Koogar

A few of you caught that Colby Pearce and I are undertaking an effort to make track frames that address everything we don't like about the options that are currently out there. Here's the custom dropout that will go on our prototypes.... I'm interested in reactions. A few comments on the design:

I started out wanting to do a bolt-on plate like Dave Tiemeyer's design (he has been very helpful in advising us), but a few considerations eventually pushed us towards improving on the slide-in insert design instead. Everyone knows how awful those can be when poorly executed and I was a bit reluctant to start out with this baggage, but I'm really happy with the end product.

The entire dropout body is oversized in thickness for strength and stiffness - the ends are 10mm wide, or about 20-25% greater than what has commonly been used on aluminum frames. The wide hoods obviate the need to squeeze down the stays at the junction to the dropout during fabrication, further strengthening the rear end of the frame. The flying buttresses are directed to the same end, but one thing we'll need to evaluate by riding and abusing the protos is the extent to which those wings will get bunged up over the course of many hypoxic gear changes. The wrench clearance is there, it's more a matter of wear and tear in practice.

The insert will have better properties than stainless steel for gripping wheel nuts, but there's a set screw as well, to eliminate any need to crank the wheel nuts down to a ridiculous torque. And it's not very obvious in the photo, but the insert rather than the dropout body will be threaded for the set screw, so that accidentally munching those threads will only require replacing the insert.

Finally, the dropout length gives almost 1.5 inches of usable range - enough to accommodate a good range of gearing without changing chains, while not getting into the realm of crazy alterations in effective chainstay length.

So, what do you think? I'd welcome feedback that I can take into consideration for future design decisions.

gycho77

I think that the dropout is too thick.
It will improve the stiffness, but trackies have to change their rear axel, because the nuts are end of the axel which mess up the axel thread overtime.
Instead of thick dropout try high quality cromoly.
However, the dropout looks good!
And you should add titanium plates

Koogar

Axles should be okay. The dropout body is substantially thicker compared to what's normally used, but it amounts to about 4mm total side-to-side additional width at the axle - should be easily accommodated, based on how long track axles generally are. Maybe I'm not understanding your concern?

I should clarify that the frame tubes are 7005 Al, so all parts welded to them must be the same material; the dropout can't be made of steel.

CP Ti is in the mix in terms of possible insert materials, but it's very expensive and probably unnecessary to achieve what we want - which is something that doesn't destroy wheel nuts, isn't destroyed by wheel nuts, and provides good grip. We're trying another material first. CP Ti may end up as an option.

Thanks for the comments!

carleton

1) I appreciate you guys actually looking to improve on a part of the bike that trackies interact with very often.

2) I think Ti should be an option (even if not the default material. It's proven and easy to find (albeit more expensive). Maybe have them as an add-on since it's a modular system. "Normal inserts $x; Titanium inserts $y" and let the consumer choose to take on the greater expense. Some won't, but some will.

3) I don't think the width will be a problem for modern rear hubs. Only older ones from the era when thin steel dropouts were the norm.

4) While you are at it, I'd consider redesigning the tensioner system. Something faster than screws. I have an idea.


I'd love to see them on a bike and a video of a wheel change in lieu of actually testing myself. Also, you are gonna have to test it without the tensioner in use. As you know, a lot of bikes have tensioners, but they often are not used.

gycho77

Ti option will be great.
Also don't you need more holes to hold the plates secure?

unterhausen

My experience with CP Ti is that it will be destroyed by wheel nuts

700wheel

You do not show how the drop outs are integrated with the chain/seat stays but the thickness at the forward end of the drop-out (where the screw hole is) looks skinny.

taras0000

Although beautiful machine work, that is a dropout that I myself would personally stay away from. I've never liked dropouts with the steel "U" shaped inserts. With aluminum as the underlying material, you can still compress the underlying material. In my opinion, the only proper and best way to do a dropout is to use a plate. Finding a way to integrate a tensioner into a plate dropout is where I would focus my energies if I were involved in something like this.

Koogar

Always great to get your thoughts, Carleton.

We've considered Ti inserts as an option and it's a possibility. The hurdle will be manufacturing volume vs part cost. The costs of custom machining have been a real eye opener, but we're committed to doing the best design we can and keeping manufacturing in the US. These prototypes were machined here in Colorado.

