Page 1 of 2 1 2 LastLast
Results 1 to 25 of 43
  1. #1
    mtbr member
    Reputation:
    Join Date
    Aug 2011
    Posts
    16

    designing a 3 speed mountain biking hub

    Okay this is going to be a long and somewhat technical post that most people will find an absolute bore fest, you've been warned!

    First off, great community here, I've lurked for a loooong time and have read an incredible amount of posts, thanks for all the info guys!

    So I have been completely obsessed with designing and building my own 3 speed geared hub for months now, I've spent literally hundreds of hours thinking about this, drawing diagrams, and researching planetary drivetrains. I'm completely serious with machining my own gear hub from scratch. (Don't worry I have the CAD and machining side covered... hopefully)

    So I have a couple questions, First, what would be your ideal 3 gears (in gear inches)? I was thinking like a 22-25" super low gear, 48-53" general mountain gearing (this would be direct), and an overdrive around 68-72". The reasons are pretty obvious, I can get up most things with 50" unless it's ridiculously steep and long. And, I am also frustrated when I spin out, so I want something that I can cruise on the roads with. What would your guys's picks be? There is obviously a massive gap between 25" and 50", but do you really need to dial in your cadence while going uphill? Mash or Spin your balls off, make your choice.

    So without even a cursory explanation as to why, other than: it's design is simpler, it would be lighter, and more efficient. What do you guys think of having the lowest gear being only accessed by reverse pedaling? i.e. when you shifted from 2 to 1 you would still be "in" 2 (direct drive) however when you back pedal you would magically get 22 gear inches of forward movement. Anyway I tried to keep this as short as I could, let me know what you guys think. And please ask questions if you want, I left pretty much everything about the design out.

    Thanks for reading,
    -Jim

  2. #2
    mtbr member
    Reputation: vaultbrad's Avatar
    Join Date
    Oct 2007
    Posts
    792
    To figure out if you like the idea of reverse pedaling try building yourself a retro-direct drive bike. I love everything about this whole idea. Please pursue this with vigor.

  3. #3
    mtbr member
    Reputation:
    Join Date
    Aug 2011
    Posts
    16
    I'm glad I got a reply, I thought I was relegated to some crazy person who thinks that reverse pedaling is at all a good idea.

    Anyway, glad to see someone is taking interest, my first adventure will be machining a single speed hub. I figure I gotta start somewhere, and I will document this so don't worry. I really would love to make it with a roller clutch, but torque specs on off the shelf roller clutches are cutting it dangerously close at 120 Nm maximum torque, and I'm not about to go down the rabbit hole of making my own roller clutch. For reference rohloff specs 100 Nm but also says that their minimum gear ratio is 2.1; so I could possibly get it to work. Does anyone have any advice about this? My second choice will be machining up essentially knock off DT Swiss star ratchets, because I'm thinking I might make use of that design in my geared hub, and in all actuality it won't be that hard.

    And yes, I will probably cobble together a retro direct bike, and since my schwinn world sport is languishing in the corner of my garage, it seems like a perfect candidate. If the reverse pedaling this turns out to be a bust (I'm not holding my breath), then I have a couple other ideas for how to achieve the balls out low gear, but none of them half as elegant.

    Thanks for the encouragement!

  4. #4
    PeT
    PeT is offline
    Occasionally engaged…
    Reputation: PeT's Avatar
    Join Date
    Jan 2004
    Posts
    1,410
    In grad school I had a "kick-back" hub on the Schwinn commuter (before it was stolen -- everything parked outside in Beserkeley gets stolen eventually), but you're talking about pedaling backwards anytime you want to spin your 25 inches? That would seem rough biomechanically, particularly after a lifetime of powering a bike by pedaling forward. Definitely something to try out in some sort of simple mock-up before committing real resources. But that being said, I've long thought that a 3-speed internal hub with the ranges you specified would be awesome, and your picks for the range is spot on (IMHO). This effort will be worth following!
    "The plural of anecdote is not data." -- Attributed to various people in a variety of forms, but always worth remembering...

  5. #5
    mtbr member
    Reputation:
    Join Date
    Aug 2011
    Posts
    16
    Ya, the problem is that with a 50" input, 25 inches is impossible without putting power through two gear assemblies or using some sort of stepped gear, which seems overly complicated for a three speed. Also, any 3 speed with this range requires two entire gear assemblies (rings, planets, and suns), which again is complicated and heavy for a 3 speed. However, driving the sun in reverse, with the planet carrier held stationary (anti-reverse freewheel), gives you your 22" out of an assembly that is also usable as a 70" overdrive.

