Idea for an easily customized frame
 

Hello everyone, I am new to the forum and I have something that I would love some feedback on. First a little background, I have started designing and building bike frames two years ago and what I am currently working on is a fully demountable road bike for myself. It is a bit basically a cross between René Herse's demountable and Caminargent's demountable bikes.

The system is a set of clamp on quick release lugs that attach the tubes to form the frame. With the exception of the chain stays, the bottom bracket, and the dropouts which will be brazed together to form what will become a rear fork. Additionally the seat stays will be single gauge steel tubes with a consistent diameter all the way down and will be bolted onto an articulating drop out and onto a corresponding articulated joint on the down tube top tube lug. The cabling on the final version will be internally routed through the down tube and will route to the front/rear derailleur and the rear brake which will be mounted on the chain stays (hopefully, pending chain clearance.)

I have completed the second draft of the lugs that so they are fairly close approximations of what will be used for the bike itself. Currently I am working on the next iteration to send off to be printed for fit and then will be having them cast to become a functional prototype of a bike (having done the calculations in autodesk it is theoretically structurally sound as an assembly but one can never be too careful.) The quick releases aren’t the final versions either they’re from Lee Valley and the lugs themselves are currently made out of plastic for fit. What I am currently thinking about the project is that it could be easily customized for others or refitted for different riders.

Any feedback, concerns, or input about the project would be greatly appreciated.

http://bikeforums.net/attachment.php...hmentid=334922http://bikeforums.net/attachment.php...hmentid=334923http://bikeforums.net/attachment.php...hmentid=334924http://bikeforums.net/attachment.php...hmentid=334925http://bikeforums.net/attachment.php...hmentid=334927http://bikeforums.net/attachment.php...hmentid=334928

nice use of a 3d printer. I am not a big fan of having that many moving parts in a frame.

That's very cool. I don't think the headtube and seat tube/seat lug needs to be detachable, and I'd probably leave the DT-HT joint brazed (it's highly stressed). I'd would also likely increase the clamping area and use two binders (or a single M8+ bolt). I'm not sure why the seat stays need to have a side to side pivot? I might be simpler to bend the socket.

I think the idea is to be able to change sizes. I'm not sure that makes sense for a single frame

Thanks for the input so far, it is invaluable. Also sorry for being MIA.

The reason for so many moving parts and I know this by reputation only is that the original Caminargent frames had issue with the lugs not being tightened correctly. Though they came from the factory correctly adjusted the major feature was their ease of disassembly, but without a torque wrench to re-bolt the assembly it could become unstable. Like attempting to add components to a carbon frame the tubes could be crushed or the assembly could be under tightened and come apart in use. With quick release cams the torque can be set and left alone.

The idea isn’t so much to have the ability to have multiple sizes on a single frame. More that the lugs could be used to build frames for different riders because the system is flexible and final assembly is greatly simplified—the frame is mostly clamped and pinned into place it only requires that the tubes be mitered instead of being brazed, filleted or welded into place. The other considerations for the frame were dead easy repairs and replacement of the parts; and being able to break the frame down for storage and travel.

The final version of the lugs will have two cam levers on either side of the lug it will exert ~900-1200lbs of force to the clamping surface depending on the lug.

The final design for the levers (not the current Lee Valley ones) have a M5 bolt through the centre to pin the assembly in place (also to make it more theft proof...) I will add a export of the final design of the lever. As for the clamping surface, I was worried about that too in the first draft of the design I made it longer but it didn't make it more effective in theory or practice, the lugs and the tubes on the second draft actually took more weight to separate than the first ones with the longer clamping surface.

I do like the idea of keeping the head tube lug the top tube and seat tube lug together in one assembly... I feel kind of dense for not thinking of that. It makes it cleaner and reduces the overall number of parts.

The seat stays have a side pivot to make assembly easier. On the first draft of the project I attempted to use a bent seat stay cap but aligning the seat stay and the dropouts during assembly wasn't fool proof so I switched to the pivoted design to make it easier.

By the way tux love the look of your mixte frame, and your demountable is very cool how did you do the sleeve by the bottom bracket and the coupling? Looks a bit like a combination of Rene Hearse and Breakaway.
http://bikeforums.net/attachment.php...hmentid=335020

If you've made some test re: the clamping surface then that's good. Was it on steel lugs? I guess if you look at handlebars clamps they aren't too big yet are sufficient.