We're planning to document and post details of our development testing. Good idea about including something on the dropouts and wheel change; I'll keep that in mind. And I agree, I want the design to work without using the set screw / tensioner. But I would always prefer to have the tensioner and use it - and anybody with a big sprint should whenever one is available, I think. It makes more sense than a million foot-pounds of torque on the wheel nut.

I'm always open to ideas, as long as all you want in compensation is an acknowledgment

Koogar

Ha, pretty much no part of the dropout is what I'd call skinny, but we'll see if you're onto something - we're going to spend some time beating the stuffing out of them. I'll offer up some pictures of the junction once I get a few tubes attached to these things....

Koogar

I can appreciate where you're coming from, Taras, because I started this endeavor with pretty much the same opinion. My bottom line: we've all been subject to crappy examples of the breed. The compression occurs from over-tightening, which is in turn a sub-optimal solution to an inappropriate insert material (most examples), possibly in conjunction with too weak a dropout body (also usually the case), lack of a tensioner, or just plain ham-handedness. I'm not just criticizing others: I've pulled my wheel a bunch of times and the dropouts on my frame have spread.

Most plate designs entail tradeoffs that keep me from thinking of them as the only proper solution. Tiemeyer's is a better design because it sandwiches the plate in the dropout body, but others don't. Then there's the reality that you're subject to the quality and manufacturing consistency of the bolts (possibly plain old chainring bolts) that you use to attach the plate. If it's a highly stressed part that I'm not in control of, it's another worry. No, you don't see dropouts falling off all the time, but it's the scientist in me. I've also been warned that lot-to-lot dimensional variance in these bolts can be a headache in practice.

I have to wonder, though - if you want a tensioner (and presumably you planto use it), why would you need to crank down on the dropout with aluminum crushing levels of torque?

Baby Puke

Not all bike racers are great mechanics.

taras0000

The aluminum gets crushed because there usually isn't a tensioner present. Your design does mitigate this. The insert doesn't hold well because it is thin, gets deformed, and most are made of cheap steel, so they wear unevenly. When you try to use stainless steel for the design, the insert is harder than the nut, so the nut doesn't bite, and the wheel slips in the dropout. To me, using inserts is a shortcut to doing something properly, and proper in my opinion is the plate. Tiemeyer had a great design (sandwiched the track end plate in the dropout), and BT uses a replaceable plate as well (not sandwiched, but has a reputation of solid construction).

If you want stiffness in your dropout design, my opinion is that a plate would work better than a hooded/shrouded track end. The reason you are machining a big lip/shroud into your design is to increase the surface area of the weld (my presumption) and to buttress the track end beyond the weld. The BT design has way more surface support than any shrouded dropout designed to be welded to tubes could have, and supported the plate almost all the way to the back.

The big upside of using a plate is significantly reduced machining costs (even if you were to included a screw adjuster). This frees you up into using whatever material you feel is appropriate for the track end because your budget isn't being eaten up by hours and tooling on a CNC. Smaller advantages are a stiffer design, being able to use/integrate standard parts (again reducing costs), easily replaceable, being able to use a more expensive/appropriate material (because of less machine time), which may increase durability/hold, as well as offering easy wrench access.

Again, these are just my opinions, but I've reached them because my BT never gave me problems in that area, I was able to get replacements made inexpensively if I had needed to, and I gave up on steel inserts after getting frustrated by them with slipping and deformation (again, probably due to not having a chain tensioner integrated), but then again, I never needed nor missed the lack of tensioner on my BT.

I may have to order some of your track ends in the end tho. They are just too pretty not to look at.

carleton

Baby don't work for free

carleton

I agree with Taras.

Think about this, BT and Tiemeyer have the best dropouts going. Neither one necessitates (or even allows) push-style chain tensioners yet they have been used by the biggest of sprinters with no problem.

Why not model after them?...even if it means spending more money?

First make it work, then start to make it less expensive.

You are starting with trying to avoid high costs, but if you never get it to work...it never works.

Remember, you are making an aluminum bike in a world dominated by carbon. Your price point will probably be much lower than a comparable carbon frame. That's wonderful. Imagine the headlines, "The Aluminum sprint bike that's beating all of the carbon wonderbikes...at half the price." You couldn't keep them in stock! BUT, it has to beat all of the carbon wonderbikes in every area to shut up the naysayers.