    Yes, how I would set it up is that in first gear pedaling forward is still your direct gear (50") and reverse pedaling would be 22". Ergonomics is my chief concern, although it's 22 gear inches and personally I would only be in it literally 2% of the time. It's really just an alternative to getting off the bike and walking up, so if it's not biomechanically efficient I really don't care. As long as it doesn't hurt my knees I'm fine. Plus, reports online about retro-direct drivetrains say that they are just as comfortable to pedal, I'm just worried about not being able to clear technical climbs while pedaling backwards, hitting a root while spinning your pedals forward might suck.

  6. #6
    ~ B A D A S S ~
    Reputation: car bone's Avatar
    Join Date
    Apr 2011
    Posts
    2,022
    I have a true precision stealth with a roller clutch /one way bearing. If you search my posts and search for true precision and/or stealth one of the posts will come up that shows the make and model of that component. The one thats in the hub is the second biggest model.

    Also that particular components max torque is dependent on the material thats outside it, basically the thickness of the material you put outside it is what determines the maximum torque noting else really (until it defelects). the specced max torque is without anything at all (from the bearing manufacturer). thicker stock there = more torque can be trasmitted. you have to crack the hub to make it malfunction.

    If you look at a true precision hub you will se that there is a thick diameter and a thin one. where the one way bearing is, its thick. really thick.
    Rule #9 // If you are out riding in bad weather, it means you are a badass. Period.

    Quote Originally Posted by iheartbicycles View Post
    Specialized sucks ass.

  7. #7
    ~ B A D A S S ~
    Reputation: car bone's Avatar
    Join Date
    Apr 2011
    Posts
    2,022
    i think its an INA HF3020 inside true precision stealth Drawn cup roller clutches - HF
    Rule #9 // If you are out riding in bad weather, it means you are a badass. Period.

    Quote Originally Posted by iheartbicycles View Post
    Specialized sucks ass.

  8. #8
    mtbr member
    Reputation:
    Join Date
    Aug 2011
    Posts
    16
    Thanks for the link and the info! I did a bunch of research and never stumbled upon your posts. I didn't realize that they just toss an off the shelf roller clutch into their hubs, I thought they had to have some proprietary roller bearing assembly in there like shimano does, but it makes sense why they wouldn't. I may be over thinking this, perhaps I should just build it with the roller bearings and if I break it, whatever. It just gives me a chance to build another.

    As far as the housing being the limiting factor; I could be wrong, but after reading the design and safety guidelines on the site you linked. It looks like your housing can be as thick as you want, but you can only put so much torque on it until you start making indentations in either the shaft or the drawn cup, which makes sense. I.E. Stiff does not equal Hard in Engineering terms.

    All of this is well and good, but true percision trusts their design enough to throw a cassette on there and put it up to the task of 22/36 low gear torque, and still have a 5 year warranty. So we will see.

    I really like the idea of a silent clutch, because the beauty of having a reverse gear is that it can always be freewheeling and all you have to do is reverse pedal to realize that low gearing. However with a ratchet system, always freewheeling means that your hub is always clicking, no matter what. If I utilized silent and near instantaneous freewheels I could essentially make a hub that only had 1 pawl that you could shift up or down to engage your sun gear and give you an over drive gear. Not to mention machining a hub this way would be waaaaay easier. The only real penalty, assuming these clutches are up to the torque, would be weight, and that would be counterbalanced by simplicity of design.

    Thanks for the info! I'll be sure to creep around your posts some more, but I think my first step, as I said earlier, is going to be building a single speed hub. I think I am going to test out a roller clutch in it and see if my 6'4" 220 frame can't destroy it.

  9. #9
    ~ B A D A S S ~
    Reputation: car bone's Avatar
    Join Date
    Apr 2011
    Posts
    2,022
    I would go with the biggest one way bearing. The axle in the stealth hub is alu with a pressed on polished steel ring/race where the bearing rolls. I'm guessing its handened steel.

    For the hub housing I recommend using stainless steel such as 17-4ph or 17-7 or even something more badass. "stiffer" than alu and hardenable by aging.
    Rule #9 // If you are out riding in bad weather, it means you are a badass. Period.

    Quote Originally Posted by iheartbicycles View Post
    Specialized sucks ass.

  10. #10
    ~ B A D A S S ~
    Reputation: car bone's Avatar
    Join Date
    Apr 2011
    Posts
    2,022
    I think you get get them without the drawn cup too, and make your own "drawn cup" how thick and hard you want.
    Rule #9 // If you are out riding in bad weather, it means you are a badass. Period.