In my demountable I used a externally butted seat tube as the downtube; I wasn't very confident in clamping and pinning a thin-walled tube. Turns out it's a perfect slip fit in a 1.250x0.035 tube, nice. I saw a similar design in a Nobilette frame. And yes I used the Ritchey idea at the seat tube; I think it's brilliant to use the seat post as part of the coupler.

If you want more ideas you can check English cycles.

Yes I did, I did a single test with a 3D printed stainless steel lug (I would use the term stainless steel loosely because it is bronze infused) of the first design and the second design. What I found is that in the first design with a longer sleeve the clamps exerted a lower clamping force consistently over a longer surface; the second had a higher force directly at the lug. So the longer sleeve took longer to fail, but it failed at a lower weight. The second design when it failed, failed at a higher weight but took slightly less time to fail. Granted both of these tests were static, and I can only rely on the computer to show the dynamic tests because I don't have a EFBe Bicycle test style jigs to help me.

And exactly if you look at a lot of the clamping on bicycles, it is done over relatively short distances with greater force. With a sleeve it is more about ensuring that the clamping is consistent over a longer area because the material can distort. Both are stable, they're just different approaches to the same problem.

And no it is a brilliant design, I just don't have the machines to do it. You were also experimenting with a S&S style coupler? Did you ditch it just because of the weight? Or was it annoying to machine as well?

Woah it looks like you've done a lot of work, awesome. I'd love to see more about the computations and SS lugs. I also wish there were fatigue testing machines easily available; I had to send some parts in CA some time ago to check my stuff.

Yeah I tried to adapt a $20 pipe coupler at first, but it was way heavy, would have required a lot of machining, and it was ugly. The S&S parts are stunning in form and function (you pay for it), but I'd say they are a bit overkill, and a bit heavier than other systems.

If the point is to make the frame easily assembled like the Caminargent then I guess you would want to avoid the brazed parts. I can imagine it would be hard to produce the lugs in different angles to accommodate different riders or wheels. If you can accept some brazing and rather focus on the storage/travel feature then I think fewer clamps would be simpler?

What are you thought in clamping thin-wall (0.9 mm and under) tubes? H-bars, steerers and seat posts are generally thick...

smugglers can fill the tubes since they're clamped joints.



I suppose the stuff burns away, so you can do lost plastic investment castings ..
in metal.

Hi, I am not an engineer, but I think what you are doing is going to get someone hurt. In 2006 I was building prototypes that would eventually become the Ravello two-piece aluminum travel frame. You have to understand that a bike frame is a type of truss. When seated on a bike, your weight is pushing the bottom bracket towards the ground as your weight is suspended between the two axles. Secondly, (especially when off the saddle and cranking) there is a whole lot of twisting force going on with the down tube with the alternating leverage of your body weight being applied to the crank arms. It would help if the down tube was square. but a junction there still will need a lot more than squeezing force to insure that it doesn't just end up pulling out. Another problem is with alignment; I don't think the frame with your junctions will stay in alignment. Just my 2 cents.

Yeah... Dynamic stresses would be great to evaluate in person rather than sending it away. But I’m just out of university and poor lol. I can see if I can get Autodesk to nicely output the data... But failing that I can get Solidworks to output and show you the data for the next iteration.

What I am really trying to strike is a balance between brazed and clamped; I want it to be easily and quickly constructed... But I also want it to be economical. The thing about angles is that for lugs as you already know there isn’t a great deal of them. A slightly increased head tube angle can make for a jumpier bike, but a properly raked fork can take that away. What I mean by that is there are multiple solutions to any problem really it is just finding them. The rear triangle was causing me pain until I realized that the angle of the seat tube was fixed, so the chain stays were also fixed to keep the wheels at the same distance. As with anything it can’t be 100% flexible and satisfy everyone.

For thinner tubes it can be an issue, it is a fine line between crushing them and securing them. I learnt that first time around. I might continue with the Reynolds 525 and see how it holds up to actual riding because the failure was beyond requirements; or I might end up using the Columbus Tubing meant for welded frames with thicker butts on the ends because the profiles are closer to what I would consider ideal (but not necessarily required.)

Handlebars and seat posts aren’t dramatically thicker they range from 1-1.5mm (just looking at examples I have in me immediate vicinity); but they are aluminum and aluminum has an entirely different material profile compared to steel. Compensating for the weaker properties isn’t done exclusively by thickening the tube mostly it is done by increasing the diameter which makes it stiffer. I don’t know about steering columns, my best guess is that the thick tube is to accommodate the threading for a threaded headset? Because having seen their carbon counter parts it isn’t entirely for strength. ß Not a strong comparison I know, carbon’s strength is directional so it could be an engineering feat.

