Fiber Q & A
Questions and answers about the amateur use of carbon fiber epoxy composites for bicycle frame construction, taken largely from my e-mail correspondence. (Thanks to all who write!) --
Tip: In most operating systems, pressing Ctrl+F will bring up the Find (On this page) dialogue box.
Newest Questions | Beginner | Materials | Number of Plies | Laminating | Bonding Metal | Painting | Warnings
Q: I was reading http://www.chainreactionbicycles.com/carblife.htm and someone has a post that says he/she fixed damaged carbon fiber with a "mixture of fiberglass shards and epoxy."
A: Compared to continuous carbon fiber, fiberglass shards and epoxy do little to add any significant strength. The fibers are not long enough, and any bond to the remaining good material is an insufficient butt joint, not the preferred lap joint. But it does fill a hole in the chainstay nicely. That helps prepare the bike for a structural repair by bringing the damaged area flush with the surrounding good structure (see below).
Q: The frame tipped over and the top tube caught the impact right on the corner of the concrete step. It didn't shatter, just chipped the hell outta the top tube. I've ridden it since then on some semi rough trails, nothing too hard, and it hasn't gotten any worse but the fear of it shattering is ever present. I was wondering if I could add some strength to the tube, and how I would go about doing that. I figure I should just wrap fiberglass material around the crack/chip (more of a big chip less of a crack). Specifically what do you think I should use, and how would I go about doing that?
A: Of course, without seeing your bike it it's hard to say what kind of repair it needs (if any), but in general repairs are accomplished as you say: wrap fiberglass around the damaged area. Here's one way to do it:
- Remove loose fibers and carbon shards (if any). Remove the surrounding paint by sanding with 80 grit to prepare the surface for bonding the patch. Wet sanding is optional but recommended to keep the dust down.
- Fill the damaged spot with Bondo or equivalent and level it with the surrounding good area before putting on the fiberglass patch. Filling the chip allows the new fiberglass to lie straight, without dipping into the chip. Fibers are strongest when they have no kinks.
- Consider using three or four plies of approximately 5 oz. plain weave glass; more if damage is severe. Add a sacrificial ply if you plan to sand and paint it. The plies should be successively smaller by about half to 1 inch (measured on the radius) to taper the thickness of the patch.
- Use epoxy resin. Follow the mixing instructions exactly!
- Laminate by putting the largest ply on the frame first, then the smaller ones concentrically like a bull's eye target. The smallest should be at least one inch larger than the damaged area (including any Bondo on the surface). Follow standard prep, cleanliness and wet out procedures for a satisfactory bond.
- While repair is still wet, wrap tape (best is heat-shrink tape, but vinyl electrician's tape also works) over the repair to compact it. Start the tape on a good portion of the frame away from the repair. If using electrician's tape, put the sticky side up (away from the repair). Spiral the tape carefully to overlap fully; you want an even number of tape layers everywhere: neither gaps of one layer nor extra overlaps of three. This avoids forming large ridges in the cured repair. Heat the heat-shrink tape after wrapping, or stretch the electrician's tape as you go.
- Make pin holes in the tape roughly a quarter inch or so apart to let excess wet resin escape.
- Let cure fully.
- Remove tape. Sand and finish if desired
Q: How did you jig up the chainstay area?
A: I made the main front part of the frame and the right stays first, keeping them aligned relative to the table. Then I made the left stay. I used a perfectly dished wheel and braced against the right side stays while it was still on the table.
Q: I assume that the foam core remains in the frame after construction. Does this increase the weight of the frame substantially?
A: Correct: the foam core is captured inside the frame as it is built. The foam does add weight, but I don't know how much. I didn't weigh it before I began adding carbon, but I'd guess that the core might be less than a pound or so. The foam's density is about 2 pounds per cubic foot. I used Styrofoam for part of the frame, and melted it after cure in order to route the internal cables. Some came out, but the frame still got heavier by 10 grams! Maybe some of the acetone was still in the frame when I weighed it.
Q: I've read almost every Web site on carbon lay-up but have no hands-on experience. I desperately wish to build my own frame but do not weld; this seems to be an alternative.
A: That's one of the advantages to composite construction: it is much, much easier to learn than welding or brazing.
Q: Would you recommend this kind of project to someone of partial experience?