    Quote Originally Posted by iheartbicycles View Post
    Specialized sucks ass.

  11. #11
    mtbr member
    Reputation:
    Join Date
    Aug 2011
    Posts
    16
    I was planning on mounting in in 6061-T6, I haven't spec'd how thick yet. The problem with the housing being stainless is that, if we're talking about the single speed, the housing is the hub shell. With the 3 speed set-up, the housing would be the planet carrier, which could be stainless, it would just be heavy. I guess I'd rather do aluminum for this reason, but we shall see. I guess, I could always press a sleeve between the shell and the bearing to add rigidity. I'll have to do some calculations to understand what is necessary there.

    The basic idea on the single speed is pretty simple, if I'm doing roller bearing. It's basically the same as the stealth hubs. I'll machine the hub shell out of a solid piece of 6061-t6 on a lathe; same with the driver. A one way roller clutch is press fit into the hub shell, and the driver is essentially something to attach the cog to (I'm thinking disc brake bolt pattern, and getting the cogs via velo-solo or some such) and a shaft that I press an inner bearing race onto (off the shelf bearing race, 52100 Chrome Steel). And then some bearings placed between the driver and the hub shell, and on the axle of course (I am also going to turn the axle myself as well). This is a rough outline of what I am going to do. If anyone wants an in-depth drawing or something, I will be making that up in CATIA over the next couple lunch breaks.

    After I have cut my teeth on the single speed hub, I am going to transfer pretty much every one of these ideas over to the 3 speed, with a ton of additions of course.

  12. #12
    ~ B A D A S S ~
    Reputation: car bone's Avatar
    Join Date
    Apr 2011
    Posts
    2,022
    The tp uses a split ring thats threaded (threads onto axle axially and locks with a small hex bolt radially) as a preaload for the bearings. In the TP the clutch is pressed in or possibly cooled in with LN or dry ice.

    Steel is 3 times as "stiff" as alu, and 17-4 about 3-5 times as strong, its a zero sum game there really, but its much harder and will last longer, I mean the "cost" here is not the materials per se, its your time, because this will take lots of time.

    also look into Latrobe bg42 steel for races and pretty much all wear parts. its stainless bearing steel, the only real stainless bearing steel designed from the start for that application, much better than 440c. Its a martensitic vim-var stainless hss/tool steel. Instead of the 52100 I mean.
    Latrobe Lescalloy® BG42 VIM-VAR High Performance Bearing Steel

    For a cog carrier/freehub of some sorts I would use sandvik 12c27 or similar (but not 420hc), its extremely fine grained, low carbon SS, very tough (since its low carbon/low Cr and very fine grained)
    Like a SS 4130 kinda. The easiest thing to do is just use the shimano cassette pattern, since then you have unlimited amounts of cogs to try. and its easy to swap cogs.

    One product I would like to see is a 3sp crank, like that good german one, Schlumpf Drive, but 3 speeds instead. But If I want to see that product I have a feeling I have to make it myself.
    Rule #9 // If you are out riding in bad weather, it means you are a badass. Period.

    Quote Originally Posted by iheartbicycles View Post
    Specialized sucks ass.

  13. #13
    mtbr member
    Reputation:
    Join Date
    Aug 2011
    Posts
    16
    Dude... The information you are laying down is awesome... I see your point with the stainless, and I agree that the actual cost of materials is only a small part of this gigantic project, so I might as well build it with the utmost quality in mind. My question is whether or not I actually need the stainless steel housing. You may be right by saying weight wise it's a wash, but machining Al is pretty damn easy. With the plan I have, everything can be put together with an interference fit; the inner bearing surface I spec’d is HRC 60-64 can be pressed onto my driver body. This means I can just turn my Al driver on the lathe, and not worry about precision grinding an inner race. My goal here is making machining and assembly easy, and making a hub that works well. It doesn't have to last multiple life times. We’ll see how this all shakes out though, once the design is finished.

    The reason I like the 6 bolt pattern is that it seems more robust for holding a single cog onto your hub, and it lends itself better to an aluminum driver, with shimano splines I would definitely take your advice and make it out of some high quality knife steel. All of the drawbacks associated with the 6 bolt pattern do not really apply to me in this instance, so I am probably just going to go ahead and do it. I can find all the cogs I need in 6 bolt pattern, and changing cogs just takes an allen wrench.