Actually... Bikes have been used for smuggling in the past, Gino Bartali used the top tube of his bike to smuggle documents to aid the Italian resistance during WWII. Quiet a cool story, then again most anything can be used to smuggle something, and generally the tactics organized crime uses aren’t necessarily humane.

Brian25 very cool, what was your role in designing the bike? It seems to be a solid system. Bikes are a fixed truss and I do know some of the complexities of them already. There was actually a very good thesis written at Ryerson about the Dynamic Forces on a Bike Frame. I would have to search it up but it outlines most of your concerns about the bike. The final plan for the frame is to pin the tubes in place through the quick release levers as a failsafe to prevent movement. It is what the holes in the lug bodies are for and the hole through the centre of the lever. And I am in Canada, if I scrape my face and smash my head we have our mighty socialized healthcare! Muhahahahhahaha... I kid, I am taking all safety into consideration but I am working off an older Idea for the project and there are still serviceable examples.

That squares (!) with the Carminargent design: they used hexagonal tubes IIRC.

Old productions frames (like the CCMs that used to be manufactured around here in TO) had parallel angles and a fixed TT lenght. The only thing that changed in the three sizes available was the head, seat tubes and seat stay lengths. Same TTs, DTs and CSs. Not ideal but it sorta works I guess. Even if you keep a fixed head and seat angles,with the same fork height & BB drop, as soon as you change the TT the DT-HT angle changes. So another reason to braze the DT-HT.

Having not round tubes clamped is great to transmit torque. Even better is splines. But having a pin/set crew can be sufficient if well designed; it's used in heavy machinery.

What are the main design objectives in rank. I can see vintage thing; personal preference; production concerns (for you or for future commercial?). Assuming demountable road bike.

Just like to have a clearer idea of why this thing sits where it does. A lot of these discussion go down a lot of alleys before we really get to the objectives.

What's with the rear drops? Are you running gears, I am always out of touch on the latest styles

Also, I assume you have looked into what happens when you cast these things since they will not come out the same size at your printed lugs. There is an art to how parts move in the casting process, so as to reduce post casting machining. Dazza and Ritchie have been through it, and possibly got some help from

Coincidentally one of my favourite frames was one of those production CCMs from 1981.

Can’t you maintain the top tube angle and down tube angle because they are shortened proportionally? That would control the reach and then the parallel angles on the seat tube and head tube would control the step over height? The head tube would be longer but the angles stay the same... The fork and bottom bracket height/drop are really what cause the problem. What I might have to do is build a fork, but I have never done that before.

The original Caminargent's tubes were octagonal aluminum, it would be great to emulate that but it is not feasible on a small scale, so I will have to make up for the lost torque by including a set bolt, while actually I guess the functionality would be closer to a clevis pin.

The main design concerns right now are; personal preference, production concerns for me (I don’t have a full shop, or much space) and possible future commercial (a few friends and family asked if they could have one made) and then vintage. The vintage ones are cool, but I am trying to improve on them slightly, because of changes in technology it only makes sense to alter some of functionality of the frame.
I’ll be running gears, the drops were designed initially to be a track style drop and accommodate a derailleur with a separate hanger. I’ve abandoned that idea and on the design I’m currently working on they are a more standard vertical dropout with the normal accommodation for a derailleur additionally I`ve switched from the socket style attachment for brazing to an internal peg style for brazing.

I am aware of the plastic parts cannot be cast as is. My grandfather was an engineer and designed oil rigging equipment that was sand cast (I know it is different process and has different tolerances) but he made me aware of the difficulties that can be encountered in these sorts of processes. For actual functional prototypes I am looking into a few different avenues, CNC machining seems like a good option, or investment casting the obvious choice, but also digital laser metal sintering for some of the smaller parts. For rapid prototype investment casting firms evaluate the part and adjust it accordingly and then they “print” a 3D wax replica, which they then use for the casting process.

Dazza and Ritchie? A quick Google search says that Dazza might be Darrell Llewellyn McCulloch from Llewellyn bikes and who is Ritchie? Would they be willing to help, that would be incredible. Any help with manufacturing these lugs would be fantastic as it has been something I’ve been struggling with to move the project on. As the stainless steel printing process isn’t very precise.

Re: the angles. Say you have a 370 mm a-t-c fork and a 540 mm c-c TT frame @ 70 mm BB drop with parallel 73° angles and 425 mm CSs. Then the DT-HT and DT-ST is around 59°, ST-CS 63.5. Change the TT to 560 and the DT angles become about 60. If you want to keep it at 59/63.5 you'd have to decrease the drop dramatically to about 85 mm and lengthen the CSs to 500 mm.

You are right, sorry I had to draw that out to make sense of it. To say the least a 500mm chain stay would be pliable. What I might have to do if I want to preserve the angle is build a fork. Though I have not done that before, so you are right that filleting the frame or designing a lug with a different angle might be a better alternative.

Yeah I don't think there is a way to have the same angles for different sizes unless you keep the TT the same. You can change the stem length (they vary from 50 to 130, approximately) but I wouldn't call the result customized.

Having lugs cast (in a single angle set) is a sizable investment and the technique is better suited to somewhat large productions I think.

I think you're on the right track if you want to recreate the Caminargent design. But there are simpler solutions if you want to have a demountable bike. As for the production aspect I'm not sure I'm following since your design already requires some brazing? And I'd say laser sintering or CNC machining is more costly than a oxy-fuel setup and some files :)

Well I guess if you built a fork specifically for the frame it wouldn't be much of an issue, because you could control the bottom bracket drop, trail, that way, so you could technically maintain all the angles (within reason, before moving to a smaller wheel size) by shortening or lengthening the head tube and fork respectively. Only problem is not being able to control the overall wheelbase and certain handling characteristics.

I would love to... but I haven't actually built lugs before, and my metal working experience is limited. Do you know if there is any primer on building lugs from scratch? Because from what I have read there are two routes, filleting the pipes together after bringing them to size, and welding them (out of the question because I don't have the tools or know how.)

Casting one off pieces isn't as expensive as you might think, they print the patterns out of wax so no tooling is required for the process. So while it does cost more than a production lug, it's only about 2.5X the price, cnc is way too much and the undercuts that would be required aren't possible. I

like the Caminargent design because it can be broken so far down that it's entirely modular, meaning that if any one part of the bike is damaged it can be replaced. I know that lugged construction can be repaired because I did it to one of my builds, but honestly it took a lot more work than I hoped it would.

The current design only requires brazing on the chain stays and bottom bracket and the dropouts... That's mostly because I am not confident in clamp fittings along the drivetrain just because of the forces acting in that area. Though what I have been looking into is using a DOW Chemical and 3M structural adhesives in the epoxy or polyurethane families (both used in aerospace and automotive for similar purposes) to replace the brazed joints, mainly because soon I won't have anywhere to braze (I don't think they like you using torches on a balcony lol)

That's true you can play with the fork height and/or leave some HT under the DT. In my example above you'd need to increase the fork height and/or HT by 15 mm. There is probably a point where it will start to look weird?

Making lugs is not that hard. The tubes come in the correct ID dimensions; mitre them and weld or braze-weld them. The annoying thing is to thin and shape them (if you need to), and more critically to bore the through hole. You could do a one-piece HT by welding the sockets directly on it. However, making a BB shell is a huge pain.

I would say that the DT-HT is under higher stress than the BB. In the former your momentum is leveraged by a fork's length under hard braking or in a collision; in the latter your force & weight is leveraged by a crank's length, and the structure is triangulated. But the tube sections are different I guess.

I don't think favouring a design over ease of repairability is wise if you have to make some compromises elsewhere. Generally the rate of frame failure is pretty low.

http://bikeforums.net/attachment.php...hmentid=336013
No not having the top tube extend below the frame. I am really bad at explaining geometry in words. Here are two frames (Seat tube 62cm and 49cm,) same geometry and same trail on the fork what happens is the head tube becomes shorter and the fork becomes longer, there shouldn't be an issue with toe clip. The chain stays stay a consistent length (in this case 400mm) and the rear triangles angles change (reason for articulating the seat stays, though I have simplified it somewhat.) So for different people the same base components can be used to construct a frame. If step over height is not directly proportionate to the reach of a rider it can be adjusted by raising the top tube and because the head tube and seat tube are parallel the length remains consistent and the top tube with lugs can be easily slipped on.

I might try my hand at making some lugs before I move then, I always thought it was a more technical process than that. I can imagine getting the four sockets on the bottom bracket in correct alignment with one another would be murder. But it's worth a good old college try.

