
photographed at random?
Cheers,
Carl Fogel
carl...@comcast.net
Dear Tim,
Bearings dent some cups, at least when banged with a hammer, squashed
in a vise, or buzzed for a few seconds with an air-hammer arrangement.
Jim Beam dents, working far too hard with a hammer:
http://groups.google.com/group/rec.bicycles.tech/msg/4aaae9db81242823
My dents, using a vise:
http://groups.google.com/group/rec.bicycles.tech/msg/834d1eb4917c2b22
My dents, using an air hammer with straw mat and wood support:
http://groups.google.com/group/rec.bicycles.tech/msg/977a3d0074b5ab4f
Cheers,
Carl Fogel
* * Chas
<carl...@comcast.net> wrote in message
news:glloi2h2uum6jkvr6...@4ax.com...
> Here's a badly indexed headset race:
>
> http://img84.imageshack.us/my.php?image=250headsetdimplesve2.jpg
>
> Here's an enlargement of the dimples:
>
> http://img84.imageshack.us/my.php?image=250gheadsetdimplesrb9.jpg
<snip>
> Cheers,
>
> Carl Fogel
This bearing cup appears to be case hardened rather than through
hardened. Case hardening was/is used on a lot of low quality headsets
Case hardening creates a surface layer up to .050" (1.3mm) deep that has
a hardness of around 60Rc. The advantage of case hardening is that it
provides a wear resistant surface while maintaining a tough crack
resistant steel substrate. It's frequently used for gears and shafts
where there is a lot of sliding contact.
Bicycle headsets are thrust bearings that are not subjected to much
rotational or sliding friction. Forces are concentrated in very small
areas at the tops and bottoms of a few ball bearings providing an
excellent opportunity for brinelling.
Angular race needle bearings are far better suited for this kind of
load.
This cup appears to have failed due to initial brinelling breaking
through the hardened layer then the bulk of the damage could be
attributed to "false brinelling" or caused by fretting or spalling.
http://en.wikipedia.org/wiki/False_brinelling
http://www.emersonbearing.com/fail_falsebrinell.htm
http://www.lubrizol.com/GreaseReference/glossary.asp#false
http://www.lubrizol.com/GreaseReference/glossary.asp#fretting
I managed a pro bike shop in the mid 1970's. We got tired of having to
frequently replace the lower bearing races on Campy headsets due to
"brinelling". We took over 20 new and used pro headsets to a friend's
machine shop and checked them with a Rockwell Hardness Tester. This was
before top quality Japanese components were readily available.
Campy NR & SR headsets were some of the worst coming in at <48Rc
hardness. Zeus and Stronglight S5 Super Competition headsets were all
>60Rc and we saw least brinelling problems with those brands. Even with
this evidence, it was a hard sell trying to get "all Campy" bike owners
to change headset brands. Slaves to fashion - something akin to wearing
stylish but painfully uncomfortable Italian shoes!
Chas.
* * Chas
<carl...@comcast.net> wrote in message
news:c8qoi2hjbadjiafck...@4ax.com...
> On 11 Oct 2006 03:19:23 GMT, jobst....@stanfordalumni.org wrote:
>
> >Carl Fogel writes:
> >
> >> Here's a badly indexed headset race:
> >
> > http://img84.imageshack.us/my.php?image=250headsetdimplesve2.jpg
> >
> >> Here's an enlargement of the dimples:
> >
> > http://img84.imageshack.us/my.php?image=250gheadsetdimplesrb9.jpg
> >
> >> You may be able to enlarge either picture even further by clicking
> >> on it.
> >
> >> I don't know whether the magnified picture of the dimples supports
> >> erosion by fretting or impact by denting or both.
The indentations appear to be equally, widely spaced indicating that the
ball bearings were held in place with a bearing cage.
I've almost always assembled headsets with loose ball bearings which in
most cases can increase the number of balls thus spreading out the load
forces.
Chas.
jim beam
> In article <glloi2h2uum6jkvr6...@4ax.com>,
> carl...@comcast.net wrote:
>
>> Here's a badly indexed headset race:
>>
>> http://img84.imageshack.us/my.php?image=250headsetdimplesve2.jpg
>>
>> Here's an enlargement of the dimples:
>>
>> http://img84.imageshack.us/my.php?image=250gheadsetdimplesrb9.jpg
>
> <snip>
>
>> After putting the bike in a vise and disassembling things, I pounded
>> uselessly on the top of the stem for a while with a 4-lb hammer.
>
> Beating on the stem would not dimple the lower headset cup under any
> circumstances.
not true. i've brinelled headsets in crashes and with a hammer.
> It'd dimple the upper headset cup and/or cone, if it was
> going to dimple anything.
>
> The photos you provide look like an excellent example of bearing wear.
> The dimples are not round, as they would be if they were caused by
> impact, and the wear trails between the dimples are nearly as bad as the
> dimples!
>
> I don't recall from previous threads whether the balls or the cups/cones
> tend to have greater surface hardness. Anyone know?
should be about the same, but on cheap stuff, expect the bearing balls
to be a little harder. it's a simple quality issue.
jim beam
you are much emboldened by the world of digital photography carl!
i'm waiting for someone to show similar pics of sealed cartridge headset
bearings showing the same features. i suspect i may be waiting a while
since the seals keep the lube in and the grit out thus removing two
major failure factors.
jim beam
absolutely.
>
> http://en.wikipedia.org/wiki/False_brinelling
>
> http://www.emersonbearing.com/fail_falsebrinell.htm
>
> http://www.lubrizol.com/GreaseReference/glossary.asp#false
>
> http://www.lubrizol.com/GreaseReference/glossary.asp#fretting
>
> I managed a pro bike shop in the mid 1970's. We got tired of having to
> frequently replace the lower bearing races on Campy headsets due to
> "brinelling". We took over 20 new and used pro headsets to a friend's
> machine shop and checked them with a Rockwell Hardness Tester. This was
> before top quality Japanese components were readily available.
>
> Campy NR & SR headsets were some of the worst coming in at <48Rc
> hardness. Zeus and Stronglight S5 Super Competition headsets were all
>> 60Rc and we saw least brinelling problems with those brands. Even with
> this evidence, it was a hard sell trying to get "all Campy" bike owners
> to change headset brands. Slaves to fashion - something akin to wearing
> stylish but painfully uncomfortable Italian shoes!
>
> Chas.
>
>
>
awesome post - testing, comparison - excellent.
jobst....@stanfordalumni.org
Carl Fogel writes:
http://img84.imageshack.us/my.php?image=250headsetdimplesve2.jpg
http://img84.imageshack.us/my.php?image=250gheadsetdimplesrb9.jpg
> So where is the "front" in the picture of the whole bearing, which was
> photographed at random?
As I said, there where the dimples are most pronounced. With lighting
is not ideal for seeing them, it seems to me that straight ahead is at
11:30. You can see that better than I, hand held with good lighting.
Jobst Brandt
carl...@comcast.net
<spamv...@bad.example.net> wrote:
>Tim McNamara wrote:
>> In article <glloi2h2uum6jkvr6...@4ax.com>,
>> carl...@comcast.net wrote:
>>
>>> Here's a badly indexed headset race:
>>>
>>> http://img84.imageshack.us/my.php?image=250headsetdimplesve2.jpg
>>>
>>> Here's an enlargement of the dimples:
>>>
>>> http://img84.imageshack.us/my.php?image=250gheadsetdimplesrb9.jpg
>>
>> <snip>
>>
>>> After putting the bike in a vise and disassembling things, I pounded
>>> uselessly on the top of the stem for a while with a 4-lb hammer.
>>
>> Beating on the stem would not dimple the lower headset cup under any
>> circumstances.
>
>not true. i've brinelled headsets in crashes and with a hammer.
>
>> It'd dimple the upper headset cup and/or cone, if it was
>> going to dimple anything.
[snip]
Dear Jim,
I think that Tim's point is that whomping down on the the stem of a
bicycle frame held off the floor in vise shouldn't affect the lower
headset cup--the impact should force the bearing assembly apart, not
together.
The situation in the vise, where nothing supports the fork except the
corroded stem, leaving no pressure on the lower bearings, is different
from the normal situation, where the tire pressing on the ground can
compress the bearings.
Cheers,
Carl Fogel
carl...@comcast.net
Dear Jim,
There's something enchanting about scrolling around a 20" high image
of a headset bearing.