    As far as 3 speed crank, I thought about it for a long time, and I couldn't think of anything. It's mainly a packaging issue. A 3 speed bottom bracket is a horse of a different color though. Your main drawback is having to build a frame around it, which depending on how you look at it may not be too much of a hurdle. I actually really like the idea of a 3 speed bottom bracket since you can run whatever you want out back. I've seen things like this but most use a gear box design like you seen in manual transmissions. Although I think you could use an epicyclic gear system.

  14. #14
    ~ B A D A S S ~
    Reputation: car bone's Avatar
    Join Date
    Apr 2011
    Posts
    2,022
    If you want to do a proto quick and dirty i would recommend 6061 alu/4130 steel/52100 (if you can get that easily) and some marine brass (it machines really fast). you will probably have to make at least one hub that turns out not so good, when its finished you will see lots of things you want to change for the next version.

    I think I saw a schlumpf drive geared bottom bracket yesterday. On flicker i just wrote schlumpf drive and there it was in a dudes hand, some real chunky bit with splined axles sticking out for crank arms.

    Any way good luck with machining the parts.
    Rule #9 // If you are out riding in bad weather, it means you are a badass. Period.

    Quote Originally Posted by iheartbicycles View Post
    Specialized sucks ass.

  15. #15
    mtbr member
    Reputation:
    Join Date
    Aug 2011
    Posts
    16
    Ya, I guess I had more of a prototype mentality. I don't want it to be a throw away, but I don't want to spend a ton of resources on something that I will probably make improvements to. I'm also keeping in mind upgrading parts down the road. I can build these things quick and dirty, but I can always build more robust housing, or change out my driver if I want. Barring an entire rework of the hub most parts I can come back to later once the proof of concept is done.

    Your suggestions are invaluable btw, I'll definitely keep you posted on how this goes.

  16. #16
    mtbr member
    Reputation: vaultbrad's Avatar
    Join Date
    Oct 2007
    Posts
    792
    I love everything about this thread. Great conversation and ideas that are often lacking on MTBR.

  17. #17
    will rant for food
    Reputation:
    Join Date
    Jan 2010
    Posts
    2,880
    Quote Originally Posted by vaultbrad View Post
    I love everything about this thread. Great conversation and ideas that are often lacking on MTBR.
    Same here.

    I like the way you think OP. I have my own musings about tailored drivetrains, but actually getting them made is pretty difficult. Nice thing is you can avoid irritants or come up with compromises you're actually willing to accept. For instance my experience with the Hammerschmidt crank, I couldn't fathom why they'd put such an anemic cable clamp on an otherwise tank-like unit. The thing was a beast, and then, what the sh!t is this little thing?

    I'd love having something like a three speed Hammerschmidt that isn't overbearingly sealed like a HS and actually spins smoothly. But getting that "third" (second) planetary gear going makes things complicated.
    Latitude: 44.93 N

  18. #18
    Hooligan
    Reputation: dirtdawg21892's Avatar
    Join Date
    Jul 2009
    Posts
    367
    this looks awesome, i'm really interested to see how this turns out. ive always thought that i could ride comfortably with only 3 well spaced ratios like you described (climbing, flat, and descending) internal hubs are great, and i would love to run one if it could stand up to the torque and excessive abuse that it would see on a mountain bike.
    Quote Originally Posted by buckfiddious View Post
    Specialized builds their bikes out of homeless children.

  19. #19
    mtbr member
    Reputation:
    Join Date
    Aug 2011
    Posts
    16
    Quote Originally Posted by car bone View Post
    For the hub housing I recommend using stainless steel such as 17-4ph or 17-7 or even something more badass. "stiffer" than alu and hardenable by aging.
    So a quick update, I think I'm moving away for 6061 any part of the hub, after I did some calculations for the housing, it's just not strong enough, I'd have to make the housing extremely thick to make it work. I think 7075 - t6 will work really well.

    Anyway, the design is all but done. I skipped right past the single speed hub, I liked the idea of doing it first, but I really just wanted to get started with the three speed. Do you guys have any interest in seeing the insides of the hub before I build it? I could put up a Youtube video or something like that. I'd love someone else to critique my design before I get carried away with myself.

  20. #20
    mtbr member
    Reputation: vaultbrad's Avatar
    Join Date
    Oct 2007
    Posts
    792
    I love seeing sequential build shots. I'd like to see the guts when you get it all together.

