Official Luthiers Forum!

This afternoon I had a problem with a CF/Spruce joint failure. I'd laminated some brace stock with .020" CF sheet from LMI S/CF/S, using West Systems 105 and the fast hardener. After curing, I shaped the brace on the belt sander and then capped it with another .020" strip of CF, full width. When I took it out of the clamps, one end was free, so being curious about the extent of the problem, I pulled up....and it separated like a zipper, leaving the cross-hatch pattern from the CF sheet in the epoxy. Now the epoxy was fine, and the remainder cured hard in the dish. Trying to figure out what happened. I've never had to clean the CF before, or rough it up given that its surface is already rough (unlike the pultruded bars and tubes).

Thoughts?

Without question you have to prepare the surface of the CF sheet.

(edit) What you described is a perfect example of the problem described below. No question about it since you described the failure as leaving all the bonding epoxy on the Spruce. The bond was GREAT to the Spruce...not so great to the CF panel.

When epoxy cures it forces some of the non-cross linked catalyst to the surface. Pretty much all epoxies do this and it always happens to some degree with epoxy molded parts. Holding to the proper mixing ratio when molding the part will minimize this but manufacturers err to the side of too much catalyst rather than too little when setting their recommended ratios. This residue, even if only a fine amount is present, can act very much like a release agent. So your CF sheet could very well have had a little of that going on. In addition, the surface of a cured layer of epoxy is inert and smooth on a microscopic scale. Although it may have LOOKED rough, it was quite smooth at a sub-visual level.

If you want cured epoxy to bond with fresh epoxy, scuff sand the panel well enough to leave only shiny depressions with 220...and then wipe with maybe three successive clean rags using acetone or MEK. Give it a good tooth and get all that residue out of there and it'll bond like downtown.

_________________
I read Emerson on the can. A foolish consistency is the hobgoblin of little minds...true...but a consistent reading of Emerson has its uses nevertheless.

StuMusic

Over clamping and starving the joint? Most likely. The Spruce is fairly porous, its best to apply straight epoxy, let it absorb into the wood, and apply more if needed, until it stays wet looking. Then clamp the pieces together with just enough pressure to keep them in contact.

Roughing up the CF wouldn't hurt, 100 grit would be my choice. I would do it as a matter of course.

MEK is the catalyst for a polyester resin, not epoxy. I don't know why it would be suggest to be used for epoxy. The film on some cured resins is called an "amine blush". Sanding it off is fine. The dust residue will be mixed up into the applied resin and not be an issue.

_________________
Expectation is the source of all misery; comparison the thief of joy.
http://redrivercanoe.ca/

MEKP is the nasty peroxide based catalyst for polyester...and something totally different than MEK.

MEK is a light solvent you can use in lieu of acetone if it's not available.

And I fully recommend using a solvent to remove migrated catalyst. Sanding might clean it off....though it could just spread it around without actually removing all of it.

Clean is king when it comes to good epoxy bonds.

_________________
I read Emerson on the can. A foolish consistency is the hobgoblin of little minds...true...but a consistent reading of Emerson has its uses nevertheless.

StuMusic

Everything Stuart said. In addition if this is a pre-cured laminate you bonding, unless it has/had a peelply on it, there's a chance it had mold release on the surface causing the release. Never trust a shiny surface to bond to.
you can test you surface for bonding by using a water break test. The water should sheet, if it beads up, you could have a problem.

_________________
Jim Watts
http://jameswattsguitars.com

Jim, this was LMI's stuff, and it isn't smooth or shiny like the pultruded rods or bar stock. It's got a rough texture, a micro waffle to it. That impression was left in the epoxy. When I went back this evening to try again, I took one of the braces I laminated and jammed a chisel in between the CF and the spruce. I split right open. So I've cleaned every last piece of CF sheet in the shop with acetone, and I scuffed the two that I used in layups this evening. Won't let that bite me again.

The best adhesive for wood/CF lamination I've found(and I have tested pert' near everything there is) is polyurethane glue....

But that said, why do you conclude it is the best?

Long-term testing. I've been testing adhesives with CF since 1998, in fact, and still have tests on-going. There are many interesting papers of tests between PU glue and others, in use in laminated beams that were(and like mine, are still on-going) conducted in Europe, and PU beats all. Why PU glue has never caught-on with the mainstream in America is beyond me, but likely has more to do with Big Business than it does with actual adhesive performance.

In short, epoxy can, and does, cold creep, so it kills the main benefit of using CF in bracing. And as we've seen, an epoxy joint can fail. An epoxy joint, even perfectly prepared, -will- fail at high temperatures. A PU joint will not fail even at temperatures that will ignite the wood....

My only experience with PU was with the old Gorilla glue about 14 years ago for some carpentry work - it was not an impressive product. I'm sure they're much improved and I would like to try some out. If I can ask, what type are you using?

_________________
Steve Smith
"Music is what feelings sound like"

Same here - curious as well.

_________________
Thank You and Best To All

The structural epoxy you'll see used in aircraft is post cured, which is not something I would recommend with any wood product.

I wish I would have copied the paper from the long-term Euro test of adhesives, as it clearly showed PU's advantages, even over high end structural epoxies when used in loaded laminated wood beams. Fact was, over the decade-long load test, the -only- adhesive that had not shown any cold creep was PU.