A: No special theoretical knowledge is needed. If you have an understanding of how bike parts need to fit together, then making the shape of the frame should be okay. If you have an understanding of how composites go together or are willing to practice a little, then the layup should go okay. It is still a lot of work, but can also be a lot of fun if you have the patience to finish it.
Q: Some of the things you did for the first bike were way too involved for me. For instance, machining your cable stays, machining your bottom bracket, and the mold of the front derailleur in fiberglass. Could these be compensated by purchasing them?
A: Metal parts are easier to purchase nowadays. Cable stops or guides and whole bottom bracket shells are available for aluminum frames from the same suppliers that sell to frame builders. These should work well for a home-made carbon frame. Try Nova Cycle Supply in Rocklin, CA. They carry lots of frame tubes, bottom bracket shells and dropouts. They only sell to bike shops or regular frame builders. Also, Quality Bicycle Products is a wholesale component distributor that carries frame building parts as well. Your local bike shop can show you what they can order from QBP.
I only made my cable stops on a lathe because I wanted them to fit into the bullet shapes I molded into the carbon for internal routing. You could use off-the-shelf parts instead.
My front derailleur mold was actually pretty easy to make, but you could avoid that by designing your frame to have a more traditional shape in that area. Then you could use a bolt-on derailleur mount from a Kestrel or other bike. Or incorporate a metal tube in that spot and use a clamp.
Q: Keep in mind I've never dealt with fiberglass or carbon fiber. I am very interested.
A: Go for it! Don't get paralyzed by thinking it has to be perfect. Compared to making metal frames, there is very little technique or practice required for composites. They are very forgiving of mistakes: you can always sand it off and do it over!
Q: I know you can get laminating and surface resins when using polyester resins, but do you get the same with epoxy resins?
A: No. Epoxies are sold in various formulations for different uses (bonding or laminating, etc.), but the distinction between laminating resin and top-coat resin as with polyesters does not apply. As I understand it, polyester resin cannot fully cure in the presence of air (oxygen?). Therefore, in order to get the final coat of polyester resin to cure, wax is added. That makes it a topcoat resin. During cure, the wax rises to the outer surface of the resin and acts as a barrier to the air. Then the polyester resin can fully cure. Epoxy can fully cure in the presence of air, so does not need any additives to distinguish laminating and top-coat resin formulations.
Q: Does it matter so much when using epoxies, so long as you prepare the cured surface correctly?
A: The surface must be carefully prepared for any kind of bond to perform as expected. When laminating with epoxy, I prepare a cured composite surface as follows:
- Mechanically clean: remove stuff like paint, old bits of dried substances, stickers and their adhesives, decals, flapping plies, scale, crud, etc.
- Chemically clean: remove grease and oil: fingerprints, grease from old bike parts, oil from the bike chain, etc.
- One last wash: rinse and scrub with acetone until white paper towels come up clean.
- If the substrate is metal, I sandblast it when possible or use 80 grit emery paper as a second choice. The idea is to develop a rough surface for the bond to key into. At the last second before it gets bonded, I rinse it with acetone.
- For aluminum, I used West System's two part etch kit. This nearly doubles the strength of the bond according to West. The idea is to convert the aluminum oxide that forms on raw aluminum. Aluminum oxide only takes fractions of a second after it is sand blasted to form, so there is no such thing as getting rid of the oxide by sanding it just before bonding. [Note, West System no longer recommends the acid etch, as its G/flex epoxy bonds to oxidized aluminum. See test results here and West System's Use Guides for recommendations on special preparation for various materials. -- John Allen]
Q: You mention using West dispensing pumps but did you also use West Epoxy?
A: I used West System epoxy for all the laminating. I used their 105 resin and either 205 or 206 hardener, depending on how fast I wanted it to kick off. See http://www.westsystem.com/ for more info on West System.
Q: Unidirectional or bidirectional weave?
A: Either. I used uni everywhere but the final visible ply; I used bidirectional for that.
Q: Ever heard of "foam-core"? It is like posterboard over a foam layer in the center (comparably to cardboard in setup but solid). It is pretty dense. I'm considering laminating sheets of 3/16" thick, up to a thickness of about 2". Do you think this will make an acceptable core?
A: The poster board may make it difficult to shape the frame easily. I always order my foam in blocks from Aircraft Spruce. I use the 2 inch thick polyurethane foam.