It gives me the hope that someone can point to the details.
I was tempted to experiment with the editor and see if I could sneak a
faint smiley face into one of the dimples, but my courage failed me.
Still, it's amazing what a modern $140 camera and computer can do,
even in the hands of the most naive amateur.
When I went to make a movie of the Park tension gauge stubbornly
deflating in synch with the tire, I had a vague notion that the camera
could only capture 3 seconds of motion and didn't realize that it
would have an audio track. It turned out that the camera captured the
sound of the air hissing out of the tire and would have happily
continued recording for twenty minutes.
Cheers,
Carl Fogel
jobst....@stanfordalumni.org
>> Here's a badly indexed headset race:
http://img84.imageshack.us/my.php?image=250headsetdimplesve2.jpg
>> Here's an enlargement of the dimples:
http://img84.imageshack.us/my.php?image=250gheadsetdimplesrb9.jpg
> This bearing cup appears to be case hardened rather than through
> hardened. Case hardening was/is used on a lot of low quality
> headsets
> Case hardening creates a surface layer up to .050" (1.3mm) deep that
> has a hardness of around 60Rc. The advantage of case hardening is
> that it provides a wear resistant surface while maintaining a tough
> crack resistant steel substrate. It's frequently used for gears and
> shafts where there is a lot of sliding contact.
> Bicycle headsets are thrust bearings that are not subjected to much
> rotational or sliding friction. Forces are concentrated in very
> small areas at the tops and bottoms of a few ball bearings providing
> an excellent opportunity for Brinelling.
> Angular race needle bearings are far better suited for this kind of
> load.
I have not found that to be true, having a needle bearing headset that
dimpled sooner than any 20 ball Campagnolo bearing. The problem is
fretting and that can't be solved by more contact such a s a roller
instead of a ball. These bearings, both roller and ball bearings work
best for rotating loads that keep a reasonable oil film between
rolling elements and their races.
A head bearing does not rotate sufficiently on a road bicycle when
costing down long descents where there are essentially no steering
motions while the fork vibrates fore and aft. It is the fore and aft
motion that causes fretting damage. Preloaded angular contact
cartridge bearings as used by Shimano, supported on a spherical plain
bearing that can tilt fore and aft, solve this problem.
> This cup appears to have failed due to initial Brinelling breaking
> through the hardened layer then the bulk of the damage could be
> attributed to "false Brinelling" or caused by fretting or spalling.
Make up your mind, was it fretting or spalling. They are definitely
not similar.
http://en.wikipedia.org/wiki/False_brinelling
This article beats around the bush and does not understand fretting
damage to rolling element bearings or in press fits, where it occurs
between shafts and sleeves mounted on them.
I think that is better explained in the FAQ item:
http://www.sheldonbrown.com/brandt/fretting.html
http://www.emersonbearing.com/fail_falsebrinell.htm
http://www.lubrizol.com/GreaseReference/glossary.asp#false
http://www.lubrizol.com/GreaseReference/glossary.asp#fretting
> I managed a pro bike shop in the mid 1970's. We got tired of having
> to frequently replace the lower bearing races on Campy headsets due
> to "Brinelling". We took over 20 new and used pro headsets to a
> friend's machine shop and checked them with a Rockwell Hardness
> Tester. This was before top quality Japanese components were
> readily available.
> Campy NR & SR headsets were some of the worst coming in at <48Rc
> hardness. Zeus and Stronglight S5 Super Competition headsets were
> all >60Rc and we saw least Brinelling problems with those brands.
> Even with this evidence, it was a hard sell trying to get "all
> Campy" bike owners to change headset brands. Slaves to fashion -
> something akin to wearing stylish but painfully uncomfortable
> Italian shoes!
BMW suffers from fretting the most among cars that I hear passing as I
ride. I was first exposed to this problem by an article from GM
Laboratories that investigate this phenomenon in the days (1950's)
when their cars were shipped farther and more securely restrained in
their RR cars than they are today. The failure of pinion gear
bearings in differentials, the one with the highest speed when
running, were fretting themselves to death in shipment. This damage
is audible as a rushing (hissing) sound coming from the rear end.
Since then I observed sport car transmissions with special ratios
gears pressed onto the main and countershafts that all had rouge in
their bores or bearing sleeves and on the shafts onto which they had
tight press fits.
Obviously a shaft under torque twists and that twist is not
transmitted ideally to the part pressed onto it. My first
introduction to this was before I understood metal elasticity as a
child watching failed Ford taper-cone rear axles being replaced. They
where they were pressed into the brake drum and retained with a large
castle nut. Later I saw that VW spline rear axles were always full of
rouge and chirped when used in reverse forcefully.
Jobst Brandt
jobst....@stanfordalumni.org
>>>> Here's a badly indexed headset race:
http://img84.imageshack.us/my.php?image=250headsetdimplesve2.jpg
>>>> Here's an enlargement of the dimples:
http://img84.imageshack.us/my.php?image=250gheadsetdimplesrb9.jpg
>>>> You may be able to enlarge either picture even further by
>>>> clicking on it.
>>>> I don't know whether the magnified picture of the dimples
>>>> supports erosion by fretting or impact by denting or both.
> <snip>
> The indentations appear to be equally, widely spaced indicating that
> the ball bearings were held in place with a bearing cage.
> I've almost always assembled headsets with loose ball bearings which
> in most cases can increase the number of balls thus spreading out
> the load forces.
How many more? Not to worry, they make their own dimples and home in
on them, especially the one in the front and rear. Fretting is the
problem rather than load distribution. It doesn't take a large load
for this to occur, so more balls is not a solution. You'll note that
light weight riders also had these problems.
Jobst Brandt
carl...@comcast.net
Emails have complained about the free image hosting site bogging down,
so here are the three original pictures at that site, plus duplicates
at another site (two have border material removed to squeeze under a
1mb upload limit).
The three originals:
Here's a badly indexed headset race:
http://img84.imageshack.us/my.php?image=250headsetdimplesve2.jpg
Here's an enlargement of the dimples:
http://img84.imageshack.us/my.php?image=250gheadsetdimplesrb9.jpg
Here's the headset dead-on:
http://img170.imageshack.us/my.php?image=253wholeheadset3za6.jpg
And here they are at another site:
Small angled view of most of the headset, trimmed to squeeze under a
1mb upload limit:
http://server5.theimagehosting.com/image.php?img=250headset%20dimplestrim.jpg
The three dimples from that picture, enlarged:
http://server5.theimagehosting.com/image.php?img=250gheadsetdimples.jpg
The headset dead-on, trimmed to squeeze under a 1mb upload limit:
http://server5.theimagehosting.com/image.php?img=253wholeheadset3trim.jpg
CF
jim beam
jobst, you're either sadly confused or deliberately trying to mislead.
fretting between the race and its housing is ABSOLUTELY NOT any damned
thing to do with the false or true brinelling evidenced between a
bearing ball and the race.
>
> I think that is better explained in the FAQ item:
>
> http://www.sheldonbrown.com/brandt/fretting.html
your faq's are /full/ of errors on a multitude of subjects. your
suppositions about bearings and their failure modes are prime examples.
remember the time you tried to b.s. about rolling element bearing
spalling and cited a journal bearing as an example?
>
> http://www.emersonbearing.com/fail_falsebrinell.htm
>
> http://www.lubrizol.com/GreaseReference/glossary.asp#false
>
> http://www.lubrizol.com/GreaseReference/glossary.asp#fretting
so what? both failure modes exist, but it's you that seems to have the
problem differentiating between them.
>
>> I managed a pro bike shop in the mid 1970's. We got tired of having
>> to frequently replace the lower bearing races on Campy headsets due
>> to "Brinelling". We took over 20 new and used pro headsets to a
>> friend's machine shop and checked them with a Rockwell Hardness
>> Tester. This was before top quality Japanese components were
>> readily available.
>
>> Campy NR & SR headsets were some of the worst coming in at <48Rc
>> hardness. Zeus and Stronglight S5 Super Competition headsets were
>> all >60Rc and we saw least Brinelling problems with those brands.
>> Even with this evidence, it was a hard sell trying to get "all
>> Campy" bike owners to change headset brands. Slaves to fashion -
>> something akin to wearing stylish but painfully uncomfortable
>> Italian shoes!