They are the two most stressed areas of a frame, it would really help if there was a definitive model for frame stress in bike frames for both dynamic and static modeling. I am not arguing that the head tube down tube area isn't under a great deal of stress.

Yeah, the frame didn't fail. It got clipped by one of those side walk snow plows... The bike's natural enemy, motorized vehicles. And maybe those ebikes I never see anyone pedaling. I like the idea of it being modular so that in the unlikely circumstance one piece breaks it can be quickly replaced. I also want the frame to pack down so that it doesn't cost a fortune to ship if I decide to take it somewhere, the more compact the better. And have it so when I build them for my friends I can just produce them quickly without issue, and ship it to them cheaply.

if you can get them 3d printed in a material that is strong enough, you can probably get by with just a couple of lower head tube lug angles. Since you have extended length sockets, I suppose the angles are all fairly important, so the bb shell angle would have to change too. To me, some sort of welded lug for those two joints makes a lot of sense. If you keep the tubes from collapsing at the clamped joints, that would improve the durability quite a bit. I'm not sure if a structural cap on the tube would be sufficient, but it would be a good start.


For some reason, the seat tube/bb joint seems to be one of the most common failure points due to fatigue. I don't know if it sees more effective cycles or what the issue is. From what I have seen there usually is an issue with the joining technique that starts the crack, but you would think that would be equally likely elsewhere.

I meant that instead of lengthening the fork for a longer frame you could extend the head tube below the down tube. You wouldn't have to customize each fork and leave a big gap between the tire and bottom of crown. Neither solution are elegant I'd say.

http://bikeforums.net/attachment.php...hmentid=336019

My suggestion would be to make a one piece HT by welding your socket/clamps on it, and also weld the TT socket on the ST (like Herse). I'd use 0.058 wall tubes without thinning them and keep a square-ish profile. For the BB I'm not sure. I tried making one... Lots of work and then I scrapped it. But it can be done; it's easier with a jig.

I capped the tube in my detachable frame. I guess it helps.

Edit. Someone has mentioned the issue of alignment by having all those clamped pieces. It's a valid concern.

The 3D printed stainless steel is a 440 alloy with a 30% bronze content (added through the sintering process.) Tensile strength of the material is 682 MPa, Modulus is 146GPa (2.3% Elongation at break), Yield is 455MPa The issue isn't its strength but the fact that it doesn't hold tolerances as well as casting. Over the areas that would be manufactured it could be off by about +-.13mm so it is taking that into consideration and adding enough material to grind away and make things fit. Additionally it has to be printed with a relatively thick wall (3mm), so how I did the initial test with it before was I ground away the extra external material on the lug to thin and smooth the walls.

A structurally cap on the end of the tube would great, also increasing the butt at the end of tube by placing another thicker diameter tube inside might be an option to prevent collapse. From what I've seen the BB ST joint fails mainly on hills when most of the weight is on the pedals, so it seems like it is just something inherent to that joint specifically.

I like the idea of a one piece head tube, the issue I can foresee is that because of the additional thickness of the lug, it would be hard to get the head tube on? The bottom bracket I'm not even going to try making... I don't have a jig or the patience to do it. I haven't quite worked out the BB yet, I might try to make one, though what I would do is get a stock BB and grind down the DT and ST sockets and have longer ones welded on (not by me) in their place and then add the clamping surface to that.

The gap actually doesn't bother me too much, from a 49-62 cm frame it looks like (depending on the tire) the gap would only increase by about 5cm, which sounds like a lot but aesthetically it doesn't look clumsy to me. It's just something that happens to bikes as they get larger. I'm a tall guy so I'm used to seeing the gap above the tire

Aligning the tubes will be done partly by the mitres and mostly by a clevis pin. I've found stainless steel threaded clevis pins (rated for a lateral load) and placed it through the lug and the tube to prevent it from shifting.

That SS alloy is pretty strong, comparable to 4130. It must have been a lot of work to thin those beasts of a wall. Yeah that thick and the steertube will barely fit if you print the HT, not to mention the HS cups. My suggestions were for welded/brazed lugs; to me it seems much simpler then printing! :)

For a welded BB you could take one of those lugless shells that have chainstay sockets. You would only have to weld the main tube clamps and I wouldn't bother opening a through hole.

IMO aligning the two holes in the lug and tube (for the pins) precisely might be tricky without good equipement?

I'd love to see the prototype with the SS lugs!