Q: I've noticed Aircraft Spruce has unidirectional roll in 3" and 12" widths. Could a 3" roll be used to wrap (like a bandage) in opposite directions (each ply) keeping every 3rd ply in line with the frame (orientation = 0)?
A: Yes. Consider using 0, +60, -60 degrees orientation. I think the 12" width will work better for that, depending on your tube diameter. I hand wrapped and rough trimmed paper patterns around the area to be laminated to guesstimate how best to cut the carbon.
Q: Have you ever applied aramid cloth [Kevlar®] to your "toys" during construction? I picked up some this past week and have been thinking about it.
A: Yes, I used a few plies as "insurance" in a one piece aero bar I made for my wife. In general though, Kevlar® or aramid is not very useful for bikes. Kestrel uses none, Trek uses none, and other composite bikes as far as I know use none (except as a marketing ploy).
Bicycles are generally stiffness-critical structures, i.e. by the time you've added enough material to make it stiff enough it is nearly always strong enough. Kevlar® is not as stiff as carbon, and consequently you are typically better off replacing each ply of Kevlar® with a ply of carbon. Kevlar® is not stiff enough to make a good lightweight frame.
As far as strength, any Kevlar® hybridized with carbon takes up very little of the strain, since the carbon is stiffer. And if you do crash hard enough to break the carbon, then you have bigger worries than whether a little Kevlar® can "save" the frame ;-). If present, the Kevlar® just holds the two pieces of carbon scrap together like a limp rope.
Kevlar® is also impossible to sand.
But Kevlar® is the material of choice for HPV fairings: they are not typically structural, and when you crash a faired HPV you can slide on a Kevlar® fairing and it won't shatter or abrade away like carbon or fiberglass will. It just gets very hot!
Q: How can I determine the amount of cloth I need, where I need it, and the best orientation for my frame design?
A: Good questions, and ones that I struggled with. I hope some day to be able to answer them by analysis.
However, to make my frame I didn't do any calculations. I used unidirectional carbon with a quasi-isotropic layup [0, +60, -60 degrees relative to the length of the frame member] and guessed at the locations of highest stress. Failures in metal frames hint at where those high-stress locations are:
- head tube/down tube area
- all tubes joined to the bottom bracket: chainstays, downtube, seat tube.
Don't neglect the middles of tubes. Although there is less stress there (butted metal tubes have thinner walls in the middle), there is still some: handlebars hit top tubes in accidents, etc. I suggest a minimum of 7 plies (two sets of [0, +60, -60] and one ply of cloth on top.
I used between 7 and 27 plies for my beam frame. A truss frame (more traditional diamond frame style) could probably do well with between 7 and 13 plies, plus a couple extra under the head tube/down tube and around the bb area.
All this assumes reasonably oversized frame tubes, like Kleins or most other aluminum frames have these days. Smaller tubes (like 1 1/4 inch OD) should use many more plies. The Specialized Allez Epic carbon downtube I cut open had 17 plies!
Q: I don't remember seeing in either of your articles on building the frame how many plies of cloth your recommend. Does it differ on various areas of the bike?
A: The formulae for calculating the number thickness and orientation of carbon plies are over my head right now. Consequently, on my carbon bike, the number of plies varies according to the stress I think the area will see, with higher stressed areas getting more plies. Here's a rundown on how many are where:
These ply counts work with the layup techniques and frame tube diameters I've used. One thing to remember is that the size of the cross section has a greater effect on strength than wall thickness. Making the diameter larger increases strength exponentially, while increasing the wall thickness only increases the strength linearly.
Q: Have you ever considered vacuum bagging for increased strength?
A: After I finished this frame I bought a vacuum pump, but have not used it much yet. I plan to vacuum bag the next frame in a mold (you may have already seen carbonTTframe.htm).
Q: Is the process you're attempting now [vacuum bagging in a female mold] a better method for producing a carbon fiber frame?
A: I believe molding two halves with a vacuum pump to provide pressure will produce a better structure and allow me to better control wall thickness. One problem with wrapping over a foam core is that it leaves ridges where the vinyl compression tape overlaps, and sanding this partially smooth for the next ply removes an uncontrolled amount of carbon. I compensated by increasing the number of plies. This problem will not exist for a molded frame, since the visible exterior surface will be against the mold face and the vacuum bag will press it hard against the mold. On the other hand, one drawback to making a molded frame is that you first have to make the mold!
