CF tube-to-tube construction, is filler/fillets "less desireable"

[wallymann]

i've seem many CF builders that work in tube-to-tube construction assemble their freshly mitered frames with a bead of epoxy, then build up with foam filler, then wrap the smoothed joints with CF.

and i've seen a smaller # of CF builders that also work in tube-to-tube construction that do away with the foam filler before applying the final CF wraps. parlee and eyewater come to mind. apparently they've figured out methods to compress the joints and do not need filler.

in any case, i'm curious if anyone would care to weigh in about whether using the filler is really that big of a downside for tube-to-tube CF constructions.

[robt57]

Seems to me additional wraps of carbon and matrix over any epoxy 'welds', then the usual vacuum curing process after, yada.

But what I know about these processes is rudimentary at best I confess.

[wallymann]

Originally Posted by robt57

Seems to me additional wraps of carbon and matrix over any epoxy 'welds', then the usual vacuum curing process after, yada.

yep, i think both processes feature CF layering and vacuum/pressure during heat curing...my question is about the foam filler.

[robt57]

So I understand, there will be additional layers over the foam, right? So this is an 'in process' shot prior to that cloth over foam wrap/vacuum/cure process?

Seems the filler is going to serve to have the top layer [assuming] have gentler flow, thus less probability of stress risers in the finished matrix in my minds eye.

If this is the context, foam=good in my mind.

[wallymann]

Originally Posted by robt57

So I understand, there will be additional layers over the foam, right? So this is an 'in process' shot prior to that cloth over foam wrap/vacuum/cure process?

Seems the filler is going to serve to have the top layer [assuming] have gentler flow, thus less probability of stress risers in the finished matrix in my minds eye.

If this is the context, foam=good in my mind.

correct...after the filler, outer layers of CF are overlayed and the whole thing is bagged or otherwise compressed when heat-curing.

here are a couple more post-foam-shaping when the outer CF is added, then post-curing/finishing.

process: tube-mitering -> tube-to-tube epoxing -> foam filler added (optional) -> CF overwrapping -> compression+heat curing -> joint clean-up

[unterhausen]

my understanding is that is pretty common in most kinds of carbon construction. We have an idea that foam is weak, but it's not necessarily so. Lots of foams are good under distributed compression loads. They build houses on top of foam nowadays.

[bicycletricycle]

This is a pretty descent way to construct frames in my opinion. It isn’t ideal but It makes custom options really accessible.

[mcteague]

Waiting for Nick Crumpton to weigh in.

Tim

[Carl Strong]

The fillets aren't typically foam. In most cases, they are a high-strength epoxy. Some builders may use a light-weight epoxy filler, but most builders I know use the full-blown "space glue" high-strength version.

The argument for the fillets is that it allows the builder to lay the overwrap over a smooth radius that minimizes "crimps". Crimps are sharp bends in the fibers. Because the fiber's greatest strength is in pure tension, you get the best performance out of it when it is flat and straight.

[sjbraun]

My Hampsten Maglia Rosa appears to have been constructed with some sort of filler/epoxy in the joints, then smoothed (sanded?) before the final carbon wraps were applied. Whatever the process, it made a really nice riding bike.

[zmalwo]

If this is the case why not use lugs. I still believe Colnago's lugged carbon frames have the strongest joints and the smoothest inside.

[prototoast]

I think Carl covered most of the bases, with respect to better fiber performance without sharp angles.

In terms of performance, while it may seem like having some sort of "filler" material may be less desirable, in many composites applications, foam core (or honeycomb, or similar) sandwich panels are deliberately chosen over straight carbon fiber for improved strength-to-weight. It can be hard to appreciate it until you've felt it in person, but you can take two flexible pieces of carbon fiber and a flexible piece of foam, and slap them all together to get a panel you can't bend with your hands.

In practice, what this means is that having some filler in there should make for a stronger/stiffer frame than without, with minimal added weight. So I don't think it should be thought of as less desirable (does anyone consider a Crumpton undesirable?). There are lots of different ways of using carbon fiber in bicycles, each with their own strengths/weaknesses.

[wallymann]

Originally Posted by zmalwo

If this is the case why not use lugs. I still believe Colnago's lugged carbon frames have the strongest joints and the smoothest inside.

mitered tubes with external wraps allow infinitely variable custom geometry. no need to spend (big) money on tooling to produce an array of lugs for lots of custom options.

if you think about it, the CF over-wraps are essentially a custom made CF "lug" added after the tubes are mitered and bonded.

[prototoast]

Originally Posted by zmalwo

If this is the case why not use lugs. I still believe Colnago's lugged carbon frames have the strongest joints and the smoothest inside.

Off the shelf lugs do not exist for carbon fiber like they do for steel builders.
This means any builder who wants to make a molded carbon fiber bike needs to design/build/commission his/her own custom lugs. This costs thousands of dollars for just one frame size, and that limits the availability of customization. It takes someone with scale for custom lugs to make any sense, and if you look at what Colnago is putting out, they still cost as much as a fully custom Crumpton or similar.

The biggest strength of tube-to-tube is that it can accommodate any geometry without additional work (much like tig or fillet brazing vs lugged for steel). And once you go down the path of lugs, it actually makes more sense to just keep going all the way to a molded frame from both a price and performance perspective... though lugged carbon sure does look good.

[Carl Strong]

Originally Posted by prototoast

The biggest strength of tube-to-tube is that it can accommodate any geometry without additional work (much like tig or fillet brazing vs lugged for steel). And once you go down the path of lugs, it actually makes more sense to just keep going all the way to a molded frame from both a price and performance perspective... though lugged carbon sure does look good.

That about sums it up. And the final point being that the difference between using fillets and not using fillets on T2T frames is one of preference. Not unlike lugged vs fillet vs TIG metal frames.