>
> BMW suffers from fretting the most among cars that I hear passing as I
> ride. I was first exposed to this problem by an article from GM
> Laboratories that investigate this phenomenon in the days (1950's)
> when their cars were shipped farther and more securely restrained in
> their RR cars than they are today. The failure of pinion gear
> bearings in differentials, the one with the highest speed when
> running, were fretting themselves to death in shipment. This damage
> is audible as a rushing (hissing) sound coming from the rear end.
rubbish. false brinelling was a problem in the 50's due to crap steel
and inferior lubricants. vehicles are shipped in exactly the same way
all over the globe half a century later, and there are no more bearing
failures. it doesn't say much for your credibility if you can't be
bothered to update your information. but then again, you wouldn't want
to have facts get in the way of a good criticism would you - your other
misplaced bleatings about anodized rims spring to mind.
>
> Since then I observed sport car transmissions with special ratios
> gears pressed onto the main and countershafts that all had rouge in
> their bores or bearing sleeves and on the shafts onto which they had
> tight press fits.
those are not bearings!!! sheesh. just because fretting occurs doesn't
mean it's responsible for the ills of the world, especially not your
apparent confusion between the noise made by a bad bearing and a car's
tire scrub.
>
> Obviously a shaft under torque twists and that twist is not
> transmitted ideally to the part pressed onto it. My first
> introduction to this was before I understood metal elasticity as a
> child watching failed Ford taper-cone rear axles being replaced. They
> where they were pressed into the brake drum and retained with a large
> castle nut. Later I saw that VW spline rear axles were always full of
> rouge and chirped when used in reverse forcefully.
so freakin' what? qr shafts fret inside their hollow axles too. it
makes no difference to their operation or longevity.
this post is typical jobstian nonsense. a small nugget of trivial truth
is the hook, but you follow it with supposition and error because you
can't be bothered to get current information.
and let's not overlook the fact that you're trying to trivialize the
posting of a guy that actually did hardness testing on bearing failures
- something you've never done. but then again, maybe we should be glad
you didn't do a dye penetrant test to determine hardness either.
jobst....@stanfordalumni.org
> The three originals:
http://img84.imageshack.us/my.php?image=250headsetdimplesve2.jpg
http://img84.imageshack.us/my.php?image=250gheadsetdimplesrb9.jpg
http://img170.imageshack.us/my.php?image=253wholeheadset3za6.jpg
http://server5.theimagehosting.com/image.php?img=250headset%20dimplestrim.jpg
http://server5.theimagehosting.com/image.php?img=250gheadsetdimples.jpg
http://server5.theimagehosting.com/image.php?img=253wholeheadset3trim.jpg
These picture reveal a few things about the nature of bearing wear
which in this case was caused some time ago. The wear tracks and
dimples are rusty in places, indicating that the bearing sat around
for enough time for the interface to rust. I don't know what the
balls did, but I suppose they also had rusty spots.
I take it that this was a low quality bearing in that is had a cage
that apparently held only 16 balls instead of the usual 20 balls
common for Campagnolo and Shimano bearings of this type. As I recall,
by removing the cage one could add two more balls raising the number
to 22.
Could you erase a couple of the above pix and replace them with ones
of the inner race including a few of its bearing balls for good
measure. That is, if anyone wants to see more of this non exemplary
quality bearing that probably doesn't represent the state of the art
of that time.
Don't you have a Campagnolo head bearing in your collection with
dimples? That would be more interesting.
Jobst Brandt
* * Chas
<jobst....@stanfordalumni.org> wrote in message
news:452d497a$0$34577$742e...@news.sonic.net...
It all depends on the hardness and quality of the steel used in the
races. Most comercial ball and needle bearing races are made of 52100
chrome alloy steel through hardened to >60Rc. Cheap case hardened
products are usually made from low carbon 1018 steel. Better quality
products use 8620, 9310 or similar alloy steels.
I too have seen a needle bearing headset race that was brinelled but it
was probably just case harened or not sufficiently through hardened.
World wide, most bicycle components are made from the cheapest materials
available using pre 1900 technology. Sealed through hardened bearings
are cheap and have been available for well over 100 years.
Crains, tank turrets, war ship turret, moveable bridges and turntables
all rotate on either flat or needle bearings which are all thrust
bearings.
>
> A head bearing does not rotate sufficiently on a road bicycle when
> costing down long descents where there are essentially no steering
> motions while the fork vibrates fore and aft. It is the fore and aft
> motion that causes fretting damage. Preloaded angular contact
> cartridge bearings as used by Shimano, supported on a spherical plain
> bearing that can tilt fore and aft, solve this problem.
>
> > This cup appears to have failed due to initial Brinelling breaking
> > through the hardened layer then the bulk of the damage could be
> > attributed to "false Brinelling" or caused by fretting or spalling.
>
> Make up your mind, was it fretting or spalling. They are definitely
> not similar.
Au contraire - they all similar failure mechanisms at the microscopic
level!
The damage could have been caused by all 4 failure mechanisms. It's hard
to tell the exact causes of the failure in a photo but it does appear to
have worn through the hardened surface of the bearing race indicating it
was case hardened rather than through hardened.
I listed the reference below because it's easy to understand and it has
some good links.
>
> http://en.wikipedia.org/wiki/False_brinelling
Here's a good link on Tribology - the science of wear which is what
we're discussing:
http://www.shef.ac.uk/mecheng/tribology/
>
> This article beats around the bush and does not understand fretting
> damage to rolling element bearings or in press fits, where it occurs
> between shafts and sleeves mounted on them.
>
Fretting damage is associated with a tight contact fit where there is
minimal motion between the parts.
Ball bearings are designed for minimal surface contact. Minor fretting
could have occured between the indented surfaces and the balls
especially if the headset has been over tightened to compenste for wear.
Chas. who?
jobst....@stanfordalumni.org
http://img84.imageshack.us/my.php?image=250headsetdimplesve2.jpg
http://img84.imageshack.us/my.php?image=250gheadsetdimplesrb9.jpg
> races. Most commercial ball and needle bearing races are made of
> 52100 chrome alloy steel through hardened to >60Rc. Cheap case
> hardened products are usually made from low carbon 1018
> steel. Better quality products use 8620, 9310 or similar alloy
> steels.
The needle bearing headsets were made as a response to fretting damage
because it was believed that it was Brinelling and increasing the
bearing contact area tenfold (rollers) would get rid of it. It did
not, and because fretting is not a function of load but of surface
contact in the absence of lubricant, it is also not solved by harder
steel.
> I too have seen a needle bearing headset race that was Brinelled but
> it was probably just case hardened or not sufficiently through
> hardened.
Hardness (yield strength) does not prevent fretting. Just consider
aluminum cranks on square steel tapers. They generate rouge in spite
of having been assembled with grease, which is just oil that is made
to nut run away.
> World wide, most bicycle components are made from the cheapest
> materials available using pre 1900 technology. Sealed through
> hardened bearings are cheap and have been available for well over
> 100 years.
That may be so but the heaviest loaded bearings in bicycles do not
dimple because they rotate. Typically cottered cranks, into which
cotters are hammered, a single 1/4" ball under the spindle taking the
shock, do not dimple. That is how good the steel of bearings were
long ago and no worse today for the reliable brands (Campagnolo).
> Cranes, tank turrets, war ship turret, movable bridges and turntables
> all rotate on either flat or needle bearings which are all thrust
> bearings.
They don't fret! They rotate. I think you'll find the individual
ball loading of a BB to be every bit as severe as the examples you
cite.
>> A head bearing does not rotate sufficiently on a road bicycle when
>> costing down long descents where there are essentially no steering
>> motions while the fork vibrates fore and aft. It is the fore and
>> aft motion that causes fretting damage. Preloaded angular contact
>> cartridge bearings as used by Shimano, supported on a spherical
>> plain bearing that can tilt fore and aft, solve this problem.
>>> This cup appears to have failed due to initial Brinelling breaking
>>> through the hardened layer then the bulk of the damage could be
>>> attributed to "false Brinelling" or caused by fretting or
>>> spalling.
>> Make up your mind, was it fretting or spalling. They are definitely
>> not similar.
> Au contraire - they all similar failure mechanisms at the
> microscopic level!
Spalling is sub-surface shear separating the hardened surface from the
substrate so that it flakes off. Fretting is spot welding by which
microscopic asperities are ripped out of the races leaving a matte
finish. That is how fresh fretting damage is distinguished from true
Brinelling, besides that the indentation is not spherical but
describes a dimple made by in-place rocking of bearing ball (or
roller).