Q: You state in your first attempt that you wrapped the first ply with electrical tape. How tight should this be?
A: The tape compacts the carbon, making a stronger structure, and squeezes out excess resin, saving weight. Therefore, it should be wrapped as tightly as possible. In my case, the limit on how tight was possible was collapse of the foam. The first area I wrapped slowly shrank in cross section right before my eyes as the epoxy warmed the Styrofoam. I now think rigid polyurethane foam would better resist this. Even so, the first ply or two of carbon could be wrapped loosely with tape to hold it in place until cure. After that the next plies of carbon could be wrapped quite tightly with tape because the cured first ply or two of carbon would be a rigid shell on the foam core. Make a simple trial part first to learn the limits for the foam you have.
Q: How thoroughly was the ply wrapped? Completely covered, partially attached?
A: Completely covered,serving two purposes:
- Epoxy that cures exposed to the air develops an amine blush that must be washed off. Then you can sand the cured surface as usual to prepare it for the next plies. On the other hand, epoxy that cures without exposure to air does not develop the amine blush. It needs only to be sanded.
- Cloth that is not compacted by the tape begins to "float" on the bulging resin that squeezes out in between the tape, and cures into a blister-like shape that should be sanded smooth. This nearly always removes the ply(s) that were left uncovered at that spot, necessitating a repair of the area.
Q: Do you wet the surface of the down tube with the epoxy resin and add the wetted cloth to it?
Q: Do you do this with each additional ply?
A: Yes, only just wet out the cloth (don't add more epoxy to the down tube). I recommend laying up about four plies during the same open time, i.e., just keep on wetting them out and adding them within the wet time of your resin (fifteen minutes?).
Q: Do you use the same 80 grit sand paper between plies or do you use something finer?
A: Yes, after the four or so plies have cured overnight, use 80 grit to rough up the cured surface in preparation for laying up the next four plies.
Q: Is it hard to remove the electrical tape (that has the pin holes in) after the epoxy has set?
A: If you wrap sticky side down, yes. I learned to wrap it sticky side up instead. It removes easily.
Q: The technique that looks the easiest would be: apply the mixed epoxy to the core. Then adding the precut cloth (dry) over the frame. Another layer of epoxy, another ply of cloth.
A: Consider wetting out the cloth on a sheet of plastic before putting it on the frame. Like this:
- cut paper patterns to shape
- cut cloth to the patterns
- mix epoxy
- wet out cloth on the workbench, then lay it on the foam core one ply at a time
- while still wet, compact the plies. I used vinyl electrical tape. Many use a vacuum bag system.
Q: From your experience, does this form a strong frame?
A: Yes, but it can be heavy because the excess epoxy is often trapped. I finally began pricking holes in the electrical tape immediately after wrapping it to let some of the excess out.
Q: Is compression required?
A: Yes. Carbon will not drape to the contours of the frame well; it is stiff enough that it wants to straighten out, not wrap around the frame members.
Q: Will I lose frame strength by not compressing?
A: Yes: uncompressed carbon will not laminate completely to itself nor to the other plies.
Q: One thing I'm not sure about is whether you allowed the resin to fully cure between plies or not.
A: I did at first, but since then I've learned it is better to apply the next plies before cure is complete. I think of epoxy as going through three stages as it cures.
- The first stage after mixing is of course the liquid stage, when I wet out the plies and lay them up.
- The second stage is gel, when the epoxy has cured enough to hold its shape (pressure can be removed), but is not yet fully cured. The West System epoxy I used took about four hours to reach this stage at room temperature. In this stage I believe you can add the next composite plies (wetted out with their own freshly mixed batch of epoxy) and still get chemical bonding in addition to mechanical and adhesive bonding between the semi-cured and freshly-mixed batches. No sanding is required. In fact, if you try to sand, the paper just gets gummed up.
- The third stage starts when the epoxy is solid. It takes about eight hours for the epoxy I used to reach this stage at room temperature. Adding new plies on solid epoxy requires washing (to remove amine blush) as well as sanding (to provide a rough surface for mechanical bonding). Fresh plies added this way adhere to the cured plies by adhesive bonding and mechanical bonding, but lack the added chemical bonding you get when epoxy co-cures as in the gel stage as described above.