So, with that being said, what's an extra $35-50 per frame for ti dropouts?

If you are gonna do it, do it right. You only get 1 chance to make a first impression.


Example:
Serenity's first impression: Not-great dropouts, seatpost issues, and seatpost clamp issues.

carleton

+1

And the sprinters are usually gorilla-handed meatheads

carleton

Ideally, you don't want to use a tensioner at all.

"What?"

Yup. That's right.

As you know very well, in a typical training session, the rider changes gears several times. To put it better, they remove and replace the rear wheel several times. Either installing the wheel after pulling the bike out of the car, switching from training wheels to race wheels, changing gears, flipping over a flip-flop hub, etc... EACH ONE of these events would necessitate fiddling with the chain tensioner. That's not ideal.

- That adds several seconds at the beginning of a chainring change.
- That adds several seconds at the end of a chainring change.
- That adds several seconds at the beginning of a wheel change.
- That adds several seconds at the end of a wheel change.
- That's another moving part to break, jam, strip, or get lost.
- That's another safety check to possibly forget to do.

You know, after being gassed after a crazy hard effort, huffing and puffing while being bent over your bike, blood rushing to your head, sweat pouring off of your nose, trying to get the gear changed for the next heat in the sprint tournament.

It really would be nice not to have to use a tensioner. THAT should be the goal. THAT would make riders very happy. That's the difference between good frames and great frames.

The fallback goal should be a tensioner that is very quick and easy to use. Not a screw (that will eventually become bent at some point) that sometimes needs a small allen wrench to adjust because it's jammed or tight.

taras0000

I don't think going to plate dropouts would be more expensive. Even if you went the Ti route, what you save on decreased machining costs is more than enough to offset the increase in the price of stock. Plain Ti plate isn't all that dear. Hell, aluminum plate is so cheap you could probably afford to include 2-3 sets of end plates with the purchase of a bike, to be swapped out by the user when they wear out. Something like this could easily be cut on a waterjet with steel or Ti fairly cheaply when it comes to machine costs and time.



See how much supported surface there is? That's where your stiffness is going to come from. To improve this rear end, I would maybe add a lip around the the steel attachment plate (that is part of the frame) to increase the stiffness between that junction and the stays. It's not an area that needs constant access, and can easily be reached by any Allen key. If I had carte blanche with a framebuilder as to how every piece was constructed on my bike, I would definitely do it this way. It's bulletproof. I would rather have something like that, something proven and then adapt it to my needs/improvements, than to take an already flawed design and spend time and money on making it bulletproof, only to still have some of the inherent design flaws present.

If your frame is going to be aluminum, then the machine time spent on the frame plate will be less than your dropout design. The track end plate can be cheaply cut on a waterjet. Although not the most economical route, Ti 6-4 plate at 6mm thick can be had on ebay for about $100US. You can probably get 10 plates out of this piece of stock. CP grade is even cheaper, more so if sourced through a supplier. That is all you need. Steel even cheaper still.

6mm Thick Titanium 6AL 4V Sheet 236" X 8" X 12" Grade 5 Plate TI GR5 Metal | eBay

Your design really is beautiful, but it doesn't solve any of the problems that I'm looking to avoid when it comes to track frames. It becomes another pretty bike for me to look at, but that I end up passing up because I see things there that are present in any bike available on the market. At which point, what else is there that's present to put it above the other bikes available?

taras0000

Then what's the point in trying to polish a turd?

You've stated that the custom machine costs of using an appropriate material is too expensive, which leave you with using crappy, pre-made inserts(1). Difference is that you have a push tensioner integrated in the design. So what happens if those threads fail? Stuck with a frame with no tensioner and crappy inserts like the other frames(3). Which ends up with having to resort to ham-handedness[B](4)[B], which crushes the underlying aluminum (aluminum is soft and weak), and that leaves us with point (2). Yes your dropouts are thick, but a soft thick material like will still compress under those inserts.

What are the trade-offs? I've never come across any in my experience. And those chainring bolts are stronger than you think. There's only five of them holding your drivetrain together, yet there are a total of six in the BT design. Why wouldn't they be good enough? How many people and stories do you hear of complaining of chainring bolt failure? Chris Hoy was putting out over 440 lb/ft of torque in his heyday, yet he was never known for breaking chainring bolts.

Koogar

Yes - I've successfully stirred the pot! You know I value the opinions to be had here, so take the following as constructive dialogue. A few global points:

1) I've actually brainstormed with my designer about doing a plate dropout as a potential second version / alternative. That's a maybe year 2 or year 3 project. But that does not mean I believe a well-executed insert design has to be inferior. We've designed these to be the best they can be and we're going to test them extensively. If it is less than dialed, it doesn't go into production. We've had some other way outside ideas we'd love to do for a show bike as well; again, any of these that see the light of day or discussion here shouldn't be interpreted necessarily as a design improvement. Sometimes it's just fun to have variety. Strat vs tele vs Les Paul - would you really choose for the world not to have one, because its features don't fit some use cases as well as others?

2) You guys are WAY off with some of your guesses about cost of production. I've got a bunch of quotes and a few "no quotes" on different designs to back this up. What's your basis? Now let me be clear: the way I think about cost is to consider what it will take to do it well, and to do it here in the US, and if that isn't feasible for a production frame, then I just won't make the frame. None of the design decisions have been taken to cut corners on cost. Cost just isn't driving the design. The one time I raised it is in relation to making Ti inserts and my opinion on that is if another material is just as good or better for the purpose, I don't want to force Ti down people's throats if it's more expensive.

Two points on which we agree I'm evaluating a material right now and it isn't stainless. If that doesn't wow us, the next thing to try is CP Ti. I might try that anyway. But consider - just for a moment - that there might be a superior material for this purpose. I'm open to the possibility. If more than one works equally well, then maybe another is an option. Or maybe not. It's not a fun prospect to sit on thousands of dollars (yes, that number is based on an actual quote) of fancy pants inserts that I can't honestly tell someone is any better.


Remember, Carleton, you and Taras are big sprinters. I'm more or less a sprinter, too. More accurately, I'm predisposed to be a sprinter but I don't focus on sprint workouts and hang out with enduros who give me poor self-esteem. We're the minority of the track population. Most people never will need to use tensioners on any frame. Nevertheless, I'll tell you right now that if I did decide to do a plate dropout design, I'd be working to integrate a tensioner anyway. If I'm going to set up for a team sprint, I'll take the extra time to back up my wheel nuts. Similar logic to why I use toe straps. Also, the tensioners (now you've got me calling them that, even though they don't provide any tension and will actually be acting in compression if they're forced to do work!) can be convenient in helping to get the wheel alignment perfect.

I'm ahead of you on your complaint about the screw - been planning on a thumb screw. There's a specific name for it I'm forgetting at the moment, that's where I rely on the designer. We've been talking about designing a custom screw, believe it or not.


And you'd be really, really unhappy if I made you a dropout using 6/4. It's too hard and slippery - eats up wheel nuts, too. My only point here is that you have to pay attention to the details, and I am.

The two areas we've felt were really deficient in existing frame options are geometry (that's the bigger thing, actually) and the dropouts. We've all been obsessing over the details of the dropouts here because they can be a real pain in the ass and they're a fun bike nerd detail, but in reality, dialed track geometry is going to do a LOT more good for a many more people. But maybe this won't be the right bike for you in any case, that's ok. I appreciate the criticism.

Bad assumptions, here, Taras. It's not a crappy, pre-made insert. It's a custom designed insert for which we're going to test materials until we're happy with the performance. And if we're unhappy with the performance of the dropout design as a whole, it doesn't go into production. I specifically had the threads put into the insert so that if you do ruin them, you only need to replace the insert.

The tradeoffs include having to rely on someone else making the chainring bolts, for one, unless I have them custom made too. Did you know that they vary enough in production that some frames have had to be modified by hand to make them fit? And again, I'm not going to claim the use of chainring bolts is dangerous. But your example about Hoy isn't relevant: the force on the chainring bolts when they're used to hold a chainring on are completely different than the lateral and twisting forces to which they're subject when they're holding the rear end of the frame together.

dunderhi

@Koogar

In looking at your design, I guess the point where the bolt hole is located would be my area of concern with your design. I'm only guessing from the your pictures, but if the bolt hole is 4mm, then there is 6mm of material from side to side and around 2mm of thickness. Using a yield strength of 45kpsi, then your design could bend around 800lbf at that location. The inserts and the axle/nut assembly will both act to reinforce this location, but it might be something to think about, that is if you haven't already thought about it and concluded that it's not a problem.

carleton

You are ignoring my point and saying that, " If I'm going to set up for a team sprint, I'll take the extra time to back up my wheel nuts." Dude, for big guys, EVERY effort is a chance for a wheel to pull.

I pulled a wheel during a rolling jump during a match sprint on a Dolan DF3. Luckily, the wheel just jammed into the chain stay and rubbed until I could slow to a stop.

For sprinters in proper sprint training, most efforts are 100%.

Again, with the Tiemeyer dropouts, I never pulled a wheel...not once. I've had chaining/cog combos that used up all of my chain and put the wheel milimeters from the chain stays and seat tube and the rear wheel didn't touch.

If Mr. Tiemeyer is part of your team, license his proven design and spend your R&D time, money, and energy elsewhere. His dropouts work. Period.

Or, use a design "inspired" by the BT.

Oh, notice a key similarity in the two?

- Solid plates for clamping area
- Held by steel chainring bolts

Seriously, man. The writing is on the wall. Making a fancy sheath/holder for the steel plates that are on dropouts that we all hate isn't the way to go.


Thumb screws (and long screws in general) are just as bad, watch:

Your dropouts are 1.5 inches of usable range. That means that he chain tensioner screw has to extend into the dropout 1.5". Then you need length to get it through the dropout (and into the rear triangle). That's another, what? 0.25"? So now we are at 1.75". Then you need some extra to allow for manipulation of the thumb screw, so another inch? So 2.75"? 3" to make it a round number.

Interesting thing about this thin 3" screw:

1) When fully engaged, if it's not 100% even with the threads, it could be bowed up or downwards if the wheel nuts slip. (This is how they get bowed). Now it's all not-straight and tough to get in and out.
2) When you have a chainring/cog combination that pulls the wheel all the way up into the dropout, the tensioner bolt has to be backed-out almost completely. But, here's the rub (literally): It does not sit evenly with the chainstay. Nope. It angles outward from the bike. So, if you are like me and wear size 46cm/12 shoes and you like to have your feet as close as possible to the bike, that means that your heel will clip this thing that's 3" long and 0.5" wide sticking outwards from the rear of your bike.



The best dropout + chain tensioner (or whatever the proper name is) is the one that does not require a tensioner to keep the wheel in place.

Tiemeyer and BT have proven that it is possible. Follow their lead...not the other guys.

Koogar

Dunderhi, thanks, man! In an earlier iteration, my designer did not want to put the spot facing in for exactly that reason, though I get that's not the only consideration. I think @700wheel may have been getting at the same thing. The conclusion was that it's not a problem in the current iteration, but actual testing takes precedence over modeling, so I'll take another look at that area and keep a close eye on it during testing. We're hopefully no more than a few weeks out from getting frames on the track. Thanks again for catching that!

carleton

Lots of edits in my post above, sorry.

Koogar

Carleton, as always there are well thought-out words and a picture to go with the point you're making But since it's past midnight here in Colorado, let me put it this way briefly before I go to bed: 1) a design that has the tension screw interfering with your shoe is not something I'd put into production, so it's not a concern, 2) I can tell you it's not a licensing issue, and 3) let me turn this around and point out that you're not entertaining the possibility that anything other than what you've known in your experience to work, will work. I'm totally open to ideas for improvement, which is why I asked in the first place. Lots of great ideas for a plate dropout. But I'm an experimenter at heart, which is why we're building prototype frames and being transparent about the process. Would you really throw $1k of dropouts into the garbage without even trying them, because you hypothesize that there's no better way to do it? Or do you think we shouldn't have tried anything other than copying a known design? Are you THAT convinced about your position? I don't have that much confidence in something I haven't tested, and I'll have no problem admitting it if I'm wrong. I fact, I'll post it all over, because it's testing that other people don't do, and I'll have the data to show that I'm wrong from trying, which is perfectly fine to me and more than others have done.