> The damage could have been caused by all 4 failure mechanisms. It's
> hard to tell the exact causes of the failure in a photo but it does
> appear to have worn through the hardened surface of the bearing race
> indicating it was case hardened rather than through hardened.
Because this is an old failure, subsequent rotational wear obliterated
fretting failure. Just the same, the wear trough looks almost like
plastic deformation except that there is no upset material adjacent to
the wear groove.
> I listed the reference below because it's easy to understand and it has
> some good links.
http://en.wikipedia.org/wiki/False_brinelling
I think this one is a bit fuzzy and allows various interpretations.
It is not written by a tribologist from what I see.
> Here's a good link on Tribology - the science of wear which is what
> we're discussing:
http://www.shef.ac.uk/mecheng/tribology/
That doesn't do us much good here. I see no text with pictures
describing fretting damage. In that line, until I wrote the item
for the FAQ, the term fretting was not known here in wreck.bike and
it was challenged as something I made up.
http://www.sheldonbrown.com/brandt/fretting.html
I think this states in plain English what occurs.
>> This article beats around the bush and does not understand fretting
>> damage to rolling element bearings or in press fits, where it occurs
>> between shafts and sleeves mounted on them.
This was my comment to Wiki.
> Fretting damage is associated with a tight contact fit where there is
> minimal motion between the parts.
Well? I think I hear echos!
> Ball bearings are designed for minimal surface contact. Minor
> fretting could have occurred between the indented surfaces and the
> balls especially if the headset has been over tightened to compensate
> for wear.
>> I think that is better explained in the FAQ item:
http://www.sheldonbrown.com/brandt/fretting.html
>> http://www.emersonbearing.com/fail_falsebrinell.htm
>> http://www.lubrizol.com/GreaseReference/glossary.asp#false
>> http://www.lubrizol.com/GreaseReference/glossary.asp#fretting
>>> I managed a pro bike shop in the mid 1970's. We got tired of
>>> having to frequently replace the lower bearing races on Campy
>>> headsets due to "Brinelling". We took over 20 new and used pro
>>> headsets to a friend's machine shop and checked them with a
>>> Rockwell Hardness Tester. This was before top quality Japanese
>>> components were readily available.
>>> Campy NR & SR headsets were some of the worst coming in at <48Rc
>>> hardness. Zeus and Stronglight S5 Super Competition headsets were
>>> all >60Rc and we saw least Brinelling problems with those brands.
>>> Even with this evidence, it was a hard sell trying to get "all
>>> Campy" bike owners to change headset brands. Slaves to fashion -
>>> something akin to wearing stylish but painfully uncomfortable
>>> Italian shoes!
Fretting is not directly related to hardness but rather to lubrication
failure, especially of similar metals. You cite texts, that I think
you should peruse to find the difference between spalling and fretting
and their relation to yield stress. Fretting can occur at low stress
as soon as lubricant is dispersed from the interface.
Lubricant is dispersed by rocking motion in which it cannot
replenished for lack of clearance generated during the motion. That
is, fretting takes place within the Hertzian contact patch into which
oil cannot penetrate. That area is essentially a press fit of graded
stress from the center of the contact to the periphery where it
decreases to zero. This causes relative micro-motion between race and
ball/roller in that zone.
>> BMW suffers from fretting the most among cars that I hear passing
>> as I ride. I was first exposed to this problem by an article from
>> GM Laboratories that investigate this phenomenon in the days
>> (1950's) when their cars were shipped farther and more securely
>> restrained in their RR cars than they are today. The failure of
>> pinion gear bearings in differentials, the one with the highest
>> speed when running, were fretting themselves to death in shipment.
>> This damage is audible as a rushing (hissing) sound coming from the
>> rear end.
>> Since then I observed sport car transmissions with special ratios
>> gears pressed onto the main and counter shafts that all had rouge
>> in their bores or bearing sleeves and on the shafts onto which they
>> had tight press fits.
>> Obviously a shaft under torque twists and that twist is not
>> transmitted ideally to the part pressed onto it. My first
>> introduction to this was before I understood metal elasticity as a
>> child watching failed Ford taper-cone rear axles being replaced.
>> They where they were pressed into the brake drum and retained with
>> a large castle nut. Later I saw that VW spline rear axles were
>> always full of rouge and chirped when used in reverse forcefully.
> Chas. who?
I see, you are afraid of taking off you mask of anonymity.
Jobst Brandt
jim beam
but it /does/ alleviate true brinelling and that's why it's used! true
brinelling /does/ occur on bikes. that's why headsets are indexed after
crashes and heavy impacts - i know because i've experienced it. /and/
i've had access to metallurgy labs so have been able to analyze properly.
>
>> I too have seen a needle bearing headset race that was Brinelled but
>> it was probably just case hardened or not sufficiently through
>> hardened.
>
> Hardness (yield strength) does not prevent fretting.
true.
> Just consider
> aluminum cranks on square steel tapers. They generate rouge in spite
> of having been assembled with grease, which is just oil that is made
> to nut run away.
utterly amazingly irrelevant. that you would cite the existence of
fretting as proof of bearing failure is beyond belief. you really don't
understand this stuff at all.
>
>> World wide, most bicycle components are made from the cheapest
>> materials available using pre 1900 technology. Sealed through
>> hardened bearings are cheap and have been available for well over
>> 100 years.
>
> That may be so but the heaviest loaded bearings in bicycles do not
> dimple because they rotate. Typically cottered cranks, into which
> cotters are hammered, a single 1/4" ball under the spindle taking the
> shock, do not dimple.
they're not hammered, they're drawn by their fastening bolt.
> That is how good the steel of bearings were
> long ago and no worse today for the reliable brands (Campagnolo).
>
>> Cranes, tank turrets, war ship turret, movable bridges and turntables
>> all rotate on either flat or needle bearings which are all thrust
>> bearings.
>
> They don't fret! They rotate.
headsets don't rotate? hate to bring you bad news there cowboy, but the
slightest degree of headset movement, one far less than can possibly be
achieved by human hand, is more than sufficient to replenish contact
lubrication.
> I think you'll find the individual
> ball loading of a BB to be every bit as severe as the examples you
> cite.
why not be correct rather than speculate like you normally do? do the math.
>
>>> A head bearing does not rotate sufficiently on a road bicycle when
>>> costing down long descents where there are essentially no steering
>>> motions while the fork vibrates fore and aft. It is the fore and
>>> aft motion that causes fretting damage. Preloaded angular contact
>>> cartridge bearings as used by Shimano, supported on a spherical
>>> plain bearing that can tilt fore and aft, solve this problem.
>
>>>> This cup appears to have failed due to initial Brinelling breaking
>>>> through the hardened layer then the bulk of the damage could be
>>>> attributed to "false Brinelling" or caused by fretting or
>>>> spalling.
>
>>> Make up your mind, was it fretting or spalling. They are definitely
>>> not similar.
>
>> Au contraire - they all similar failure mechanisms at the
>> microscopic level!
>
> Spalling is sub-surface shear separating the hardened surface from the
> substrate so that it flakes off.
no, spalling is fatigue. do some electron microscopy some time. i have.
> Fretting is spot welding by which
> microscopic asperities are ripped out of the races leaving a matte
> finish.
that is true, but the extent is determined by steel quality, surface
finish [to some extent] and lubrication quality.
> That is how fresh fretting damage is distinguished from true
> Brinelling,
er, true brinelling surfaces are not necessarily any different. the
easy way to tell them apart is if the service history is known. the
scientific way is to section the pieces and check for plastic deformation.
> besides that the indentation is not spherical but
> describes a dimple made by in-place rocking of bearing ball (or
> roller).
er, we're looking at multiple factors here. or at least the rest of us
are. it's not clear what /you/ are looking at.
>
>> The damage could have been caused by all 4 failure mechanisms. It's
>> hard to tell the exact causes of the failure in a photo but it does
>> appear to have worn through the hardened surface of the bearing race
>> indicating it was case hardened rather than through hardened.
>
> Because this is an old failure, subsequent rotational wear obliterated
> fretting failure.
it can just as easily be asserted that wear obliterated true brinelling.
> Just the same, the wear trough looks almost like
> plastic deformation except that there is no upset material adjacent to
> the wear groove.
there is almost certainly plastic deformation in this instance.
>
>> I listed the reference below because it's easy to understand and it has
>> some good links.
>
> http://en.wikipedia.org/wiki/False_brinelling
>
> I think this one is a bit fuzzy and allows various interpretations.
> It is not written by a tribologist from what I see.
so go ahead and "correct" it. see how long your explanation stands in a
forum that allows editing. your "faq's" do not.
>
>> Here's a good link on Tribology - the science of wear which is what
>> we're discussing:
>
> http://www.shef.ac.uk/mecheng/tribology/
>
> That doesn't do us much good here. I see no text with pictures
> describing fretting damage. In that line, until I wrote the item
> for the FAQ, the term fretting was not known here in wreck.bike and
> it was challenged as something I made up.
it largely /is/ something you made up. assertions that headsets do not
true brinell are incorrect. assumption that false brinelling is the
/only/ failure mechanism is also incorrect. you may believe what you
say, but the fact that you don't believe what others say doesn't make
you correct or make others wrong. particularly in the context of
someone evidencing huge conceptual misunderstandings, [fracture,
fatigue, deformation, "strength", etc.], then doesn't have the wit or
the gonads to correct mistakes when shown.
>
> http://www.sheldonbrown.com/brandt/fretting.html
>
> I think this states in plain English what occurs.
you "think", but you won't correct when wrong.
>
>>> This article beats around the bush and does not understand fretting
>>> damage to rolling element bearings or in press fits, where it occurs
>>> between shafts and sleeves mounted on them.
>
> This was my comment to Wiki.
>
>> Fretting damage is associated with a tight contact fit where there is
>> minimal motion between the parts.
>
> Well? I think I hear echos!
/your/ examples are out of context - evidence that you don't understand
what you're talking about. or that you're trying to deceive.
jobst, again, spalling is fatigue. fretting is a type of wear. you're
still not evidencing you have any idea how either are related to yield
since you say bearings shatter before they deform.
>
> Lubricant is dispersed by rocking motion in which it cannot
> replenished for lack of clearance generated during the motion. That
> is, fretting takes place within the Hertzian contact patch into which
> oil cannot penetrate. That area is essentially a press fit of graded
> stress from the center of the contact to the periphery where it
> decreases to zero. This causes relative micro-motion between race and
> ball/roller in that zone.
at the microscopic absolute contact point level, that may be true, but
the truth is, there's no way a bike headset bearing can be kept
sufficiently still for that to ever be an issue. improved lubricants
also take a substantial bite out of that 50's bug-a-boo, as do improved
steels. you're fudging to make it look like you know what you're saying.
>
>>> BMW suffers from fretting the most among cars that I hear passing
>>> as I ride. I was first exposed to this problem by an article from
>>> GM Laboratories that investigate this phenomenon in the days
>>> (1950's) when their cars were shipped farther and more securely
>>> restrained in their RR cars than they are today. The failure of
>>> pinion gear bearings in differentials, the one with the highest
>>> speed when running, were fretting themselves to death in shipment.
>>> This damage is audible as a rushing (hissing) sound coming from the
>>> rear end.
>
>>> Since then I observed sport car transmissions with special ratios
>>> gears pressed onto the main and counter shafts that all had rouge
>>> in their bores or bearing sleeves and on the shafts onto which they
>>> had tight press fits.
>
>>> Obviously a shaft under torque twists and that twist is not
>>> transmitted ideally to the part pressed onto it. My first
>>> introduction to this was before I understood metal elasticity as a
>>> child watching failed Ford taper-cone rear axles being replaced.
>>> They where they were pressed into the brake drum and retained with
>>> a large castle nut. Later I saw that VW spline rear axles were
>>> always full of rouge and chirped when used in reverse forcefully.
>
>> Chas. who?
>
> I see, you are afraid of taking off you mask of anonymity.
ad hominem as an attempt to discredit is weak. reliance on the credit
of others [jobst....@stanfordalumni.org] is disgraceful.
carl...@comcast.net
Dear Jobst,
I'm afraid that your comment about rust may be a--
(sorry, can't resist it)
--genuine red herring.
The long, curved reddish areas are just an artifact of the lighting.
In the dead-on picture, the reddened sections along the outer curve
are pretty much matched by simmilar reddened sections along the inner
curve.
Also notice in the dead-on picture that shorter yellow sections on the
outer curve are again matched by yellow sections on the inner curve.
In the picture of the tilted headset, the same reddish curves can be
noticed.
The camera has a red light and a flash, which I ignore, since I know
nothing about how the camera works. It works feverishly to make up for
my ignorance of lighting and exposure and agrees to let me focus
manually up close only with the greatest reluctance--no zooming
allowed.
The headset cup is either shiny or darkened, not rusty. Most of the
tiny black spots come off with my fingernail. Possibly one or two of
them are dark rust instead of grime.
I've tilted and peered at the headset to check this, but I don't know
how to take a picture that won't have some red reflecting off it.
(Possibly the red-eye remover feature of the software would work?)
As for your other requests, I'm puzzled. I don't recall any previous
comments about dimples that claimed that the balls or the brand
mattered.
What will a picture of 16 ball bearings tell us about the dimples? And
what's special about a Campagnolo cup in respect to dimpling?
Cheers,
Carl Fogel
jobst....@stanfordalumni.org
>>> The three originals:
http://img84.imageshack.us/my.php?image=250headsetdimplesve2.jpg
http://img84.imageshack.us/my.php?image=250gheadsetdimplesrb9.jpg
http://img170.imageshack.us/my.php?image=253wholeheadset3za6.jpg
http://server5.theimagehosting.com/image.php?img=250headset%20dimplestrim.jpg
http://server5.theimagehosting.com/image.php?img=250gheadsetdimples.jpg
http://server5.theimagehosting.com/image.php?img=253wholeheadset3trim.jpg
> I'm afraid that your comment about rust may be a genuine red
> herring.
> The long, curved reddish areas are just an artifact of the lighting.
I'm not interested in the red stuff but the dark nearly black material
along the wear track. That isn't steel, but an oxide discolored by
lubricant tar. It is these traces that I called rust, and I'm sure
they are to a large part made of rust.
> In the dead-on picture, the reddened sections along the outer curve
> are pretty much matched by similar reddened sections along the
> inner curve.
Forget the red stuff.
> Also notice in the dead-on picture that shorter yellow sections on
> the outer curve are again matched by yellow sections on the inner
> curve.
> In the picture of the tilted headset, the same reddish curves can be
> noticed.
> The camera has a red light and a flash, which I ignore, since I know
> nothing about how the camera works. It works feverishly to make up
> for my ignorance of lighting and exposure and agrees to let me focus
> manually up close only with the greatest reluctance--no zooming
> allowed.
> The headset cup is either shiny or darkened, not rusty. Most of the
> tiny black spots come off with my fingernail. Possibly one or two
> of them are dark rust instead of grime.
Well let's call it darkened, but it is wear debris in the wear track.
> I've tilted and peered at the headset to check this, but I don't
> know how to take a picture that won't have some red reflecting off
> it.
> (Possibly the red-eye remover feature of the software would work?)
> As for your other requests, I'm puzzled. I don't recall any
> previous comments about dimples that claimed that the balls or the
> brand mattered.
I have no questions about that, only whether the balls also have this
dark wear debris on them.
> What will a picture of 16 ball bearings tell us about the dimples?
> And what's special about a Campagnolo cup in respect to dimpling?
As I said, the race looks like it had rust tracks and I wanted to see
the inner race and a few typical balls to see if they also showed
these signs. I don't need to see all of them, only some typical ones.
Jobst Brandt
jim beam
one more thing our esteemed alumni mechanical engineering expert seems
to be missing: that bearing is chrome plated. chrome is brittle and
flakes, particularly when the surface is crazed by say, speaking
hypothetically of course, a passing loaded bearing ball. just look at
the shiny plated areas vs. the colors of the bearing track and its
environs, particularly the transition between the two - the nickel
substrate is clearly exposed.
bottom line carl, it matters not what the exact mechanism is - all that
matters is ease of repair, economics of repair and prevention for the
future. replacing with another bearing of comparable quality will give
comparable results. replacement with a superior bearing, particularly
if of the sealed cartridge variety, will give superior results. unless
you're masquerading as an expert, failing, and somehow desperately
trying to save face with bizarre red herrings and baseless conjecture,
you shouldn't be worried about it either.
jta...@nospam.hfx.andara.com
<spamv...@bad.example.net> wrote:
>>
>> That may be so but the heaviest loaded bearings in bicycles do not
>> dimple because they rotate. Typically cottered cranks, into which
>> cotters are hammered, a single 1/4" ball under the spindle taking the
>> shock, do not dimple.
>
>they're not hammered, they're drawn by their fastening bolt.
>
Wrong again, jimbo.
"hammer or press" - and most of us used a hammer.
* * Chas
<jta...@NOSPAM.hfx.andara.com> wrote in message
news:8pvri2pe76b74v6a2...@4ax.com...
No, a BFH! ;-)
We had a VAR cotter press tool at our shop in the 70's. It was about the
size and configuration of a 24" pair of bolt cutters and worked great.
Chas.
Jasper Janssen
wrote:
>Tim McNamara wrote:
>> Beating on the stem would not dimple the lower headset cup under any
>> circumstances.
>
>not true. i've brinelled headsets in crashes and with a hammer.
But did you do do it to the *lower* headset bearing by beating on the
*stem*? Beating on the stem should only impact the *upper* headset.
Jasper
Jasper Janssen
>Charles who? writes:
>> Cranes, tank turrets, war ship turret, movable bridges and turntables
>> all rotate on either flat or needle bearings which are all thrust
>> bearings.
>
>They don't fret! They rotate. I think you'll find the individual
>ball loading of a BB to be every bit as severe as the examples you
>cite.
A warship or a tank turret typically stays still an awful long time while
it's being transported, either under its own power or on trucks/railroad
cars. That should be an analogous situation, at the very least to those
cars being transported while stationary on a railroad or road car.
Jasper
* * Chas
"Jasper Janssen" <jas...@jjanssen.org> wrote in message
news:u19ti21oeujsiob73...@4ax.com...
Yes and they use huge roller bearings along with some ball bearings. The
forces created when they fire their weapons are generally transmitted to
the hull or chassis rather than directly into the bearings but they are
somewhat analogous to what is happening to the bearings and races of a
bike headset as it goes over a rough road or trail.
Chas. Who?
jobst....@stanfordalumni.org
>>> Cranes, tank turrets, war ship turret, movable bridges and
>>> turntables all rotate on either flat or needle bearings which are
>>> all thrust bearings.
>> They don't fret! They rotate. I think you'll find the individual
>> ball loading of a BB to be every bit as severe as the examples you
>> cite.
> A warship or a tank turret typically stays still an awful long time
> while it's being transported, either under its own power or on
> trucks/railroad cars. That should be an analogous situation, at the
> very least to those cars being transported while stationary on a
> railroad or road car.
I'm not familiar with shipping such assemblies. The only large
assemblies I shipped were at least as heavy as gun turrets, but the
bearings were assembled at the installation. These were support
structures with bearings for 100+ ton magnets at SLAC (Stanford Linear
Accelerator Center). I can't imagine shipping a rotation drawbridge
with its bearings assembled.
Other items mentioned can be locked in place to prevent motion enough
to not have fretting damage, either because there is no significant
motion or the loads are insufficient to displace bearing lubricant.
That would be the case with shipping bicycles, for instance.
Jobst Brandt
Tim McNamara
jobst....@stanfordalumni.org wrote:
Pro bike mechanics have reported headset damage to bikes that have never
been ridden, but have spent hundreds of miles on roof racks that clamp
the fork through the dropouts. There's no impact to blame at all, just
the vibrations transmitted through the car.
jobst....@stanfordalumni.org
I was thinking more of bicycles assembled in boxes not exposed to wind
and vibration that are typical of outdoor racks on motor vehicles. I
don't doubt that in the least. Even up-side down with the front wheel
into the wind, lengthwise on the roof of a car could cause such
motions.
Jobst Brandt
Donald Gillies
>As I said, the deepest dimples are generally centered about the front
>of head set races when the bicycle is steering straight ahead.
>Therefore, when photographing it you could put that position at the
>top of the picture so that 12:00 is straight up and is also the front
>of the bearing.
This fact demonstrates nothing at all, other than the fact that
bicycles are steering straight ahead 99.9% of the time.
>If this were road shock induced rather than fretting the dimples would
>be uniformly distributed (assuming the head tube had square ends). As
>should be obvious, the forward rim of the bearing bears most of the
>wheel loads and wears accordingly. The cone opposite the cup shown
>will have matching marks, leaving no doubt about in use orientation.
Again, this argument makes no sense. If the flexion in the headset
was from road shock, the balls on the side would move parallel to the
cones / plane of turning, whereas the ones front and back would move
up and down the cones, which to me implies that wear would most likely
occur only in the front and back of headset cones.
>This is more evident on a better quality bearing and if the bicycle is
>used with heavy braking will have dimples in the front AND rear
>quadrant. My dimpled bearings had nearly symmetrical wear in that
>manner, while the side quadrants showed almost no visible wear.
In my experience (20 bikes sample space), long head-tubes and
lighter-weight steels (i.e. head tube flexion), and heavier riders -
greatly contribute to headset wear - assuming the headset is properly
adjusted.
Coasting has zero impact on headset wear, imho.
- Don Gillies
San Diego, CA
jobst....@stanfordalumni.org
>> As I said, the deepest dimples are generally centered about the
>> front of head set races when the bicycle is steering straight
>> ahead. Therefore, when photographing it you could put that
>> position at the top of the picture so that 12:00 is straight up and
>> is also the front of the bearing.
> This fact demonstrates nothing at all, other than the fact that
> bicycles are steering straight ahead 99.9% of the time.
I think you don't see the problem. When steering, even straight
ahead, normal motions are large enough to move bearing balls in their
races farther than their contact area, thus replenishing lubrication
(the lubrication is oil contained in grease).
I don't see how you can be so sure of what you say and not explain why
you believe so. It rings of idle argumentation, but I could be wrong.
So give that some more thought. With the load being primarily on the
forward quadrant of the bearing and the fork vibrating fore and aft,
bending at the base of the steer tube, the greatest load and movement
occur at the forward sector of the bearing. After all, it is a load
and micro-motion related effect.
>> If this were road shock induced rather than fretting the dimples
>> would be uniformly distributed (assuming the head tube had square
>> ends). As should be obvious, the forward rim of the bearing bears
>> most of the wheel loads and wears accordingly. The cone opposite
>> the cup shown will have matching marks, leaving no doubt about in
>> use orientation.
> Again, this argument makes no sense. If the flexing in the headset
> was from road shock, the balls on the side would move parallel to the
> cones / plane of turning, whereas the ones front and back would move
> up and down the cones, which to me implies that wear would most likely
> occur only in the front and back of headset cones.
That is the way it occurs. That is why I mentioned that dimples are
deepest in front and slightly less in the rear, fading to barely
visible at the sides where motion and load are small compared to the
other two zones, having no significant load during these excursions.
In any event, even if this were all imagined, the change in design
that countered these effects, works. Radially pre-loaded cartridge
head-bearings on a spherically swiveling base has essentially stopped
this sort of wear in Shimano bearings. You don't think they went to
that trouble if that wasn't a problem, one that was obvious to other
observers.
>> This is more evident on a better quality bearing and if the bicycle
>> is used with heavy braking will have dimples in the front AND rear
>> quadrant. My dimpled bearings had nearly symmetrical wear in that
>> manner, while the side quadrants showed almost no visible wear.
> In my experience (20 bikes sample space), long head-tubes and
> lighter-weight steels (i.e. head tube flexion), and heavier riders -
> greatly contribute to headset wear - assuming the headset is
> properly adjusted.
> Coasting has zero impact on headset wear, imho.
Can you explain why you believe that. Coasting is the only time head
bearings do not make steering motions large enough to replenish
lubricant and occurs while there is significant vibratory action fore
and aft in the fork, a direction for which the bearing is not
designed. It's too bad that the Engineering article I read about
damaged wheel and differential bearings at GM was so long ago.
However, this is as close as I can find right now:
Jobst Brandt
jim beam
it's not a question of whether you read a document 50 years ago or
whether bearings can be made to fret today. it's a question of whether
fretting is the only headset damage mechanism for bikes, and the fact
is, it's not.
> However, this is as close as I can find right now:
>
> http://tinyurl.com/v84ab
fretting exists. that's not the mystery. what's the mystery is how you
managed to analyze these bearings in such a way as think you can
definitively rule out one of the primary damage mechanisms, that of true
brinelling. since i have had such bearings analyzed at a lab and have
determined that true brinelling does indeed occur, i have to conclude
that whatever mystery method you used in your own analysis was either
incomplete, or just plain wrong. personally, i don't believe you ever
did any actual analysis, you merely made supposition. just like some of
your other amazing but incorrect suppositions about anodizing and
fatigue, spoke squeezing and fatigue elimination in stainless steel,
etc. etc.
Johnny Sunset aka Tom Sherman
"jim beam" wrote:
> ...
> one more thing our esteemed alumni mechanical engineering expert seems
Did your application to Stanford come back marked "rejected", jim beam?
--
Tom Sherman - Here, not there.
jim beam
> "jim beam" wrote:
>> ...
>> one more thing our esteemed alumni mechanical engineering expert seems
>> to be missing....
>
> Did your application to Stanford come back marked "rejected", jim beam?
>
right - it wasn't a mere mere sub-department of the engineering faculty.
* * Chas
<carl...@comcast.net> wrote in message
news:glloi2h2uum6jkvr6...@4ax.com...
> Here's a badly indexed headset race:
>
> http://img84.imageshack.us/my.php?image=250headsetdimplesve2.jpg
>
> Here's an enlargement of the dimples:
>
> http://img84.imageshack.us/my.php?image=250gheadsetdimplesrb9.jpg
>
> it.
>
> I don't know whether the magnified picture of the dimples supports
> erosion by fretting or impact by denting or both.
>
> visible before on RBT.
>
> The headset belonged to my faithful Schwinn LeTour '98. It spent about
> two miles per day coasting down a 65 mph highway on 120 psi 700x26
> tires at 30-50 mph, depending on the wind. Occasionally, it drifted
> over the rumble strips and buzzed like a jackhammer. Its other daily
> 13 miles were at about 20 mph over somewhat cracked and bumpy
> pavement. No impacts or accidents were severe enough to bend the fork
> back.
>
> After years of cruel neglect, my faithful Schwinn was retired and
> replaced by its twin brother (easier than maintenance).
>
> While fiddling with my camera this evening, I wondered about headset
> indexing and remembered what a shameful example I had close at hand.
>
> After putting the bike in a vise and disassembling things, I pounded
> uselessly on the top of the stem for a while with a 4-lb hammer.
>
> Then I gave up and took more drastic measures that resembled a
> partial-birth abortion. By sawing through the exposed but hopelessly
> frozen stem, I extracted the stem and the intact lower race without
> pounding on it.
>
> Cheers,
>
> Carl Fogel
I picked a used late 80's Colnago steel fork and frame from a local
frame builder. The alignment of the frame and fork were checked and
adjusted on a surface plate. The head tube and fork crown were in good
shape and didn't need to be faced. I installed a Campy Chorus alloy
headset. The races were pressed in and on with Campy headset tools. I
adjusted it perfectly with a slight preload and re-adjusted it after
several short test rides.
I took the bike out for it's maiden ride today and after an hour or so
found that the headset had become slightly loose and needs to be
re-adjusted again (1/16 of a turn or less on the top headset race).
It occurred to me that a slightly loose headset could cause premature
damage to the bearing races. A well adjusted headset in an accurately
set up frame with a slight bearing preload should distribute impact
loads throughout all of the ball bearings preventing this kind of
failure.
A slightly loose headset will allow movement during normal road
vibration and impact shock. This can easily lead to indentations in the
bearing surfaces especially in the front of the lower headset and crown
races.
My guess without tearing the headset apart is that the steel races are
seating themselves into the aluminum cups resulting in a loose headset.
When this Crampy Chorus headset starts to develop indentations, I'll
replace it with my last Stronglight S5 Super Competition steel headset.
I used to use these on cyclocross and MTB bikes that I built and never
saw a failure.
Chas.
jobst....@stanfordalumni.org
> I picked a used late 80's Colnago steel fork and frame from a local
> frame builder. The alignment of the frame and fork were checked and
> adjusted on a surface plate. The head tube and fork crown were in
> good shape and didn't need to be faced. I installed a Campy Chorus
> alloy headset. The races were pressed in and on with Campy headset
> tools. I adjusted it perfectly with a slight preload and
> re-adjusted it after several short test rides.
> I took the bike out for it's maiden ride today and after an hour or
> so found that the headset had become slightly loose and needs to be
> re-adjusted again (1/16 of a turn or less on the top headset race).
> It occurred to me that a slightly loose headset could cause
> premature damage to the bearing races. A well adjusted headset in
> an accurately set up frame with a slight bearing preload should
> distribute impact loads throughout all of the ball bearings
> preventing this kind of failure.
Yours is the classic analysis of head bearing loads but it is
incorrect in several ways. Any clunk sound from bearing clearance
comes from the upper bearing, it being the loose one, while the lower
bearing remains in contact from the load of the bicycle and makes no
complaints. The other is that the lower bearing is receiving road
shocks through the cushion of a pneumatic tire and cannot be damaged
by riding in this condition.
Head bearing failures or "indexing" is caused by fretting damage as is
outlined in:
http://www.sheldonbrown.com/brandt/indexed-steering.html
http://www.sheldonbrown.com/brandt/fretting.html
To visualize the effect on the bearing, assume you had a new and
undamaged head bearing, turned the bicycle upside down (in a repair
stand that supports the bicycle frame) and hammered on the tire. Can
you imagine producing dimples in the bearing from this action? I am
certain that you cannot cause dents in a head bearing race, even
slight ones, by this method.
Meanwhile those of us who recall cottered cranks know how they were
routinely installed by hammering the cotters into the cranks that were
supported by one or at most two 1/4" bearing balls. These spindles
did not sustain (Brinelling) dimples from those sharp and uncushioned
hammer blows. The nut on the cotter served to keep it from working
its way loose. It could not be used to seat the cotter.
> A slightly loose headset will allow movement during normal road
> vibration and impact shock. This can easily lead to indentations in
> the bearing surfaces especially in the front of the lower headset
> and crown races.
I think you are repeating myth and lore with no evidence that this
story has any merit. Meanwhile I had upper head bearings that had
fretting dimples to verify that it is caused by fretting.
> My guess without tearing the headset apart is that the steel races
> are seating themselves into the aluminum cups resulting in a loose
> headset.
I don't know what caused your extra clearance but it is not from
Brinelling as one is lead to believe from pseudo science. I would
look for a not fully seated lower inner race on the fork crown. That
is where most rapid clearance growth after installing a new bearing
have arisen in my experience.
> When this Crampy Chorus headset starts to develop indentations, I'll
> replace it with my last Stronglight S5 Super Competition steel
> headset. I used to use these on cyclocross and MTB bikes that I
> built and never saw a failure.
You are looking for help in the wrong place. You need a design change
and that is offered by Shimano and Cane Creek, among others.
Jobst Brandt
jim beam
JOBST, YOU'RE MISSING THE VITAL AND VERY IMPORTANT POINT THAT BEARINGS
CAN BOTH TRUE AND FALSE BRINELL. CRASHES AND IMPACT CAN LOAD BEARINGS
SUFFICIENTLY TO TRUE BRINELL THEM. I KNOW THIS FROM EXPERIENCE AND
ANALYSIS.
citing yourself and your own omission does not and will not ever change
this fact. stop being so stubborn and ignorant and update your faq's.
if you can't/won't/are too lazy to update them, have them posted to a
wiki where others can do your work for you. sheesh.
>
> To visualize the effect on the bearing, assume you had a new and
> undamaged head bearing, turned the bicycle upside down (in a repair
> stand that supports the bicycle frame) and hammered on the tire. Can
> you imagine producing dimples in the bearing from this action? I am
> certain that you cannot cause dents in a head bearing race, even
> slight ones, by this method.
imagination vs. fact. that's /so/ convincing.
>
> Meanwhile those of us who recall cottered cranks know how they were
> routinely installed by hammering the cotters into the cranks that were
> supported by one or at most two 1/4" bearing balls. These spindles
> did not sustain (Brinelling) dimples from those sharp and uncushioned
> hammer blows. The nut on the cotter served to keep it from working
> its way loose. It could not be used to seat the cotter.
that's b.s.
>
>> A slightly loose headset will allow movement during normal road
>> vibration and impact shock. This can easily lead to indentations in
>> the bearing surfaces especially in the front of the lower headset
>> and crown races.
>
> I think you are repeating myth and lore with no evidence that this
> story has any merit. Meanwhile I had upper head bearings that had
> fretting dimples to verify that it is caused by fretting.
you can't differentiate by sight between true and false brinelling any
better than you can use a dye penetrant test to determine that anodizing
causes fatigue or that spoke squeezing eliminates fatigue in stainless
steel.
>
>> My guess without tearing the headset apart is that the steel races
>> are seating themselves into the aluminum cups resulting in a loose
>> headset.
>
> I don't know what caused your extra clearance but it is not from
> Brinelling as one is lead to believe from pseudo science. I would
> look for a not fully seated lower inner race on the fork crown. That
> is where most rapid clearance growth after installing a new bearing
> have arisen in my experience.
>
>> When this Crampy Chorus headset starts to develop indentations, I'll
>> replace it with my last Stronglight S5 Super Competition steel
>> headset. I used to use these on cyclocross and MTB bikes that I
>> built and never saw a failure.
>
> You are looking for help in the wrong place. You need a design change
> and that is offered by Shimano and Cane Creek, among others.
>
> Jobst Brandt
if you bothered to read what he said, he doesn't claim brinelling caused
the looseness - he conjectured that initial loosening was such as could
allow false brinelling - and it could. modern bearings with correct
lubrication and correct pre-load do not [cannot] false brinell. some
bearings are shipped either stripped down or PRE-LOADED to alleviate the
false brinelling that occurs when loose. you /do/ know that, right?
peter
> It occurred to me that a slightly loose headset could cause premature
> damage to the bearing races. A well adjusted headset in an accurately
> set up frame with a slight bearing preload should distribute impact
> loads throughout all of the ball bearings preventing this kind of
> failure.
>
> A slightly loose headset will allow movement during normal road
> vibration and impact shock. This can easily lead to indentations in the
> bearing surfaces especially in the front of the lower headset and crown
> races.
My experience has been the exact opposite. The times that I've
encountered indexing of a headset it has been when tightened to the
point of having at least some preload. OTOH, I've ridden many
thousands of miles with a slight, but definitely noticeable, looseness
in the headset bearings and have never had these develop the
indentations that cause indexing.
This appears to support Brandt's contention that the problem is due to
fretting rather than Brinelling. When adjusted with some preload and
ridden in a straight line the lubricant can be squeezed out of the
contact zone and result in fretting from vibration without significant
movement. In a loose headset there will be enough movement to
replenish the lubricant.
Johnny Sunset aka Tom Sherman
Citation?
Johnny Sunset aka Tom Sherman
"jim beam" wrote:
>
> JOBST, YOU'RE MISSING THE VITAL AND VERY IMPORTANT POINT THAT BEARINGS
> CAN BOTH TRUE AND FALSE BRINELL. CRASHES AND IMPACT CAN LOAD BEARINGS
> SUFFICIENTLY TO TRUE BRINELL THEM. I KNOW THIS FROM EXPERIENCE AND
> ANALYSIS.
> if you bothered to read what he said, he doesn't claim brinelling caused
> the looseness - he conjectured that initial loosening was such as could
> allow false brinelling - and it could. modern bearings with correct
> lubrication and correct pre-load do not [cannot] false brinell. some
> bearings are shipped either stripped down or PRE-LOADED to alleviate the
> false brinelling that occurs when loose. you /do/ know that, right?
Note the sentence with both capital AND lowercase letters.
jim beam
> jim beam wrote:
>> Johnny Sunset aka Tom Sherman wrote:
>>> "jim beam" wrote:
>>>> ...
>>>> one more thing our esteemed alumni mechanical engineering expert seems
>>>> to be missing....
>>> Did your application to Stanford come back marked "rejected", jim beam?
>>>
>> actually, no. my alma mater had a full materials faculty in its own
>> right - it wasn't a mere mere sub-department of the engineering faculty.
>
> Citation?
>
* * Chas
<jobst....@stanfordalumni.org> wrote in message
news:45342d88$0$34559$742e...@news.sonic.net...
It's rather presumptuous to assume that I was looking for help. I was
making an observation and supposition about one possible cause of
indentation related headset failure not the mechanisms of the failures.
When I ran a pro bike shop, most bikes that came in for service or
repairs had headset problems. Most of them had loose headsets. Most
Campy NR headsets had indentations in the lower headset cups and crown
races and most of those indentations were in the frontal areas.
That's why we were curious enough to do Rockwell Hardness testing on
various brands of headsets. We found that Campy components of that era
were around 48Rc versus several other brands that were >60Rc.
Over a 5 year period, we found very few failures in the harder headsets.
Also, I've rarely seen any indentations in the upper races.
There is no one single failure mode in bicycle headsets! Case hardened
races or lower hardness races are probably more susceptible to failure
developing out of indentations whereas harder races are more affected by
other types of wear.
As I mentioned, the headtube and fork crown surfaces were inspected
before and after installation. The Campy Chorus crown race is solid
steel. The top and bottom races are steel pressed into aluminum cups. My
supposition is that the headset loosened after the steel races seated
further into the aluminum cups after riding. There's only about .005
play but there was a slight preload before I went on the ride.
Also, I've pounded on lots of cotter pins and never saw indentations in
the BB cups or spindle.
Chas. Who?
* * Chas
"peter" <prat...@comcast.net> wrote in message
news:1161049157.3...@i3g2000cwc.googlegroups.com...
There is no indexing in this headset yet. It loosened after only an hour
of riding and I was musing about the cause without tearing it apart.
Headsets are thrust bearings and most thrust bearings require a preload.
What no one seems to address here is effect the quality of the steel in
the races and the type and degree of hardness has on indexing. I've
caused indentations in the races of a newly installed cheap case
hardened headset by overtightening when replacing them on a customer's
bike.
I've only run across a few damaged bearing surfaces in through hardened
steel headset races over 55Rc hardness.
I've seen damage in most case hardened headsets. The zone of hardness
may be only .005" to .010" thick and less than 55Rc. The steel under
the case can be 30-35Rc or less which would easily allow indentation to
occur.
My point being, most of this discussion has consisted of arguments about
various failure modes when the real problem lies in the bearing material
and hardness of headset races.
In addition, the old reference to false brinelling in bearings of
automobiles being transported doesn't take into consideration advances
in lubrication technology. Oil and grease are both liquids that can be
easily squeezed out between bearing surfaces. Modern lubricants contain
extreme pressure additives that create solid molecular boundary layers
to reduce or prevent metal to metal contact.
Additives like molybdenum disulfide remain in place for long periods
without breaking down and can work when almost dry.
http://en.wikipedia.org/wiki/Molybdenum_disulfide
Unless someone used a very low quality lubricant in their headset like
Vaseline or KY, Jobst's argument about lubrication failure is pretty
weak.
Chas. Who?
jta...@nospam.hfx.andara.com
Tell us if you still say that crank cotters are correctly installed by
tightening the nut.
jta...@nospam.hfx.andara.com
<spamv...@bad.example.net> wrote:
>> The nut on the cotter served to keep it from working
>> its way loose. It could not be used to seat the cotter.
>
>that's b.s.
>
Why do you claim in contrast to all experience and writing that crank
cotters were correctly installed by tightening the nut?
jim beam
absolutely. that's why we haven't had motor vehicle transportation
bearing failure in 50 years. regarding materials, vacuum degassed
steels have also proven to be highly effective in this situation.
* * Chas
"jim beam" <spamv...@bad.example.net> wrote in message
news:Lamdnb1QTIjuTqnY...@speakeasy.net...
Editorial:
Only a very small portion of the bicycles ever produced have utilized
high quality components. Worldwide, most bikes use the cheapest
manufacturing techniques, materials and components available.
I have a box of old headset parts that appear to have been drawn or
stamped, case hardened and chrome plated. These were from the bike boom
days and early 1980's. There's no sign of any grinding or machining on
them.
Bearing components made of premium grade steels are machined in the soft
condition to near finished size, hardened through heat treatment and
then precision ground to the final dimensions. This adds to the
manufacturing costs but provides longer bearing life.
I've seen very few bearing failures in properly hardened bicycle
components manufactured in this manner except where abrasive materials
entered the bearings. The use of seals or sealed bearings can eliminate
most of these kinds of failures.
Chas.