  21. #21
    will rant for food
    Reputation:
    Join Date
    Jan 2010
    Posts
    2,880
    Quote Originally Posted by jimhead88 View Post
    Do you guys have any interest in seeing the insides of the hub before I build it? I could put up a Youtube video or something like that. I'd love someone else to critique my design before I get carried away with myself.
    Absolutely.
    Latitude: 44.93 N

  22. #22
    ~ B A D A S S ~
    Reputation: car bone's Avatar
    Join Date
    Apr 2011
    Posts
    2,022
    Quote Originally Posted by jimhead88 View Post
    So a quick update, I think I'm moving away for 6061 any part of the hub, after I did some calculations for the housing, it's just not strong enough, I'd have to make the housing extremely thick to make it work. I think 7075 - t6 will work really well.

    Anyway, the design is all but done. I skipped right past the single speed hub, I liked the idea of doing it first, but I really just wanted to get started with the three speed. Do you guys have any interest in seeing the insides of the hub before I build it? I could put up a Youtube video or something like that. I'd love someone else to critique my design before I get carried away with myself.
    The 7075 t6 is about twice as strong (yeild) as 6061, but not any stiffer, its about 80GPa or so I guess for both of them. the 7075 will be more "brittle" and prefers to crack instead of bending. however you would need a twice as big load to do that in the first place. I heard 2000 series is good if you are expecting the failure to be from fatigue.

    My precious #1 | Flickr - Photo Sharing! read the text there, very cryptic imo, its a hub??
    designing a 3 speed mountain biking hub-4518667704_69d61854b8_b.jpg
    Rule #9 // If you are out riding in bad weather, it means you are a badass. Period.

    Quote Originally Posted by iheartbicycles View Post
    Specialized sucks ass.

  23. #23
    mtbr member
    Reputation: vaultbrad's Avatar
    Join Date
    Oct 2007
    Posts
    792
    That's a schlumpf 2 speed unicycle hub

  24. #24
    mtbr member
    Reputation: vaultbrad's Avatar
    Join Date
    Oct 2007
    Posts
    792
    Bump for updates. How goes your progress?

  25. #25
    mtbr member
    Reputation:
    Join Date
    Aug 2011
    Posts
    16
    It goes well, I've been rediculously busy and have been pretty well unable to do much during my knee surgery (damn you soccer!), but I have been finding time to finish off the design.

    tl;dr I ran in to some design issues, and am sacrificing shifting under load and smoothness of shifting for mechanical simplicity.

    I ran into an unforeseen road block when I was dimensioning things out in CATIA. I was planning on doing an internal clutch system on the axle, like you see on Rohloffs and Alfines. However, since my sun and planet gears are close in size I am able to scale everything way down. Which is great for size and weight (all told the hub body will be about 60mm in diameter), except that with a module 1 22t sun gear you really don't have the space for a clutch system inside of the gear itself. Not wanting to scale the whole thing up, I explored some other options.

    What I think I am going to end up doing is making the whole sun gear slide from side to side and use a dog clutch system on either side of the sun gear to get my desired power transfer. One side would be gear 1 & 2 (reverse pedal and direct gear). Then when it slides to the other side it will mesh with teeth on the axle giving you overdrive. This dog clutch system is essentially the exact same thing as the gear change on a manual transmission. I'm not a huge fan of the gear itself moving, however it greatly simplifies the internal mechanics so I am probably going to stick with it. This will unfortunately make changing gears under load pretty well impossible, but I haven't quite found another option that I like. The real problem is the space. This design sacrifices smooth shifting under load, for greatly simplified design and size requirements.

    This problem kind of encapsulates almost every decision I have made from the start of this project. I am not making it fool proof and I'm potentially sacrificing some ergonomics, but I gain a simple and robust design.

Page 1 of 2 1 2 LastLast

Similar Threads

  1. 9 speed/10 speed, Shimano/SRAM, Road/Mountain Compatibility Issues
    By Hundun in forum Drivetrain - shifters, derailleurs, cranks
    Replies: 15
    Last Post: 04-10-2014, 03:17 PM
  2. Winter mountain biking, not fat biking, on front range?
    By Training-Wheels in forum Colorado - Front Range
    Replies: 5
    Last Post: 09-23-2013, 08:29 PM
  3. Designing mountain bikes...
    By Rusty_Venture in forum General Discussion
    Replies: 25
    Last Post: 04-13-2011, 02:42 PM
  4. Designing mountain bike frames
    By Rusty_Venture in forum All Mountain
    Replies: 8
    Last Post: 04-13-2011, 09:49 AM
  5. Designing mountain bike frames
    By Rusty_Venture in forum Frame Building
    Replies: 15
    Last Post: 04-13-2011, 06:19 AM

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •