Any of the carbon fiber / wheel experts want to comment on filament wound rims? It seems like the process is the next generation, at least by the claims (both stronger and lighter).
From what I can tell there are a few manufacturers starting to go this route:
Profile Design
FSE
Schmolke
I’d like to know if there are downsides?
Filament winding should be a more cost effective approach because there are less manufacturing steps and likely less scrap.
Picture pulling a carbon fiber through a resin bath and then wrapping it around some tooling and then curing to make a part. All activities performed by the part maker.
Using Prepregs involves spreading fibers casting or applying a resin film, b-staging winding up, sheeting the Prepregs, laying them up then applying to some tooling to cure. Typically involves multiple parties and many off-line processes. You do this because thinner layers of resin impregnated filaments / fibers produce a higher quality composite. This higher quality composite may or may not be required for a given application. Laying up Prepregs in different fiber orientations can create a quasi isotopic material (filament winding is limited somewhat in the orientations you can achieve during the fiber lay down).
If filament winding is sufficient then great. My experience is that these parts are thicker than their Prepregs counterparts to offset the greater risk that a damaged fiber tow could result in a failure (load is not spread out as far). So I cannot reconcile how this makes for a lighter wheel.
(filament winding is limited somewhat in the orientations you can achieve during the fiber lay down).
FSE wheels claims the ability to vary fiber orientation by 45 degrees in a continuous run.
If filament winding is sufficient then great. My experience is that these parts are thicker than their Prepregs counterparts to offset the greater risk that a damaged fiber tow could result in a failure (load is not spread out as far). So I cannot reconcile how this makes for a lighter wheel.
The claim is that you can use less resin, essentially. Which is the heavy material in the process.
I’m not an expert, I’m just relaying marketing-schtick which may or may not be valid.
It does seem to be true that some of the filament-wound rims are about 20% lighter than Zipp or Enve rims (best-in-class). However I don’t know if that 20% came at costs in durability or safety, e.g. if it’s an apples-to-apples comparison. (And that’s just rim weight, not wheel weight - a much smaller difference since rims are already really light.)
Typically 60% fiber volume maximizes mechanicals like tensile strength.
Carbon fiber density is 1.79 g/cm3
Epoxy density is 1.20 - 1.30 g/cm3
Composite density typically around 1.55 at the 60% fibe Volume.
So, fiber is the lighter of the two for applications where you want to maximize mechanicals like tensile strength. Not sure exactly what is required for a wheel but I imagine a fair amount of toughnere is added to the epoxy.
I can see getting +/-45 with the winding for sure.
I would love to see the process at work to better understand the layup and perhaps benifit over Prepregs.
The only filament wound rims in the market right now are those made by us, Venn http://www.venn-cycling.com and Lightweight. We launched our filament wound rims and technology in 2015.
FSE’s suppler does not use filament winding, but tape wrapping using a process not similar, or analogue to filament winding. I am not sure why FSE claims that they use our process. We are not supplying them.
Schmolke rims are woven, just like the Munich Composites rims which were first to market with a machine made rim. They are made by braiding a hollow shape which is then infused with resin in a process called RTM. This is what Time bikes use for their tubes.
The limitations of wet filament winding are resin content which for us is a stable 40% and longer development time for high Tg resins as we cannot use them off the shelf due to having to formulate the resin to have a certain viscosity.
This issue is resolved by using a towpreg process where the tow already contains a predefined amount of resin, but this negates several important benefits of wet filament winding which are: low input costs and no supply chain constraints.
Here is a video thanks to GCN showing what a filament wound rim looks like.
V.
…Also which Profile Design rims are filament wound? I cannot find them…
They mention a “semi-automated carbon lay-up” which is becoming increasingly common, but this does not mean filament winding. One of the more common automated methods employed in China is tape wrapping of the tire bed/braking surface element, and automated wrapping of the rim body in a radial fashion around a solid mandrel - ie, not winding around a mandrel, but applying the laid up carbon fiber to the mandrel.
V.
My experience is that these parts are thicker than their Prepregs counterparts to offset the greater risk that a damaged fiber tow could result in a failure (load is not spread out as far). So I cannot reconcile how this makes for a lighter wheel.
In our experience with road and MTB rims, damage to the tow does not happen, but delamination is very obvious - you can observe individual tows being mobile. With our UD rims delamination is hidden but can be felt (rim becomes softer) or is observed after the rim is cut up.
So from our experience filament wound rim is no more susceptible to damage to UD rims. Woven rims may retain the appearance of function better, but impact resistance rests primarily with the resin as it is off axis to the fibers, so even woven structures fail, just less visibly so.
