Official Luthiers Forum!

There is no spruce or other wood in the top. I tried standard parallel bracing early on. The theory was that I wanted it to sound like a "normal" guitar. It ended up leaving the sound too tight and banjo like and didn't prevent the catastrophic collapse of the top. I've attached a picture of one of the dead top experiments. Sharp eyed readers can probably see where this one failed just past the neck end of the treble side brace.

I have thought of laminating a decorative wood veneer on the top, but don't know whether that would deaden the sound. Walnut Burl, maybe? Although I have a lot of headroom there to experiment. The top weighs about 150g. Not sure how much added weight wood veneer would add. My spruce carved tops have been in the 450g range.

There is some "bracing" in the sense that the layup is reinforced. I won't say exactly how, a girl has got to have some secrets The mold is an inversion of the carving program I developed on my home cnc machine for my wood archtop plates although I suppose it could have been done by hand carving. Had to modify the arch a little to account for the differences CF/wood. For example, a "recurve" only seemed to introduce a weak spot without any obvious benefit.

The layup is a bit laborious and time consuming, but compared to hand carving a plate much less so. CF cloth is hard to cut cleanly. The edges want to fray constantly. The material comes in long rolls, and I have a very small shop. I seem to either make way too much epoxy (relatively expensive stuff!) or too little. Squeegeeing the epoxy has a tendency to push the weave around, which is fine except for the presentation layer. I learned I need to wear a painter's suit and mask whenever I cut the stuff. It itches raw skin like fiberglass. The materials also are comparable in price with a good piece of spruce, which means it stinks to mess up a top.

In short, working CF was a steep learning curve. I did very basic layup technique which is certainly not rocket science. Still, it had its challenges. Regular CF hand layup:

1) Mold prepped with PVA and wax
2) CF sheets laid up one at a time. Epoxy squeegeed on until soaked, next layer.
3) Peel ply applied
4) Breather cloth applied
5) Vacuum bagged overnight
6) 3 day cure.


I don't have any pictures of my work-flow, but they wouldn't really be illustrative. All you would see is CF in a mold in a vacuum bag.

That looks great--thank you for posting. I've no doubt a carbon fiber archtop can sound great. The learning curve looks to have been very steep--bravo for persevering!

Is this a process--making the top--that could be automated easily? It looks like once you've got the formula down, you could reproduce it pretty easily. Have you thought about selling the tops?

I honestly hadn't even thought of selling just the soundboard as a supplier to other luthiers. With the right equipment they could be produced in quantity. Right now I only have one mold and one vacuum bag. Each top takes three rounds in the vacuum bag (top layers, bottom layers, and laminated together) so it is currently a minimum three day process. It used to take me about a week to carve a top from Spruce, so I wasn't too concerned about speeding up the process. If there was a need, I could invest in a vacuum table and could easily machine multiple molds.

Very impressive! Looks really awesome, too. A CNC router and a vacuum pump open up so many doors to making all kinds of stuff with carbon. Garbage bags make totally passable vacuum bags combined with a heat sealer.

Having spent a...few...dollars on carbon fiber experiments/disasters, I've got a couple ideas that might help people trying to do this:

The weave will flatten out if it's under pressure against an external mold, and all the little gaps will be filled with resin. Heat also helps A LOT, so even if you can just put a hot lamp or something on the part while it's being vacuum bagged the epoxy will flow a lot better and be less likely to leave voids on the mold surface. Low viscosity epoxy plus heat are huge when vacuum bagging, because your pressure is limited. The spray mold release tends to be a lot easier to use and put down in a thin layer than wax.

Carbon cuts pretty well with rotary cutters (with the round razor wheel) vs scissors, though they do sell special scissors for cutting carbon that have one serrated blade so the fabric doesn't slide and fray as much. You can stop the edges from fraying by spraying it with a tiny bit of spray adhesive, like Super 77. It's not enough to mess with your end product.

It's worth weighing the epoxy when you wet out the fabric to minimize the amount you need to squeeze out, we weigh the epoxy and then squeegee it out onto the cloth on a piece of vapor barrier over our tabletop. We do it one layer at a time and then stack them to keep things even. The cloth is much easier to handle when it's 'dryer', and it also makes for less waste with the epoxy.

_________________
Bob Garrish
Former Canonized Purveyor of Fine CNC Luthier Services

Thanks for your feedback, Bob. Couple of quick questions, if you don't mind sharing:

1) What epoxy are you using? West 105? Is there something with a lower viscosity that you are using?
2) What is your epoxy/cloth weight ratio? 1:1?
3) What kind of "vapor barrier" are you using on your worktable to wet out the carbon fiber?
4) Which spray mold release do you use? I've heard bad things about them not releasing and that is why I was using traditional wax.

Thanks!

1- We're using Restech by a company called Fabtech, but it's just what we had on hand when we started with our current system. When we restock, we'll likely use the Adtech 820 from Soller Composites. We get all our materials from Soller already, it's just a hassle to get epoxy due to the hazmat shipping. The Adtech is thinner than West. The Restech is gloopy, but at elevated temperatures it flows much better.

2- We're using 1:1, then running it through a pasta strainer, and we get a decent bit of squeeze-out on our molds. We're mostly bladder molding with carbon tube, though. I'd say the same starting ratio likely won't hurt though, it's pretty standard.

3- We're just using standard vapour barrier (big roll of plastic from the hardware store) to cover the table so we don't cover it in epoxy.

4- Loctite Frekote 770-NC for mold release. So long as they're properly applied, we've never had any brand of spray mold release fail.

_________________
Bob Garrish
Former Canonized Purveyor of Fine CNC Luthier Services