West system has an excellent online description of cure stages.
Q: Is there any metal under the carbon fiber?
A: Only the head tube sleeve and rear dropouts (co-molded with the frame) and bottom bracket shell (bonded in after cure).
Q: On your web site, you have indicated that aluminum and carbon fiber react so as to cause cathodic corrosion. Is this also the case with steel or cromoly steel and carbon fibre?
A: Yes, it is a problem with steels as well as with aluminum alloys. The easy solution is to prevent contact. One way to do this is to include a single light ply of fiberglass in the layup as an electrical insulator between the metal and carbon. Trek did this with its bonded aluminum-lug carbon-tube bikes. Specialized did too, and Kestrel uses small patches of fiberglass in the fork and frame where metal parts attach. Aerospace structures do the same. See NASA's Corrosion Control and Treatment Manual.
Q: Bonding of the Aluminum BB shell. Have you had any failures in BB/epoxy bonding since using the cycle?
A: No failure yet after four years of riding! The exterior surface of the aluminum shell has a very heavily knurled surface. I think this helps the epoxy to capture the shell mechanically .
Q: Did you do any surface treatments to the BB prior to the epoxy bonding?
A: I cleaned it with acetone. Now (2001) I know better, and would also use West System's two part aluminum etch kit. [Note, West System no longer recommends the acid etch, as its G/Flex epoxy bonds to oxidized aluminum. See test results here and West System's Use Guides for recommendations on special preparation for various materials. -- John Allen]
Q: What brand of epoxy did you use?
A: For bonding I used 3M's DP460, dispensed through its DuoPak squeeze gun cartridge. See 3M's web site for more info.
Q: How can I paint over carbon-fiber wheels (e.g. Specialized Tri-Spoke, disc wheels and Zipp wheels)? How hard is it to do custom paint on carbon fiber? What materials do I need? Any recommended readings or URLs ???
A: Yes, you can paint over carbon fiber. Call JB for a pro job, or do it yourself. In my experience this is how it's done:
- Remove stickers and their adhesives (I use DX-330 or acetone).
- Hand sand off some of the old clear coat or paint, if present. Decals under the clear coat can be sanded smooth this way. Stop sanding immediately at any spot if you touch carbon. It is incredibly easy to remove carbon by hand sanding. Watch especially near sharp edges (airfoil trailing edges on wheels, frames and forks) and parting lines on molded parts (often along the center line of frames and wheels, often along the sides of forks). Wet sanding helps keep the dust down.
- Clean the surface mechanically and chemically as follows:
- - Mechanically clean: remove stuff like paint flakes, old bits of dried substances, stickers and their adhesives, decals, flapping plies, scale, crud, etc.
- - Chemically clean: remove grease and oil: fingerprints, real grease from old bike parts, oil from the old bike chain, etc.
- - One last wash: rinse and scrub with acetone until white paper towels come clean.
- Prime using Fill'n'Sand or similar. Fill'n'Sand is just a high-build primer intended to hide small imperfections, which, depending on your level of perfectionism, is optional ;-). Any similarly described product should do.
- Paint using regular paint (Imron, Deltron, PPG, spray can, etc.) as usual. Do not exceed 100 degrees F. Some pinholes may appear. I think this may be outgassing. Bake at lower temps or let the paint dry at room temp to avoid.
See West System's recommendations for preparing a newly cured epoxy surface for paint.
- Some of the chemicals used in the procedures described here are highly toxic. Follow safety instructions carefully.
- Do not media blast with any media! Not sand, not plastic, not glass beads, not walnut shells, not steel or ceramic shot. Nothing! Not at even the lightest pressure or for even the shortest time. It is incredibly easy to eat away the carbon and epoxy. Before you know it your part is ruined!
- Do not use chemical paint strippers! They will attack the epoxy resin that holds your part together. Acetone, alcohol and other solvents are okay, but they don't remove paint.
- Do not use heat! Some epoxies undergo glass transition as low as 150 degrees F or so. Keep it under 100 to be safe. If it is too hot to touch with your bare skin, it is too hot. That means no power tools (buffing wheels, sanding discs, etc.) and no heat gun paint strippers, etc.
Home back to Damon Rinard's Bicycle Tech Page
Copyright © 1999-2001
|Articles by Sheldon Brown and others|