new life for worn Ti cogs

[forrestw]

I (stupidly, I know) bought a used 11-23 Record cassette last year and promptly found that while the 4 Ti cogs looked OK on first glance, they were in fact worn down and the 17-21 cogs skip if I stand up on 'em.

So, being ornery and having ready access to a mill w/ rotary table I'm gonna cut new surfaces into the teeth tomorrow and see how it rides. It looks to me like there's enough material there to do this once, probably not 2x.

Dunno if I'm compounding my initial 'stupid' here. The whole job, including building the necessary fixture is going to come to about 5 hours which isn't a bad return on investment if it works.

I'll f/up here whether the results look like success or failure :-)

[Blue Jays]

Good luck with your project!

[palincss]

Good luck replicating those complex tooth shapes with the ramps and cutouts and all. And by the way, isn't machining Titanium supposed to be rather difficult and very hard on tools?

[flydhest]

Good luck replicating those complex tooth shapes with the ramps and cutouts and all. And by the way, isn't machining Titanium supposed to be rather difficult and very hard on tools?

Don't listen to palincss, but by all means, take pictures!!

[Kontact]

How are you going to cut them? Forward along the chain's pathway, effectively thinning the teeth?

[jr59]

Please let us know how this comes out!

[dd74]

Hmmm...very interesting. Hopefully, you'll be successful, potentially making you the go-to guy for others with worn Ti cogs. Good luck! Keep us posted.

[oldpotatoe]

I (stupidly, I know) bought a used 11-23 Record cassette last year and promptly found that while the 4 Ti cogs looked OK on first glance, they were in fact worn down and the 17-21 cogs skip if I stand up on 'em.

So, being ornery and having ready access to a mill w/ rotary table I'm gonna cut new surfaces into the teeth tomorrow and see how it rides. It looks to me like there's enough material there to do this once, probably not 2x.

Dunno if I'm compounding my initial 'stupid' here. The whole job, including building the necessary fixture is going to come to about 5 hours which isn't a bad return on investment if it works.

I'll f/up here whether the results look like success or failure :-)

If it doesn't work.

http://www.etsy.com/listing/63513578/bicycle-cog-wind-chimes

[RPS]

So, being ornery and having ready access to a mill w/ rotary table I'm gonna cut new surfaces into the teeth tomorrow and see how it rides. It looks to me like there's enough material there to do this once, probably not 2x.

Dunno if I'm compounding my initial 'stupid' here.
Yeah, pretty much in my opinion; unless you just want to have fun doing it. ;)

The main problem goes beyond your ability; it’s because wear reduces tooth pitch, and machining will almost certainly reduce it further. You’ll end up with the same number of teeth in a smaller diameter, hence the chain can’t engage and carry load correctly.

[AngryScientist]

i dont see how this can end well. essentially when the cogs wear, they lose material. the only way to correct this, as i see it would be to build the surface back up by adding material, then re-machining back to original shape. simply machining without adding material is almost certainly not going to solve the problem.

you're likely going to ruin a chain in the tinkering process.

[forrestw]

How are you going to cut them? Forward along the chain's pathway, effectively thinning the teeth?
Yes, thinning the teeth in the tangential axis, here's the setup http://web.mit.edu/forrestw/www/images/index_setup_sm.jpg

Rotary table and a list of index positions for the number of teeth and a boring head setup to cut the correct radius matching the chain rollers.

Good luck replicating those complex tooth shapes with the ramps and cutouts and all. And by the way, isn't machining Titanium supposed to be rather difficult and very hard on tools?
Shifting ramps are already there. I'm not certain they'll work perfectly after re-cutting the bearing part of the teeth but I bet they work. Yeah hard on tools, that's what grinders are for :-) 6/4 and 3/2.5 TI are pretty hard but then so his hill climbing, you just gotta do it right.


