How a carbon wheel is made + video

One of the things about carbon fiber manufacturing is that there really isn't much margin for error. And while robots can do some things fairly well, the type of work involved in layup is not all that well suited to a machine. There's a lot of 'finessing' things into place and making sure things line up with other things that is very dynamic.

If you look at a typical "made in China" garment, you almost always find things like excessive thread used to finish edges, skips and loops in things etc... they make a lot, so they can throw away a lot.

A lot of that is because of a reliance on machinery.

Now imagine that on the inside of a carbon rim. You can't see it to check it. And remember, it's not just like one garment, it's like several garments in layers and stitched together. And they all have to be correct. Which is why Ultrasound or MRI is often used to check things like that.

Take that illustration the other way and now you've got something like a tailor made garment. Button holes are tidied up, stitching is careful and even. The number of garments made is fewer, but the percentage of product that is no good because of inattention and general carelessness is much, much lower.

It's probably worth noting that simply the act of filming this video ruined one rim at the forming process because it disrupted the timing from the cold room to the primary forming stage. Something slipped and the bead lip was around 1mm too shallow for a section around 5cm in length. Oops. That's an expensive frisbee....

Good question.

That machine can test wheels with several different modes. The video shows vertical with bumps and angled with bumps and with strong downward pressure switching rapidly from left to right.

The bump height is shown briefly. The wheel can be run on the bumps on the angle or vertical. Another factor is time.

The test factors on the wheel are not the same for every type of test. For example, one normal type of test that is done during the prototyping stage is 1000km per run, checked thoroughly on the digital tension checker (shown at the end with the square screen monitor) which makes a spreadsheet before the wheel is re-trued and put back on the machine. Each subsequent pass measures how much each spoke goes out of true and if by the 10th pass, there is a good record of how badly the spokes are fatigued. That test uses 50% more weight than a standard test and has a pre-programmed route that simulates a variety of terrain and usually repeats around every 30 minutes.

Wheels that take that level of testing are not sold to anyone since this is designed to be destructive to the wheel. And it is part of the prototyping process, so the wheel wouldn't be saleable anyhow.

For wheels that are for production and sale, there are usually 6 passes on the side press (which actually pressed 4 times before being re-checked, but shortened for this teaser... see the stage with the guy with the dishing tool). By then, the spokes are usually well seated.

Every wheel is also checked and has a spreadsheet built.

Wheels that do not pass are sent back for 6 more passes on the fine tuning and side press.

Then, they are greenlit by QC. From there, the rule is 4 wheels randomly chosen from 10. The testing is not destructive to the wheel and the wheel is checked on the spreadsheet again, then re-trued and regreased before sending it out.

I am not sure how many km the test is.

At that point, few wheels show any significant issue - and this is mirrored by the general response from riders that the wheels usually remain true.

I have personally had wheels built by professionals who have worked for Olympic teams that have required a re-true after a couple of weeks and were otherwise amazing wheels.

Only a wheel with thorough de-stressing and spoke seating can do this without a re-true after a couple of days/weeks though.

Heh, yep. Good eye on that.

The video includes footage from the layup and forming of 4 different rims. One of which was damaged because of delaying a crucial step and is now in our showroom as a 'bounce tone demo' model... Carbon rims have a tone that rings out almost like a bell. We'll try to include an audio byte of it some day. Carbon that has compacted well will sound like a bright note and bounces very well. Carbon that hasn't tends to thunk or sound a bit dead. Kind of like an instrument.

Also shown in the video are some alloy Rosa wheels (spoke lacing) and a 50mm carbon tubular for a prototype Road Disc type wheel that was on the table that day. It was just too difficult to get consistent light into the clincher bed, so I opted for the tubular. The filming for the wheelbuilding segment shows 4 different Bracciano Caliente Carbon Clincher wheels. I did my best to stick to just 4 of the batch through the whole process, but I couldn't tear down a built wheel just to get a shot of the rather shy lady putting the spokes in the hub for example. Some things required a bit of time to let people relax with a camera with a big honking lens pointing at them... These are not actors. Meanwhile, it's actually a factory and things are still moving on apace behind me.

Please note that this is just a teaser for a bigger video with more detail. I hope to finish within 2 weeks.

You'll also see a sequence in that vid showing an MTB wheel getting loctite'd even though the carbon clincher does not get loctite. I was going to put it in this one, but it needs a bit of explanation.

You probably would have noticed it too. I hope you liked it though.

Nothing compared to a car though, I would expect the carbon-footprint of a bike to be less than 1% of that of a car - just in the building and transportation of them. A car weighs 150x more, costs 10-50x more and has 100x more parts (30,000 in a car!!!).

Don't know about the cost 10-50x more. A top end bike is pretty much the same as a bottom end car

I think I must have missed something. Why does this indicate that the carbon footprint of a Chinese or Taiwanese built bike is significantly different from anything else that you might buy?

The factories in Taiwan are generally fairly close together as there are two main districts for manufacturing. And they are fairly close to a major port. Well compared to distances traveled in delivery...

Parts are trucked sensibly as well, maximizing space for every stage. It only starts to become more carbon-footprinty when there's a lot of wasted space in packaging.

Importing anything by sea is going to have carbon footprint, but remember that there is always a process of materials too. For example, if you are using German steel, or Swedish Steel (ie Pro-Lite Spokes), or Japanese Carbon Fiber (ie the carbon used in the vid for the layups), or Aluminum Alloy.... it is all moving around a lot.

The only way around that is to use product generated locally. IE flax fiber (where does the resin come from?) and maybe locally sourced aluminum...

But then, you are going to end up using less efficient manufacturing methods and vastly increasing the cost of labor.

Not to mention creating jobs of a type that are not well liked by much of the population. How many kids in the UK go into 'assembly line work'? Probably less than go into "sitting at home on welfare"... I'm from Canada and I don't know anybody in my home town working in a 'make stuff' type of job.