I use Excel-One:

http://www.excelglue.com/excel-one.html

or Elmer's "Probond" or "Ultimate". They're the same identical product, other than the packaging; makes it impossible to get Home Depot or whgoever, to price match, as they'll argue that it's not the same product.

The most important thing is to use it when fresh; I buy a new bottle, and glue up as many sets of bracing as I can, until I run out of wood, CF, or glue. I use a LOT of clamping force, and it has to be even clamping force. The foaming action actually tries to push the laminates apart, and the foam has no strength, so solid, and even, clamping is important. I leave 'em clamped for about 24 hours, therefor, I only make one set of braces per day when I'm in glue up mode. And yes, surface prep is important, and I'll use alcohol or MEK to prepare the CF surfaces, even though they are lightly textured and appear clean. Wear leather gloves! CF splinters are nasty!

Thanks Mario. I can see where this would be quite useful for laminates. On the other hand epoxy will probably still be best for installing CF bars in the neck or applying CF tow to the top and bottom of the brace.

_________________
Steve Smith
"Music is what feelings sound like"

I just made up a few sets of CF reinforced braces with epoxy. I believe Mario has me convinced to use poly in the future. Good info Mario!!!

_________________
Joe Beaver
Maker of Sawdust

Just for the record I tested a cutoff of the epoxied braces I had made. While I didn't have the 'zipper' situation it did come apart with a little less wood clinging to the CF than I had hoped for. I used Smiths all wood epoxy

_________________
Joe Beaver
Maker of Sawdust

Kind of a side question - is this or something similar also a problem when using TOW? I imagine it wouldn't be since there's no epoxy on TOW.

_________________
http://www.birkonium.com CNC Products for Luthiers
http://banduramaker.blogspot.com

Epoxy is not a problem on Tow and you would not want to use polyurethane glue directly on tow since it would not have a low enough viscosity to wet it out.

Some years ago I read an article in the Experimental Aircraft Association's mag 'Experimenter', which is aimed at folks who build their own airplanes. It was all about 'surface energy', and the importance of having high surface energy in an epoxy joint. Basically, the strength of the bond depends critically on getting a good chemical bond with the surface. Old surfaces that have not been worked in a while will not have open bond sites and will bond poorly. Working the surface breaks chemical bonds, and allows the new epoxy to latch on strongly. Note that it's not 'tooth' that counts; it's a chemical bond, and not a mechanical one that we're looking for, so scraping is as good as sanding, if not better. This is the reason that fiberglass sub-assemblies come with peel plys: you remove the peel just before spreading the adhesive, to provide a 'freshly worked' surface. The same thing applies to wood. of course: the Forest Products Lab found back in WW II that you get a better bond if the surface has been worked within 15 minutes. An easy test for surface energy is to spritz on a mist of water droplets. If the polar water molecules are attracted to open bond sites on the surface it will spread out into a film. If it sits there in drops, the surface energy is low.

Everything Al said, except I'll note that it applies to -all- curing adhesives, not just epoxy. I say curing to rule out pressure sensitive adhesive, for example, which works differently. Also: clean clean clean, but exposing the fresh surface should always be the last step before gluing.

Scraping and sanding would be essentially equivalent if cleanup was always done properly. In practice, cleanup is rarely done properly. The dust created from sanding has high surface energy, and can cause a misleading spritz test as the water wets the dust instead of the surface, and the adhesive will do the same thing. When sanding wood pre-glue, best practice is to clean off the part with dry air at high pressure (90PSI+) with the nozzle within an inch of the work to keep the pressure concentrated.

In more demanding applications (gluing metal, for example) the surface prep can get really exotic. Best method I found for gluing aluminum was two cleans with acetone followed by a pass with a torch with a blue-flame, though I didn't try any chemical etches which might have worked even better.

_________________
Bob Garrish
Former Canonized Purveyor of Fine CNC Luthier Services

Great thread. I didn't know about mold release agents. I usually sand my CF bars for neck reinforcement to get the right fit. Guess I was doing the right thing without even knowing it.
Thanks for all the info gents.

_________________
It's not what you don't know that hurts you, it's what you do know that's wrong.

TOW covers a wide range of configurations from a few hundred strands of fiber to many tens of thousands of strands. I'm assuming you mean the tows that come with a relatively large number of strands such that they may be suitable for wet layup?

When manufacturing carbon fiber, the fiber strands are run though a chemical bath called a "sizing"...kind of like a primer. The coated fiber essentially reduces surface tension of the liquid matrix and helps it seep along the fiber surface better...promoting a thorough wet out. Sizings can also be designed to facilitate the weaving of tows into fabrics by altering how the tows slide past each other.

The thing is....sizing is specified to accommodate epoxies, polyesters, vinylesters, urethanes, ...among many other matrices. The sizing that comes with the vast majority of CF tow that you're likely to get your hands on is going to be epoxy compatible. So..yes and no...if you try to wet out a tow with urethane it won't be the viscosity preventing wet out...it'll be because the sizing isn't appropriate.

_________________
I read Emerson on the can. A foolish consistency is the hobgoblin of little minds...true...but a consistent reading of Emerson has its uses nevertheless.

StuMusic