The main problem ... wear reduces tooth pitch, and machining will almost certainly reduce it further. You’ll end up with the same number of teeth in a smaller diameter, hence the chain can’t engage and carry load correctly.
a. no, primary wear is tangential, radial wear (i.e. at the low spot between teeth is negligible. The reason for this is that most of the load is taken by the first 2 teeth and the vector is principally tangential). You'll see this in the picture of the cuts. b. even if there were significant radial wear, the new surface I'm cutting is at the correct pitch diameter. This does think the teeth

i dont see how this can end well. essentially when the cogs wear, they lose material. the only way to correct this, as i see it would be to build the surface back up by adding material, then re-machining back to original shape. simply machining without adding material is almost certainly not going to solve the problem.

you're likely going to ruin a chain in the tinkering process.
I'm sacrificing a little strength but unless the teeth are really not strong enough with the reduced tangential thickness, they should hold up for another season or more. I won't damage a chain because when I test it out tomorrow, either it shifts well enough to work or it doesn't and i write off the time. Based on what I saw doing the cutting, however I think it's gonna work ok. TBD

[Dekonick]

You call that proper equipment? :beer: :banana: :o :)

[forrestw]

I think these are gonna ride ok, you see in this photo the 19t cog below as-worn and the 17 with the drive side of the teeth re-cut on the correct pitch diameter with a loss of about .015" of tooth thickness, i.e. 20%.

http://web.mit.edu/forrestw/www/images/teeth_sm.jpg

or the full image -- http://web.mit.edu/forrestw/www/images/teeth.jpg

What I observed while cutting the new wear surface is, as you can see the heads of the teeth are fairly worn from shifting activity, I also saw that (no surprise) the teeth in the shifting-ramp area are far more severely worn and I think this affects the nearby teeth. it almost makes me think fondly of the days of friction shifting. I'll see how these ride and shift tomorrow.

[sbparker31]

I have been following this thread, but haven't really understood it. After looking at the pictures, it all makes sense now. I think it probably will work! And furthermore, I bet you can recut the teeth several times, as I suspect you will have to remove a significant amount of material before the teeth are weakened to the point of breakage. And it seems clear that as long as each tooth is cut the same amount, you are not changing the distance as per the chain, only "shifting" the chain by the amount of the cut.

[jr59]

Ride report????

Extended ride report???

[soulspinner]

My guess is youll have more noise and still may have a skip. With less material the chain will ride up on top even more . Best of luck...............I do like the wind chime though, Peter. :beer:

[Blue Jays]

Excellent pics really helped explain the idea.
Hope that it works perfectly in your testing!

[oldpotatoe]

My guess is youll have more noise and still may have a skip. With less material the chain will ride up on top even more . Best of luck...............I do like the wind chime though, Peter. :beer:

Campag cogs are CP titanium, BTW-

[RPS]

Your reply to RPS:
a. no, primary wear is tangential, radial wear (i.e. at the low spot between teeth is negligible. The reason for this is that most of the load is taken by the first 2 teeth and the vector is principally tangential). You'll see this in the picture of the cuts. b. even if there were significant radial wear, the new surface I'm cutting is at the correct pitch diameter. This does think the teeth
Using the same logic then there is little reason to hardly ever replace chainrings, right?

Chain load over the first few teeth is even more tangential, and the wear is hardly radial. The tooth pockets mainly wear from a circular to oval shape but due to symmetry the spacing is equal (essentially what you are cutting). Is that the way you see it?

[forrestw]

Campag cogs are CP titanium, BTW-

Pete, inclined as I am to believe your knowledge, they sure didn't feel like CP when cutting, the degree of hardness was more than I'd attribute to work-hardening.

CP Ti isn't much stronger than 7075 aluminum, tho it probably is more wear-resistant.

Your source for this info?

[martinrjensen]

you know of course that there is no possible way this can work unless you eliminate one complete tooth, then you won't have the same cog....

[RPS]

you know of course that there is no possible way this can work unless you eliminate one complete tooth, then you won't have the same cog....
If you mean like taking a 21T and cutting to a new 20T so that the correct pitch diameter was maintained that’s not possible because there is not enough material there. If one tried, at some point around the circumference of the cog the top of one new tooth would be at the valley between two existing teeth, which means you’d have to make the cog a lot smaller than one tooth smaller. And as can be seen from the pictures, existing cogs don’t have enough depth.

I think what he is attempting is similar (but not quite the same because it’s not entirely comparable) as what tandem owners occasionally do with their timing chainrings when they swap them so that the chain rides against lightly worn tooth areas. Even though the rings can be worn on one side of teeth, the other side usually has almost no wear. By reversing, you can extend life by using the unworn side as if it were a new ring. In a way he is trying the same basic idea but by cutting new teeth profiles.

[bshell]

It sure seems logical. And like a worthwhile experiment just to see what happens.

I wouldn't have thought to try this but I also don't have or know how to use his tools.

I bet it works!

[forrestw]

I went out today and found a moderate length 10% grade and tested each of the 4 low cogs that I resurfaced, shifting, climbing out of the saddle and full-on sprint. Before re-cutting the cogs this was causing chain-skip; that's all gone now. Shifting feels pretty much standard and that's the only thing I was really thinking might be compromised. I'm pretty good about backing off pressure on the drivetrain when I shift and so to whatever extent the lateral strength has been reduced that won't be a worry.

So I'll put in a longer ride tomorrow AM but since stomping on it is the failure-case I'm not too worried about a regular 20 mile ride. If that goes ok then Saturday will be 150 and I'll report again if it's held up :-).

One thing I forgot to estimate in the time to do this was de-burring the cut surfaces, there was an extra hour there but even given this, I gotta say I'll be looking over the 'bay for more used Ti cassettes.

Lastly, getting in close and machining new wear surfaces necessarily involved seeing up close how the shifting ramps and tooth geometry makes this whole indexed shifting thing works, a cool bonus.

[soulspinner]

Good for you, you made it work, those ti cassettes are expensive :crap:

[MattTuck]

bravo!

[forrestw]

Your reply to RPS:

Using the same logic then there is little reason to hardly ever replace chainrings, right?

Chain load over the first few teeth is even more tangential, and the wear is hardly radial. The tooth pockets mainly wear from a circular to oval shape but due to symmetry the spacing is equal (essentially what you are cutting). Is that the way you see it?

It's been my experience that chainrings last a whole lot longer than cogs. The 'approximately tangential' nature of the forces breaks down in primarily 2 cases; the first being small tooth-count / diameter and the second and more severe being the case of a worn chain. (but the force distribution is very different in each of these). Anyhow, I think that's explains the longer life of chainrings even though they're made of much less wear-resistant material.

And yeah the spacing of the force-bearing surfaces should be identical as new. (except perhaps that by my observation my machine work was a little more precise than Campy's :-))

[Mark McM]

It's been my experience that chainrings last a whole lot longer than cogs. The 'approximately tangential' nature of the forces breaks down in primarily 2 cases; the first being small tooth-count / diameter and the second and more severe being the case of a worn chain. (but the force distribution is very different in each of these). Anyhow, I think that's explains the longer life of chainrings even though they're made of much less wear-resistant material.

That, plus the chainring teeth generally see a fraction of the number of load cycles as do the rear sprockets. Just for example, if you used a 48 tooth chainring and a 16 tooth rear sprocket, the rear sprocket would rotate 3 times each for each rotation of the chainring. So the teeth on the rear sprocket would see 3 times as many load/unload cycles than do the chainring teeth.

[Blue Jays]

Congratulations!
Refinishing worn cassettes might be a business just yet.
Keep us posted on further testing, forrestw...

[forrestw]

That, plus the chainring teeth generally see a fraction of the number of load cycles as do the rear sprockets. Just for example, if you used a 48 tooth chainring and a 16 tooth rear sprocket, the rear sprocket would rotate 3 times each for each rotation of the chainring. So the teeth on the rear sprocket would see 3 times as many load/unload cycles than do the chainring teeth.

Maybe if you're riding a fixed gear, even with a corncob on the back a 10 spd cassette's got 2x more teeth on the rear than the front so unless you're a rider that really focuses on a small subset of the rear cogs then it's actually the reverse -- the ring teeth see 2x more load cycles than the cog teeth.

[thwart]

Congratulations!
Refinishing worn cassettes might be a business just yet.
Keep us posted on further testing, forrestw...
This is great! Recycling in the best sense.

Once you've set up the machining tools, I wonder if it could indeed be possible to make a small business out of this... ?

[Mark McM]

Maybe if you're riding a fixed gear, even with a corncob on the back a 10 spd cassette's got 2x more teeth on the rear than the front so unless you're a rider that really focuses on a small subset of the rear cogs then it's actually the reverse -- the ring teeth see 2x more load cycles than the cog teeth.

I'm not so sure the effect can be so easily discounted. Most riders wear out the 2 or 3 sprockets in the middle of the cassette well before they do the sprockets at the ends, so when you consider that the wear is split between 2 (or 3) chainrings, the teeth on the most used sprockets in the middle of the cassette may still see more load cycles than the chainring sprockets (although not multiple times more).

[ultraman6970]

true

[CunegoFan]

Back in the day when cassettes were made from individual cogs did people flip the cogs when the cassette was worn? Campy has that funky spline that makes flipping Veloce cogs impossible, but it seems like that it would be a pretty cool feature if cassettes were made so that you could get double use out of them.

[forrestw]

Rode the B2B yesterday, 148 miles +riding to / fromt he stert on the re-cut cassette it shifted perfectly every time, ground out 7500' of climbing, a lot of it out of the saddle and pushing hard on the Ti cogs.

They look fine in terms of wear this morning, I think I will make this my regular riding cassette for this season.

Oh and not surprisingly all of the prettiest bikes on the ride were Serottas :-)

[thwart]

OK, now need your address to send my used ti cogs... are you good with PayPal for funds transfer?

Just kidding...

... or am I?

[AngryScientist]

great work. i had my doubts, but you can argue with success. anytime you give new life to an otherwise trashed part, its a good thing. :beer:

[jr59]

I hope you KNOW that this may very well turn into a part time.....errr...full time job....errrr...I mean hobby! :D

What was that address again!



Great job with this.

[Nil Else]

Not sure whether there's enough profit margin for a business but... competition or partnership? http://forums.thepaceline.net/showthread.php?t=92193 :)

[Kontact]

Not sure whether there's enough profit margin for a business but... competition or partnership? http://forums.thepaceline.net/showthread.php?t=92193 :)
Seems like that would require a bunch of people who are cheap enough to want to have their cassette machined, but big enough spenders to buy a disposable Ti cassette in the first place.

Something like developing a market for used Gucci underpants.


I'm glad it worked for Forestw, a machinist with time on his hands.

Actually, the marketability of this might be for developing a jig that allows you to do it with simple tools - kind of like a chainsaw sharpening tool.

[Jason E]

I think I'd want it reviewed by someone else other then the tinkerer before sending in a cassette.

If it is a fresh cassette after cutting, we should be able to throw it onto a bike with a new drive-train and have a rider evaluate.

Pez?

[thwart]

Seems like that would require a bunch of people who are cheap enough to want to have their cassette machined, but big enough spenders to buy a disposable Ti cassette in the first place.

Something like developing a market for used Gucci underpants.

I'm glad it worked for Forestw, a machinist with time on his hands.

Actually, the marketability of this might be for developing a jig that allows you to do it with simple tools - kind of like a chainsaw sharpening tool.You do have a point, but new steel/Ti 10 spd cassettes are ~ $300-$400. Many got the cassette when they bought a gruppo or entire bike, and didn't pay those individual part prices.http://www.amazon.com/Campagnolo-10-Speed-Titanium-Bicycle-Cassette/dp/B0015EI3Y0

If you can get another season out of your Record cassette for something under $100, I suspect there may be a market.

And you have to hand it to someone who's willing to do the work to prove a point---I wouldn't dismiss it as 'a machinist with time on his hands'.

[Jason E]

You do have a point, but new steel/Ti 10 spd cassettes are ~ $300-$400. Many got the cassette when they bought a gruppo or entire bike, and didn't pay those individual part prices.http://www.amazon.com/Campagnolo-10-Speed-Titanium-Bicycle-Cassette/dp/B0015EI3Y0

If you can get another season out of your Record cassette for something under $100, I suspect there may be a market.

And you have to hand it to someone who's willing to do the work to prove a point---I wouldn't dismiss it as 'a machinist with time on his hands'.


Another good point... What would it cost?

You are going to spend $8 to mail it, He is going to charge $10 to ship it back, so, Forrestw, is this the sort of thing you would do for $80, including all mark-up?

What would you charge, excluding re-packageing and reshipping? And how many hours of time does it take you, including driving it back to the post office?

[Grant McLean]

You do have a point, but new steel/Ti 10 spd cassettes are ~ $300-$400. Many got the cassette when they bought a gruppo or entire bike, and didn't pay those individual part prices.

for those still riding 10 speed, i'd just buy a new centaur cassette for $65.

http://www.ribblecycles.co.uk/sp/road-track-bike/Cassettes-Road-Campagnolo-10Spd-Centaur-UD-Cassette/CAMPCASS515

-g

[Kontact]

You do have a point, but new steel/Ti 10 spd cassettes are ~ $300-$400. Many got the cassette when they bought a gruppo or entire bike, and didn't pay those individual part prices.http://www.amazon.com/Campagnolo-10-Speed-Titanium-Bicycle-Cassette/dp/B0015EI3Y0

If you can get another season out of your Record cassette for something under $100, I suspect there may be a market.

And you have to hand it to someone who's willing to do the work to prove a point---I wouldn't dismiss it as 'a machinist with time on his hands'.
I wasn't being dismissive at all. Take a look at my first post - I realized what he was doing right off and that it could work.

I was simply making the point that the amount of time it takes to do this right and the tools necessary may be prohibitively expensive to do as a retail service. 70+ precision cuts in a hard metal could easily equal the cost of the cassette. I certainly spend time on my own projects that I couldn't offer at the shop.


Anyway, I've heard the Ti part of the cassette is available separately. Anyone bought one?

[soulspinner]

for those still riding 10 speed, i'd just buy a new centaur cassette for $65.

http://www.ribblecycles.co.uk/sp/road-track-bike/Cassettes-Road-Campagnolo-10Spd-Centaur-UD-Cassette/CAMPCASS515

-g

Except for race wheels this is the way to go..............

[RPS]

I think I'd want it reviewed by someone else other then the tinkerer before sending in a cassette.

If it is a fresh cassette after cutting, we should be able to throw it onto a bike with a new drive-train and have a rider evaluate.

Pez?
From an economic standpoint, a key factor is how long the rework will last compared to an original. For example, if only half as many additional miles, the cost may be even harder to justify. Another factor is wear on other non-titanium cogs.

[forrestw]

I had no particular idea of doing it as a business, it was in fact a bit of a rhetorical question as I was in the process of setting up the machine as I posted.

That said, other people have made enough money on my good ideas in the past, I'm definitely hip to realizing any money to be made on this one and I think it's pretty easy to do, and I did drop a note to one forum member who'd expressed interest.

The first big cost consideration to the entrepreneur in this case is what will be the legal responses of the cassette manufacturers. I know if the speedplay guy made cassettes and I made a business of machining new wear surfaces in his product I could expect a cease & desist letter from his lawyer. I have no idea how campy/shimano management would be responding and then again, unlike the fellow offering speedplay rebuild kits I would not be providing any new parts.

The second big consideration of course is the possibility of a customer claiming that a modified cassette caused injury. Both of these could trump the simple economics.

To answer questions posed about why/how/cost.

First, lighter components as you move from Centaur/105 to (Super)Record/DA cost in the range of $2/gram saved and often (particularly here) with less, not more durability so in this case double that to $4/gram factoring in the 1/2 life span of Ti cogs.

That's the playing field. True, Centaur cassettes are <$100 and Chorus which are still all-steel are just a few FRN$ more. Whether Rec/SR or DA, the extra cost of Ti cogs is about the same on a per-cog basis and they charge more for cassettes with larger cogs. (corollary: Campy's pricing for Ti cog pairs is higher per-cog than buying a new cassette)

Second, I'm not going to submit trial cogs to any 3rd party for testing (I might offer a lower price to the first few customers). I've climbed category 1 & 2 ascents on this rig and you can take my word on it or not as you like that every shift went fine and there was no visible damage to the cassette after.

Third, I expect that not-too-badly worn cogs could be machined 2 times with an expectation that they will last as long with each pass as they did the first go. Even badly worn cogs should still be good for one full-life machining. Drive chain design originated with far lower strength materials over 100 years ago and the basic geometries have not changed. For this reason there is substantially more material in cog teeth than is necessary from a strength perspective.

THIS IS WITH CAVEATS. A rider with a habit of shifting under full power application on reduced-web teeth might result in teeth bending. I haven't seen any evidence of it and have shifted during out of saddle climbing on 15% grades but I have a long standing habit of being as easy on my drivetrain as possible so others might get different mileage.

Fourth, in my first go each cut needed to be made 'by hand' as the Ti cogs flex significantly under the loading involved in machining The tooling can be improved to get to a point where this job could be done faster, possibly even with a CNC but that would would be a substantial investment. There's no conceivable tooling that's going to make this like sharpening a chain saw :-).

Fifth, Campy steel cogs are available for individual replacement at reasonable cost and in any case don't wear anywhere near as fast as Ti, even in much smaller tooth-count.

Lastly, I only use campagnolo chain in a campy drivetrain. My experience is that campy's 10 spd chain 'stretches' far less than shimano or SRAM chain. Because worn chain is the primary cause of rapid wear in cog teeth I think it's a terrible idea to run anything less than the best chain available on Ti cogs. (I haven't tried KMC's best offering and may do so someday)

-----

So all that said, if any forum members are interested in having their worn Ti CAMPY 10 speed cogs refurbished I'm game to take some initial orders. There's no reason not to work on 11 speed but I want to get started on 10.

Pricing will be based on tooth count, obviously working on a 13-29 cassette will cost more than an 11/23.

Contact me via PM if interested for pricing and schedule.

My initial warranty: If unhappy with the cassette operation, return it to me for analysis and recycling and receive a full refund of the FRN$ paid.

fw

[Jason E]

What sort of insurance do you have to cover this?

Do you have an entity set up to run this through?

I believe there are some holes in your model, but it is up to you if you want to take that risk (losing money) of course. That said, prior to offering a service like this, where you are modifying a piece of engineered drive-train, removing all manufacturers warranties on a wear item that is often under load, you really should have some real insurance and protect yourself and your possessions.

One of these breaks in a sprint or on a climb and someone breaks something, whether a shifter while falling or a bruised scrotum on a downtube, you are really want to be set up.

Dentists and Lawyers are a litigious group... (the Dentist part was humor)

[DRietz]

Jason, did you happen to read his last post?

Where he outlines that he can't accept liability?

PS: the dentist part tickled my inner BikeSnob. :D

[forrestw]

...modifying a piece of engineered drive-train, removing all manufacturers warranties on a wear item that is often under load, you really should have some real insurance and protect yourself and your possessions

So far in this thread you've suggested: there's not enough money to make it profitable; that I should do one on free and send it to someone for testing; that it would be stupid to do this without insurance (I'm pretty sure no insurer in his right mind would take this on so that point is fairly well moot).

Here's a basis on which I'd do it on spec if anyone wants to send me 2 worn Ti cassettes, I'll send one back for test and keep one for my own testing. We stipulate in writing the that these cassettes have no commercial value (and hold me harmless etc).

With no FRN$ having been received I'm pretty sure there's no way to attach liability.

Also, per DRietz if I do this for FRN$ there will be a big-ass disclaimer acknowledging cycling is dangerous; the buyer accepts all responsibility for using the modified parts; possibility of injury or death etc.

[Dekonick]

So far in this thread you've suggested: there's not enough money to make it profitable; that I should do one on free and send it to someone for testing; that it would be stupid to do this without insurance (I'm pretty sure no insurer in his right mind would take this on so that point is fairly well moot).

Here's a basis on which I'd do it on spec if anyone wants to send me 2 worn Ti cassettes, I'll send one back for test and keep one for my own testing. We stipulate in writing the that these cassettes have no commercial value (and hold me harmless etc).

With no FRN$ having been received I'm pretty sure there's no way to attach liability.

Also, per DRietz if I do this for FRN$ there will be a big-ass disclaimer acknowledging cycling is dangerous; the buyer accepts all responsibility for using the modified parts; possibility of injury or death etc.

Perhaps add to the disclaimer that the re-worked cassettes are for display only or only for indoor trainer use ;)

Kidding - I believe forrestw is an engineer and knows what he is doing. If I used Ti, I would give him a shot before tossing a cassette in the trash... worst case it skips right? I mean lets be honest... no different than if a tooth chips off a 'regular' cassette - I have had that happen to me before.

[bobswire]

Necessity is the mother of invention. Well done forrestw. I've run 32 tooth Mtn cassettes using a short cage road 5600 105 derailleur, I still hear folks saying a Campy 10 speed shifter won't work with Shimano/Sram drive ( yes they will and function smoothly using 9 speed cassette).
I'm a tinkerer always have been and like a challenge, especially when someones tells me it won't work.
Wish I would have kept a metal working mill (mini) and lathe I had a few years ago. At the time I was working full time and didn't have the time for outside projects.