Gonna make a pimpin carbon disc wheel from scratch

I picked up these obscure disc wheel hubs on eBay. Not going to say where I work but there's carbon and autoclaves. Don't think I've got any realistic way of making a lenticular mould, so going to lay up a couple of flat skins for each and see what happens. One is 33mm between the flanges and the other 23mm. I think the 33 will make a pimpin-ass 20" disc and the other a narrow 23mm flat disc maybe with a foam core for more stiffness. UD plies on the outside because stiffer and more gangster.

Input plZ. No idea what I'm doing here.

1 Attachment

Subbed for potential amazingness/carcrash

Awesome! Shouldn't be any more sketchy than your obscure ratty russian disc.

That thing is really solid. I'll be using far higher quality composites, stiffer fibres and much stronger adhesives.

The weakest link will be the bonding to the hub, no? The benchmark for pimpin discs:

https://www.youtube.com/watch?v=7trn91xkJ0w

I'll be using 9323 to stick it together, which is one of the strongest adhesives in lap shear. I need to do some sums, but it might not be the adhesive or the bond which is the weak point. My outer skins are going to be M55J UD fibre which is notoriously poor at sticking to anything. The weak point might be the boundary between the fibres and the outer layer of resin in the pre-preg - because that's how it typically fails when you peel it apart.

older disk wheels used rivets through the hub flanges in to the carbon sides so maybe thats worth considering in addition to a bond. the hardest part as i see it to this project is true-ness of the wheel in roundness when creating the central hole for the axle. Bonding to the rim is a critical element, the mid 90s through to mid 00s mavic comete suffered from a small bond area which has been know to delaminate. I feel the properties of the glue dont need to be a strong as you think if the surface area is there, an overly hard/brittle formula may sheer sooner than something with a bit more flex given the vibrations and stresses it will be subject to but thats just my own personal view. A foam core is a good idea for a flat sided disc. Ive seen the inside of an old Ambrosio glass fiber disc is made a bit like that/ old skool surfboard way. My Cinelli Carbon Fiber again a Gipiemme creation has the best sound of any disc i have used both old or modern and is probably the most rigid too being full carbon, lenticular and hollow inside. Prepreg might be over complicating things. Have you seen the instructibles on the DIY disc wheel? vac bagging the layup against glass or a very flat surface to make each side.

There isn't much point bonding and riveting as the two methods can't work together. the bond will rigidly take all the stresses and only if it breaks will the rivets take load - so they'd both need to be designed to take all.. plus rivet placement is likely to be most useful at the critical bonding regions... the best bet with this from my experience (manufacturing composites PhD) would be to make a good bond and try to estimate the peel stresses, using rivets to account for that as more of an anti peel aid and no design then to take load at all

I've only got access to pre-preg. Besides, it's way easier than wet layup!

For the smaller wheel I've got a glass plate. For the bigger one we've got sheets of aluminium sandwich specifically for laying up flat items and I can stick some PTFE down onto it to prevent the surface pin holing.

Bonding the two halves together and getting them parallel is no problem. Truing it up on the hub should be no problem with a dti.

Riveting isn't a good idea in my mind, because you lose fibre continuity and I think 9323 is strong enough. It's a tough adhesive, specifically for sticking aerospace parts together and tabbing high strength test specimens.

Any ideas what to use as a mould for a hollow lenticular are welcome.

a mould for a hollow lenticular

I have access to a plastic vakuum machine and have thought of making some lenticular wheel covers. For that I dreamed up using plywood or mdf cut as a reverse bowl by a "cnc" woodshop. I have not got a quote for that yet, so have no idea of cost....

Couldnt you do something similar?

Yeah I've thought about some of those ideas over the years. Also getting a bowl spun in aluminium.

Here's a thing, if you partially vac a piece of thin plastic film into a round hole, it'll pull the surface into a spherical surface. It's how they make some mirrors for telescopes. For your covers you could do that, pour plaster of paris in and let it dry - then you've got a former. Obvs make it bigger than the actual wheel because then you'll be able to trim it back.

Having said that, I wonder if I could do something similar here with a well-sealed MDF box with the hole cut in the top and probably a vac bag for the film. The key thing would be how to get a good enough surface to laminate onto. And I'd have to take a female mould off it. Hmmm...

I tried doping fabric with resin, aircraft covering film, a foam filled core shave into shape but none of them were really that great to be honest. A local firm the deals in plastics explained the costs of a one off mould or vacuum forming which put me off. Ive considered clay, all sorts really. The most plausible one seemed to be, as mentioned above, to have one turned in wood.

Yeah it's all a bit too expensive for a one-off. I could take a mould off my HED disc and replicate it in proper composites but it's a bit bashed up!

I used to build competitive model airplanes a couple years ago. We mostly had positives milled from MDF (cheap), prepped those and then took the negative from them. Once worked with a mould directly milled out of some sort of PU (was probably Ureol or sth, not sure), was pretty nice to work with as well. Obviously, with MDF and lots of prepping, our moulds weren't 100% exact, but we're talking way sub-mm here. Could see if I can find some pictures (or old moulds) if interested.

Rim purchased. Got a 30mm wide Alex single wall alloy BMX rim in black.

Rough plan is to lay up a pair of skins with the rim on top of the release film inside the vac bag. It should leave a slight indent in the surface which will make centring the two skins a bit easier.

Looking at the strengths and stiffnesses. The M55J UD strength will be about 1700MPa in tension and about 700MPa in compression. That's a bit more than I remembered and much stronger than the woven I had in mind, so I will probably sack off the woven plies. Stiffness is about 290GPa in tension and 260GPa in compression, or about 25-30% stiffer than steel. I think I'm going to lay it up so the fibres run tangentially from the hub rather than radially and alternate the direction with each ply. This gets around needing the woven layer to bind the UD together in the 90 direction. It should be pretty stiff and way strong enough.

This is a gert heavy rim at 520g. I thought by going cheapo single wall with a million spoke holes it would be a bit lighter, but it seems to be made of lead or something.

2 Attachments

Bit of a slow burner.

This is not a skin. Going to use it as my ply cutting template once I've cut it to shape. It'll be the same shape as the plies on wheels like the Fluidisk, so sort of two opposite pie wedges with a wide bit in the middle and fibres running radially for maximum strength and stiffness. Technical shiz.

1 Attachment

It's what I do.

Proceed.

What about some form of honeycomb foam flat disc? Sugino CHCD style

id have thought..

flat sided- create an inner structure for strength, although they can be hollow it would help reduce chances of the sides delaminating under load/flex of the wheel in general.
lenticular- the tension of the sides can be sufficient if done well. - i rode on a wheel made out of a rim with a surface made from a polyester bed sheet doped in resin and it didnt break however front disc wheels are death at the best of times and it didnt have the best handling characteristics given uneven weight distribution.

I picked up these obscure disc wheel hubs on eBay.

Running two rear hubs on one bike is going to be RAD however you build the rims
2 x drive trains = 2 x faster

@BigH @neil-bah for this little 20" disc I don't think I'll need a core. It's so small and wide I think it'll be plenty stiff with M55J or HM63 fibre and the rim bonded in. It won't delam with HM63 but it might be a risk with M55. I'll definitely need a core with a flat 700C disc, but foam will be stiffer and easier to work with than honeycomb core. I have both, so see what pans out, but I'd still love to do a fat hollow lenticular if I can come up with a suitable mould.

There's not enough technical chat and/or jargon in this thread.