A problem with gears.



Martin Wilson wrote:
> On Mon, 13 Dec 2004 20:36:20 +0900, James Annan
> <[email protected]> wrote:


> >Personally I think you're not firing on all cyclinders if you
> >prefer to make up some feeble straw man rather than just read
> >what I said, which was neither ambiguous nor inaccurate.

>
> Make up some straw man? Never heard of this before and frankly don't
> know what your talking about.


James is suggesting, I think, that the argument you refuted was not the
one he was putting forward but rather a superficially similar one you
set up just so you could knock it down. That's my understanding of a
"straw man". Whether that's what you were actually doing I'll leave for
the two of you to sort out.

> The OP has high levels of wear on the sprocket most used that has
> the most load applied. The wear on the other sprockets is much,
> much lighter and from the sound of it almost non existant. I assume
> the same grit and dirt moves with the chain and goes round all the
> sprockets and not just the smallest. I don't know the percentage of
> how much the smallest rear sprocket is used compared to the other
> gears. I suspect quite high but theres possibly more chain/teeth
> contact in the lower gears at any one time. Obviously cadence is
> factor too. Surely though if load wasn't a factor the wear and tear
> of the chain and sprocket teeth would be more general across the
> whole range of cassette gears.


In the OP's case the wear is greatest on the smallest sprocket because
that's the one he uses most of the time. First the chain wears causing
it to "stretch". Once it has stretched sprocket wear begins to
accelerate. The more stretched the chain, the faster the sprocket
wears. The sprockets that wear most are the ones that are used most.
I've personally worn out 2 and 3 in a 6 sprocket cluster by continuing
to use a stretched chain. 4 was less worn and 1 (the largest) and 5 and
6 (the smallest) were hardly worn at all.

I believe distance covered is far more significant in this than either
load or cadence. No matter what gear you are in each tooth of the
engaged sprocket meshes with a chain link once per wheel revolution.
The point about the larger sprockets having more teeth in contact with
the chain is misleading, I think, because the stretched chain no longer
meshes properly with the teeth; the load is concentrated on the top
teeth until the sprocket has worn sufficiently to conform to the
stretched chain.

--
Dave...
 
Martin Wilson wrote:
> On Mon, 13 Dec 2004 20:36:20 +0900, James Annan
> <[email protected]> wrote:
>
>
>>Martin Wilson wrote:
>>
>>
>>
>>>Personally I think your not firing on all
>>>cyclinders if you think load isn't a factor with regard chain and
>>>cassette wear even if its not the main factor but hopefully if you
>>>read some real world experiences you might think differently.

>>
>>Personally I think you're not firing on all cyclinders if you prefer to
>>make up some feeble straw man rather than just read what I said, which
>>was neither ambiguous nor inaccurate.
>>

>
>
> Make up some straw man? Never heard of this before and frankly don't
> know what your talking about.


The straw man is your suggestion that I, or anyone else, is claiming
that load is not a factor at all in chain and cassette wear. Of course
it's _a_ factor (I have already written "I agree that in principle,
increasing the load will increase the wear"), just a small one in
comparison to how clean or dirty the chain is. As has now been
explained to you several times over, with supporting evidence.

James
--
If I have seen further than others, it is
by treading on the toes of giants.
http://www.ne.jp/asahi/julesandjames/home/
 
Donny wrote:
> soup wrote:


> The two smallest cogs on the cassette are separate from the rest so
> hopefully I could just buy replacements for these?


Theoretically, but I don't think you'll be able to get them easily or
economically.

> I guess I'll need a new chain as well?


I'm certain you do. The question really is whether you also need a new
cassette. If you just change the chain you'll know soon enough as it
will skip worse than ever if the cassette is worn out. Are you happy
about tackling these jobs yourself or do you need some advice?
--
Dave...
 