With regard to load spreading, as you implied, it is the entire laminate structure that has to carry the load, and in this way filament wound structures are largely no different to UD or other structures made by using non-crimp fabric as they are also not supported by other fibers off-axis. Only the resin (interlaminar tensile strength actually) holds them there.
…for the curious, why then use filament winding, UD or non-crimp fabrics, well because you obtain the maximal stiffness, or strength possible of the part for the fibers that you are using. Any crimping (for example woven structures, fabric, braiding) reduces the stiffness and strength (due to shear stress on the matrix) of the part, all other things being equal.
V.
Thank you for the explanation.
Delamination is absolutely an issue in poorly consolidated parts and without c-scanning or thermography on 100% of the parts it is difficult to rule out the possibility of voids, which could lead to delaminations.
It sounds like I wrongly assumed most wheels were UD Prepregs. Crimp is a property killer. I had assumed fabrics were only used on a surface ply for cosmetic purposes with the bulk layers all being UD.
At 40% RC by weight, I imagine you are in the 55% fiber volume range which is still more than acceptable for optimizing composite properties.
Thank you for the innovation!
Thanks V. Awesome explanation.
Thanks for the explanation! Filament winding of tube shapes seems very straightforward, but wheels seem much more complicated. Would you mind answering a few questions?
How do you wind onto a wheel? I would think that the wheel would have to pass through the fiber every revolution. I assume the wheels are not made in two halves and joined together after.
How do you get highly variable wall thicknesses, such as the inner diameter and tire side of the rim vs. the sides?
How do you get the fibers to conform to convex shapes along the mandrel without a secondary molding process?
Sorry - that’s a lot of questions, but I’m interested in the process.
Thank you for the explanation!
In your prior post, I was a little confused about what sort of resin you use. Does it have a high glass-transition temperature (e.g. >200C) or no? In theory, using filament winding you can have a lower resin content and therefore use more carbon for a given weight (therefore a stronger rim) correct?
What sort of testing do you guys do from an aerodynamic perspective? In the scope of modern aerodynamic rim profiles, are you at all limited in what sorts of shapes you can manufacture?
Thanks again for chiming in. People here on ST love details about the technical nitty-gritty.
I would think filament winding may be a more repeatable process, but it’s not without it’s challenges. I’m more familiar with composite over wrapped pressure vessels which are also filament wound. Too much or too little tension can cause the tow to shift, too much resin can cause things to move around during cure, etc. And you’d still have the stress rupture issues, though for rims I don’t see it as challenging as tanks.
It sounds like I wrongly assumed most wheels were UD Prepregs. Crimp is a property killer. I had assumed fabrics were only used on a surface ply for cosmetic purposes with the bulk layers all being UD.
No, you were not wrong. Bulk of any fiber used in hand made rims (and frames) is UD fabric. woven fabric is used only in sections where you need to conform to tight geometry changes, or want to achieve “easy” isotropy. The other scenario where woven fabric may appear is as the contact layer with the bladder as the first layer, or with the mold as the last layer. Woven fabrics’ shortcomings (strength/stiffness) are a positive during pressure molding as woven fabric will move and stretch a little, and thus reduce the potential of wrinkles forming. They also let solvents degass better during baking so you get less bubbles or surface defects. Currently popular woven fabric is twill fabric which I think is what leads people to assume the rims are machine made as this type of fabric looks very different to the usual 90/0 weave. Twill fabric also has an advantage that you can warp it radially without having to cut it.
This is an example of a generic Chinese rim finished in twill fabric: 
Thanks for the explanation! Filament winding of tube shapes seems very straightforward, but wheels seem much more complicated. Would you mind answering a few questions?
How do you wind onto a wheel? I would think that the wheel would have to pass through the fiber every revolution. I assume the wheels are not made in two halves and joined together after.
How do you get highly variable wall thicknesses, such as the inner diameter and tire side of the rim vs. the sides?
How do you get the fibers to conform to convex shapes along the mandrel without a secondary molding process?
Sorry - that’s a lot of questions, but I’m interested in the process.
Good questions, and unfortunately I cannot tell you the details as already even what details we release end up adorning other companies’ websites and product lines.
But, I can tell you the following, winding toruses is not theoretically that complicated. This is how electronic toroidal inductors are made: https://www.youtube.com/watch?v=82PpCzM2CUg
Some factories in China use a similar process to wind rims, but this results in fiber alignment geometry problems as unlike in that example of a square profile torroidal transformer core in the video, they are winding a curved rim. Thus the carbon tape that they are winding also ends up curved and fanned - you get higher density of fibers at the inner section of the rim, versus the outer section of the rim and this problem gets worse for deeper rims. The curvature of the fibers is visible thus the structure (rim) that is made using a simple torus winding process has very poor mechanical properties. Only a straight fiber is a strong and stiff fiber. Curved fibers are a defect.
This is why we made a big deal of being able to make a 50.7mm rim and still retain our fiber alignments and geometry.
There is a secondary molding process to set the final shape and get a good surface finish, but this step cannot correct incorrect fiber placement, it only sets the final shape of the rim.
Here is a presentation I gave at Cyclitech 2015 in Brussels. It shows you the rim cross section: https://my.visme.co/…-presentation-515ad6
The rim body is filament wound, tire bed is separately wrapped (soon filament wound too) and the spoke hole reinforcement section is manually laid up and then the whole thing is co cured. We are working on the process continually towards achieving full automation, while not resorting to a crimp based process.
In your prior post, I was a little confused about what sort of resin you use. Does it have a high glass-transition temperature (e.g. >200C) or no? In theory, using filament winding you can have a lower resin content and therefore use more carbon for a given weight (therefore a stronger rim) correct?
What sort of testing do you guys do from an aerodynamic perspective? In the scope of modern aerodynamic rim profiles, are you at all limited in what sorts of shapes you can manufacture?
Thanks again for chiming in. People here on ST love details about the technical nitty-gritty.
Our current resin fails in testing at 240 C/464 F so this is a high Tg resin.No, wet filament winding’s main shortcoming is higher resin content. We need to find a balance between trying to minimise the amount of resin, while achieving good tackiness so that tows stick to each other during winding. Too little resin and we end up with slip which is not good as the fibers end up being curved at best. We have a stable process at 40% resin content.We had our 35mm rim in a wind tunnel and it did very well versus 50mm rims in that test. You can see the results here: http://www.venn-cycling.com/index.php/rims/filament-wound-rims. I developed the Venn 507 rims using our advanced CFD capability (LBM with LES - acronym time 
) so they work very well with 25-26mm wide tires. I will put them into the Flanders Bike Valley wind tunnel later this year as our office is right there.The process limitation is in depth. Each new depth increase means a lot of head scratching about how to solve the winding pattern so that all fibers remain straight. I think we have a solution for the next depth beyond 50.7mm so that may come later this year or next year.
I would think filament winding may be a more repeatable process, but it’s not without it’s challenges. I’m more familiar with composite over wrapped pressure vessels which are also filament wound. Too much or too little tension can cause the tow to shift, too much resin can cause things to move around during cure, etc. And you’d still have the stress rupture issues, though for rims I don’t see it as challenging as tanks.
Yes, it is a challenging process, but this is why I like it, a lot:
filament winding is theoretically the best way to use continuous fiber reinforcement as the fibers are laid up in tension, so each fiber is facing the same direction, and are never crimped. Even UD fabric cannot achieve this as at micro scale fibers are not perfectly aligned. This is especially true for wet filament winding as fibers are free to orient under tension due to the resin being in liquid statematerial costs are minimal. We purchase raw carbon fiber tow - the only processing was to add sizing (chemical treatment to keep the tows together while not sticking to each other), and resin in bulk.we have no material shelf life issues which can affect prepreg in the marketwe have no material shortages as we do not rely on availability of fabric or prepregthere is far less waste since we are not cutting round shapes out of square aspect cloththe process can be automated thus removing the largest source of error in production of continuous fiber reinforced fabrics - human senses and hands. Humans are not good at repetitive work and even a small error in fiber alignment or laminate overlap leads to changes in the mechanical performance or outright failure.
Cool, thanks for that. What tire did you guys use for testing?
…I am sorry I do not remember. It was done by a third party in a group test. We just supplied the wheelset and were not asked for any more input. From memory it was a 23mm tubular, but unsure what brand/model.
http://blog.velocite-bikes.com/2015/10/venn-rev-35-wheels-aero-data/
Apologies for the slight post necromancy, but we have updated our filament winding process so I am adding to this thread.
Some information about the updated VAR filament winding process for manufacturing of carbon rims is here: https://www.venn-cycling.com/en/venn-technology/#var
Victor
Apologies for the slight post necromancy, but we have updated our filament winding process so I am adding to this thread.
Some information about the updated VAR filament winding process for manufacturing of carbon rims is here: https://www.venn-cycling.com/en/venn-technology/#var
Victor
Thanks for the update.
Looks like filament winding is catching on. Are you familiar Scorpion, Winspace, Hunt wheels? It appears that Scorpion is the supplier to Winspace and Hunt brands, while also selling their own brand. What is interesting is that only Winspace sells 65mm deep rims while the other 2 do not. In your opinion, do you think 65mm deep rims using filament winding is still an issue today?