On Tue, 14 Dec 2004 11:59:16 +0000, Martin Wilson
<[email protected]> wrote in message
<[email protected]>:

>Make up some straw man? Never heard of this before and frankly don't
>know what your talking about.


http://www.datanation.com/fallacies/index.htm

Guy
--
"then came ye chavves, theyre cartes girded wyth candels
blue, and theyre beastes wyth straynge horn-lyke thyngs
onn theyre arses that theyre fartes be herde from myles
around." Chaucer, the Sheppey Tales
 
Martin Wilson wrote:
> I assume the same grit and dirt moves with the
> chain and goes round all the sprockets and not just the smallest. I
> don't know the percentage of how much the smallest rear sprocket is
> used compared to the other gears. I suspect quite high but theres
> possibly more chain/teeth contact in the lower gears at any one time.
> Obviously cadence is factor too. Surely though if load wasn't a factor
> the wear and tear of the chain and sprocket teeth would be more
> general across the whole range of cassette gears.


Grit on the sprocket will not actually wear it, this is not a corrosion
effect. The wear comes when that grit is ground into the sprocket by the
chain. It doesn't matter how much shite is on the other sprockets if the
chain never touches them.

Jon
 
Jon Senior wrote:
> Martin Wilson wrote:
>> I assume the same grit and dirt moves with the
>> chain and goes round all the sprockets and not just the smallest. I
>> don't know the percentage of how much the smallest rear sprocket is
>> used compared to the other gears. I suspect quite high but theres
>> possibly more chain/teeth contact in the lower gears at any one time.
>> Obviously cadence is factor too. Surely though if load wasn't a
>> factor the wear and tear of the chain and sprocket teeth would be
>> more general across the whole range of cassette gears.

>
> Grit on the sprocket will not actually wear it, this is not a
> corrosion effect. The wear comes when that grit is ground into the
> sprocket by the chain. It doesn't matter how much shite is on the
> other sprockets if the chain never touches them.


I think most sprocket wear is caused by chain elongation. The argument
goes that grit within the chain wears it out and elongates it.

~PB
 
dkahn400 wrote:
> Donny wrote:
>
>>I guess I'll need a new chain as well?

>
> I'm certain you do. The question really is whether you also need a new
> cassette. If you just change the chain you'll know soon enough as it
> will skip worse than ever if the cassette is worn out. Are you happy
> about tackling these jobs yourself or do you need some advice?


I think I'll go for a new chain to start with then. What determines the
type of chain to get, it is just the number of sprockets on the cassette?

Cheers,
Donny
 
On Mon, 13 Dec 2004 20:36:20 +0900, James Annan
<[email protected]> wrote:

>Martin Wilson wrote:
>
>
>>
>> Personally I think your not firing on all
>> cyclinders if you think load isn't a factor with regard chain and
>> cassette wear even if its not the main factor but hopefully if you
>> read some real world experiences you might think differently.

>
>Personally I think you're not firing on all cyclinders if you prefer to
>make up some feeble straw man rather than just read what I said, which
>was neither ambiguous nor inaccurate.
>
>James


So how exactly have I moved the goalposts to create this so called
feeble straw man?

All I've ever said is load is a factor, not the main factor but an
important factor and obviously under certain conditions (extreme
weight or gearing perhaps the major factor especially where a chain
hasn't been well maintained). I've never argued against grit/dirt
being a major factor. However there seems to be plenty of examples of
people where load has been the main factor for them in excessive chain
and sprocket wear and other sources where dirt and grit are the main
factor.

">The chain almost certainly has stretched, but this has nothing to do
>with the weight of the rider or the amount of hill climbing done."


This statement clearly states that the writer does not perceive load
to be a factor at all in chain roller wear.
 
Martin Wilson wrote:
> ">The chain almost certainly has stretched, but this has nothing to
> > do with the weight of the rider or the amount of hill climbing
> > done."

>
> This statement clearly states that the writer does not perceive load
> to be a factor at all in chain roller wear.


That was me, I think. I still don't believe that load is a significant
factor in chain wear in normal bicycle riding. When the rollers are
full of grinding paste (oil + grit) they will wear rapidly with any
rider. If they are clean and lubricated inside they will wear much more
slowly with any rider.

--
Dave...
 
Donny wrote:

> I think I'll go for a new chain to start with then. What determines
> the type of chain to get, it is just the number of sprockets on the
> cassette?


Basically, yes. Go for the lowest available number that fits. You
cannot use a 7/8 speed chain on a 9 speed cassette, for example, but
you can use a 9 speed chain on a 7 speed cassette. A 7/8 speed chain,
however, would be slightly better.

There's also a tremendous range of prices. I personally don't think
there's an awful lot to be gained in spending £30 on a chain rather
than £10. Sachs/SRAM chains are very good. These come with a special
link that enables you to break and remake the chain without special
tools, once you have the knack. You will, however, still need a chain
link extractor when you first put the chain on if you need to remove
any links, and of course to remove the old chain.

--
Dave...
 
Donny wrote:
> I think I'll go for a new chain to start with then. What determines
> the type of chain to get, it is just the number of sprockets on the
> cassette?


It is really. Cassette manufacturers tell you to use only their make of
chain but that's not essential in practice.

SRAM chain is pretty good. PC41, PC48, PC58, PC68 suits 8-speed and
7-speed; PC49 etc suits 9-speed. Make sure you get a "Powerlink"
connecting link as they're not always supplied with the chains.

~PB
 
dkahn400 wrote:
> Martin Wilson wrote:
>
>>">The chain almost certainly has stretched, but this has nothing to
>>
>>>do with the weight of the rider or the amount of hill climbing
>>>done."

>>
>>This statement clearly states that the writer does not perceive load
>>to be a factor at all in chain roller wear.

>
>
> That was me, I think. I still don't believe that load is a significant
> factor in chain wear in normal bicycle riding. When the rollers are
> full of grinding paste (oil + grit) they will wear rapidly with any
> rider. If they are clean and lubricated inside they will wear much more
> slowly with any rider.
>


Load creates the normal force on the bearing surface. This is one the
fundamental factors in determining wear rates. It's quite basic
engineering knowledge. Look up "tribology" references.

--
Joe * If I cannot be free I'll be cheap
 
Martin Wilson wrote:

> On Mon, 13 Dec 2004 20:36:20 +0900, James Annan
> <[email protected]> wrote:
>
>
>>Martin Wilson wrote:
>>
>>
>>
>>>Personally I think your not firing on all
>>>cyclinders if you think load isn't a factor with regard chain and
>>>cassette wear even if its not the main factor but hopefully if you
>>>read some real world experiences you might think differently.

>>
>>Personally I think you're not firing on all cyclinders if you prefer to
>>make up some feeble straw man rather than just read what I said, which
>>was neither ambiguous nor inaccurate.
>>
>>James

>
>
> So how exactly have I moved the goalposts to create this so called
> feeble straw man?


By posting - as a follow-up to me - a message saying "I think your not
firing on all cyclinders if you think load isn't a factor with regard
chain and cassette wear" when not only had I not said such a thing, I
had clearly stated that on the contrary I believe that load IS a factor.
I know from my own experiences (not just reading about others) that
tandem chains wear out faster than single bike chains under similar
riding conditions. However, this very large difference in load still has
a relatively small effect compared to the dirtiness of the chain, which
can certainly cause a factor of 10 difference in life (even without
going as far as enclosed oil baths or extremely dirty MTBing - I'm just
talking about a fair weather rider who cleans the chain every time
versus someone who rides on the road in all weather with minimal
maintenance).

James
--
If I have seen further than others, it is
by treading on the toes of giants.
http://www.ne.jp/asahi/julesandjames/home/
 
JLB wrote:

> Load creates the normal force on the bearing surface. This is one the


> fundamental factors in determining wear rates. It's quite basic
> engineering knowledge. Look up "tribology" references.


If there's no load the chain won't turn at all so there'll be no wear.
Is that your point?

--
Dave...
 
dkahn400 wrote:
> JLB wrote:
>
>
>>Load creates the normal force on the bearing surface. This is one the
>>fundamental factors in determining wear rates. It's quite basic
>>engineering knowledge. Look up "tribology" references.

>
>
> If there's no load the chain won't turn at all so there'll be no wear.
> Is that your point?


You are missing the point.

Perhaps you cannot visualise the forces on individual components of the
chain that are moving in relation to one another.

You wrote earlier

"That was me, I think. I still don't believe that load is a significant
factor in chain wear in normal bicycle riding. When the rollers are
full of grinding paste (oil + grit) they will wear rapidly with any
rider. If they are clean and lubricated inside they will wear much more
slowly with any rider."

The load is a *necessary* but *not* *sufficient* condition, and where
bearing surfaces in relative motion make contact, wear rate will
increase with increased load. Note also that where foreign particles
like bits of grit are present, the contact points of the grit and the
chain components is part of the bearing area; it is the area on which
the load bears. The wear that occurs is due to huge local stresses,
easily above the yield stress, acting on tiny areas of the components as
the rough surfaces from earlier damage and bits of grit are pushed into
and dragged across other surfaces. At a microscopic level the effect is
like chisels digging into the metal surface. And of course the damage is
related to the load; just as you cannot chisel chunks out of a block
of wood unless you drive the chisel into the wood.

Your statement "I still don't believe that load is a significant
factor in chain wear" suggests you refuse to accept basic mechanical
engineering knowledge.

--
Joe * If I cannot be free I'll be cheap
 
On Wed, 15 Dec 2004 17:13:49 +0000, JLB <[email protected]>
wrote:

>Your statement "I still don't believe that load is a significant
>factor in chain wear" suggests you refuse to accept basic mechanical
>engineering knowledge.


It might suggest that except for the fact that you conveniently cut
four words from the end of my statement. What I actually said was, "I
still don't believe that load is a significant factor in chain wear in
normal bicycle riding."

--
Dave...

Get a bicycle. You will not regret it. If you live. - Mark Twain
 
Dave Kahn wrote:
> On Wed, 15 Dec 2004 17:13:49 +0000, JLB <[email protected]>
> wrote:
>
>
>>Your statement "I still don't believe that load is a significant
>>factor in chain wear" suggests you refuse to accept basic mechanical
>>engineering knowledge.

>
>
> It might suggest that except for the fact that you conveniently cut
> four words from the end of my statement. What I actually said was, "I
> still don't believe that load is a significant factor in chain wear in
> normal bicycle riding."
>

Fine. Put it back in. And explain what the difference is between a
bicycle chain in normal bicycle riding and a mechanical system of
components subject to conventional well-known tribological wear
mechanisms. If you can't, you are still pissing into the wind.

--
Joe * If I cannot be free I'll be cheap
 
On 16/12/04 12:03 am, in article [email protected],
"JLB" <[email protected]> wrote:

> Dave Kahn wrote:
>> On Wed, 15 Dec 2004 17:13:49 +0000, JLB <[email protected]>
>> wrote:
>>
>>
>>> Your statement "I still don't believe that load is a significant
>>> factor in chain wear" suggests you refuse to accept basic mechanical
>>> engineering knowledge.

>>
>>
>> It might suggest that except for the fact that you conveniently cut
>> four words from the end of my statement. What I actually said was, "I
>> still don't believe that load is a significant factor in chain wear in
>> normal bicycle riding."
>>

> Fine. Put it back in. And explain what the difference is between a
> bicycle chain in normal bicycle riding and a mechanical system of
> components subject to conventional well-known tribological wear
> mechanisms. If you can't, you are still pissing into the wind.


The arguement is not over whether load is a factor, it clearly is, but over
the significance of the load.

Which variable would have most effect on the chain wear rate? High load or
high dirt?

At which point we gather lots of data and start to evaluate them with
respect to wear rate.

If you ignore the dirt on the chain, what is the correlation between load
and wear? (R = x)

If you ignore the load on the chain, what is the correlation between dirt
and wear? (R = y)

Now, if y is much bigger than x (much closer to 1) then I would clearly
have to support the statement that dirt is significantly more important wrt
chain wear than load.

Data anyone?

...d
 
David Martin wrote:
> On 16/12/04 12:03 am, in article [email protected],
> "JLB" <[email protected]> wrote:
>
>
>>Dave Kahn wrote:
>>
>>>On Wed, 15 Dec 2004 17:13:49 +0000, JLB <[email protected]>
>>>wrote:
>>>
>>>
>>>
>>>>Your statement "I still don't believe that load is a significant
>>>>factor in chain wear" suggests you refuse to accept basic mechanical
>>>>engineering knowledge.
>>>
>>>
>>>It might suggest that except for the fact that you conveniently cut
>>>four words from the end of my statement. What I actually said was, "I
>>>still don't believe that load is a significant factor in chain wear in
>>>normal bicycle riding."
>>>

>>
>>Fine. Put it back in. And explain what the difference is between a
>>bicycle chain in normal bicycle riding and a mechanical system of
>>components subject to conventional well-known tribological wear
>>mechanisms. If you can't, you are still pissing into the wind.

>
>
> The arguement is not over whether load is a factor, it clearly is, but over
> the significance of the load.


The original statement said load was not significant, which is absurd.
You have changed the debate.
>
> Which variable would have most effect on the chain wear rate? High load or
> high dirt?


The chain will wear if there is load even with no dirt whatsoever,
because the lubrication of a bicycle chain will in practice not be
perfect. Whenever the steel components make contact and there is any
normal force there will be wear. The wear will produce particles that
will themselves accelerate the rate of wear.
>
> At which point we gather lots of data and start to evaluate them with
> respect to wear rate.
>
> If you ignore the dirt on the chain, what is the correlation between load
> and wear? (R = x)
>
> If you ignore the load on the chain, what is the correlation between dirt
> and wear? (R = y)


There might even be a point where there is so much dirt it decreases the
rate of wear by providing a better spread of load and therefore lower
stresses across the chain components.
>
> Now, if y is much bigger than x (much closer to 1) then I would clearly
> have to support the statement that dirt is significantly more important wrt
> chain wear than load.
>
> Data anyone?


No. Zebedee has had enough.


--
Joe * If I cannot be free I'll be cheap
 
On 16/12/04 12:57 am, in article [email protected],
"JLB" <[email protected]> wrote:

> David Martin wrote:
>> On 16/12/04 12:03 am, in article [email protected],
>> "JLB" <[email protected]> wrote:
>>
>>
>>> Dave Kahn wrote:
>>>
>>>> On Wed, 15 Dec 2004 17:13:49 +0000, JLB <[email protected]>
>>>> wrote:
>>>>
>>>>
>>>>
>>>>> Your statement "I still don't believe that load is a significant
>>>>> factor in chain wear" suggests you refuse to accept basic mechanical
>>>>> engineering knowledge.
>>>>
>>>>
>>>> It might suggest that except for the fact that you conveniently cut
>>>> four words from the end of my statement. What I actually said was, "I
>>>> still don't believe that load is a significant factor in chain wear in
>>>> normal bicycle riding."
>>>>
>>>
>>> Fine. Put it back in. And explain what the difference is between a
>>> bicycle chain in normal bicycle riding and a mechanical system of
>>> components subject to conventional well-known tribological wear
>>> mechanisms. If you can't, you are still pissing into the wind.

>>
>>
>> The arguement is not over whether load is a factor, it clearly is, but over
>> the significance of the load.

>
> The original statement said load was not significant, which is absurd.
> You have changed the debate.


It is not a change in the debate. Load having no effect is very different to
load having a significant role.

Significant in this case is relative to other factors. If dirt relates to
chain wear very well, but load poorly (evne though in the controlled
environment when the other factors are constant there is a clear linear
relationship) the load is not a *significant* factor.

>> Which variable would have most effect on the chain wear rate? High load or
>> high dirt?

>
> The chain will wear if there is load even with no dirt whatsoever,
> because the lubrication of a bicycle chain will in practice not be
> perfect. Whenever the steel components make contact and there is any
> normal force there will be wear. The wear will produce particles that
> will themselves accelerate the rate of wear.


Of course it will. This misses the point though. The point is how fast this
occurs and whether external dirt is significant in dramatically accelerating
the process.

>>
>> At which point we gather lots of data and start to evaluate them with
>> respect to wear rate.
>>
>> If you ignore the dirt on the chain, what is the correlation between load
>> and wear? (R = x)
>>
>> If you ignore the load on the chain, what is the correlation between dirt
>> and wear? (R = y)

>
> There might even be a point where there is so much dirt it decreases the
> rate of wear by providing a better spread of load and therefore lower
> stresses across the chain components.


Indeed. We can all speculate. Easy to do in the absence of data. (how many
angels on the head of that pin?)

>>
>> Now, if y is much bigger than x (much closer to 1) then I would clearly
>> have to support the statement that dirt is significantly more important wrt
>> chain wear than load.
>>
>> Data anyone?

>
> No. Zebedee has had enough.


Boingggg!

Time for bed.

...d