Repairing a wrecked Trek Remedy 29 9.8 Carbon Frame
Repairing a high-end carbon fiber bicycle frame.
Disclaimer: I do not recommend anyone try this themselves. Improper repairs will lead to life-threatening accidents. Inhaling carbon fiber dust after cutting or sanding is carcinogenic. I suggest you get your bicycle frame professionally repaired — Calfee Designs offers carbon frame repairs for ~$500, for example, and given how time-consuming and risky this repair can be it’s probably well worth the money.
I’ve been attempting to build a carbon fiber bicycle frame for years. My perfectionism has always gotten the best of me, so when I saw a wrecked frame of one of my dream bikes for sale I decided to at least attempt a carbon fiber repair.
First, let’s see what the frame looked like when I received it. It looks like it had sustained a hard impact by a sharp object, bottom bracket area was probably hit by a sharp, pointed rock at high speed. Delamination is visible, the frame is actually punctured in one spot, fibers are torn and top layers abraded. I was hoping the damage was localized to what’s visible but it turns out it had spread a bit further and I had to remove more material.
After the first inspection I had to find out the extent of the damage and remove the damaged material. So the first step was to wet-sand the affected area around the impact. I find it is easier to judge delamination when wet-sanding, the water will permeate between the separated layers and squeezing the sanded edge will weep water when delamination occurred. The only way to tell for sure is to x-ray the frame probably. Warning: Sanding carbon fiber without proper protection may lead to lung cancer! Always wear a proper respirator — a regular dust mask won’t help with the fine carbon fiber dust!
After wet-sanding we have ourselves a pretty nice, big hole. Pretty large but manageable.
As a side-note: I got to peek into the frame’s interior through that nice, big hole and it looks great. Trek is known for their carbon fiber (CF) production prowess and this frame does not disappoint. Even the difficult-ish bottom bracket area looks good, no creases or folds, walls are smooth as you’d hope. CF frames sure have come a long way this past decade…
Here is the process of how I intended to repair the hole. A good carbon fiber laminate should at least be/have:
- strong enough,
- bond well with existing structures,
- good layer compaction (no air! or resin between carbon layers),
- no excess resin (extra weight and doesn’t help strength),
- good shaping.
In order to achieve some of those goals I figured it’d be best to use an internal bladder against an external mold/barrier/shaper (“cover plate” from here on). The bladder expands and presses the carbon layers together, pushes out (some) of the excess resin and should push the fibers to conform to our externally dictated shape. I was also going to laminate “inside” and “outside” the frame in two steps, so as to sandwich the edge of the hole between two layers for best adhesion. Below is a quick mockup of the intended process.
A piece of plastic (“cover plate”), tightly wrapped in tape, will be used to keep the exterior shape of the frame for the carbon fiber to expand against when the internal tube is inflated — think of this as your “mold” if you will. A plastic soda bottle was used here but I found a piece of a plastic shampoo bottle to work better. The shampoo bottle was softer and easier to ply around the frame to follow its shapes.
Initially, I had planned to use an internal cover plate as well. The idea was to laminate the carbon fiber on the internal plate which is prepped with a mold-release agent and tied to a steel cable for later removal. Then the internal plate would be bent, shoved into the hole and positioned from the outside. That bending and shoving into hole did not work at all, so I abandoned the idea mid-process and inserted the CF layers on their own onto the tube (which did not have the mold-release treatment… I had fun getting the tube out).
Steel shifter cables (and a wide ziptie around the far end of tube) were used to pull the tube into the frame. My frame conveniently had a drain hole below the bottom bracket, otherwise you may have to drill a small hole to get the tube where you want it.
In order to make sure the tube does not expand out of the large holes on both sides of the headtube, it was wrapped tightly in packing tape. I like to use a layer of cling wrap first to avoid tape residue.
Unfortunately, the tube exploded during the first test inflation at around 15psi. It was sliced open by some sharp edge or cable stop boss inside the frame. So all sharp edges and internal cable port bosses were masked with a lot of duct tape and cloth.
Now that I had a process and the internal bladder in place, it was time to get ready for the wet lay-up. It’s best to prepare all your materials beforehand, otherwise you’re going to get yourself a sticky mess.
First, make sure that all bonding surfaces are clean and roughed so that your laminate will adhere to the frame. I sanded the inside edges around the hole again, about 1.5 inches around the hole.
Before laminating comes the tricky part: figuring out a layup schedule. You’ll need to do your research. For example, the German site R&G used to have an Excel-based laminate calculator, ther are other web-based solutions. I really can’t help here, one needs to carefully consider the desired strength, safety margins, … The existing frame structure may be helpful.
I will leave out specifics so that my layup cannot be copied without research. Composites-Central has some good resources (e.g. here and as suggested go read velocipedesalon.com threads).
I measured the existing wall-thickness around 1.6–2.1mm so I went for several layers wide layers for support that were larger than the hole and fewer layers barely larger than the hole. One inch of overlap around the hole should be reasonable but I decided to have wider support so my fabric was cut a bit more than 1.5 inches wider than the hole. These are the internal layers, I added three additional layers externally in a later step. The external layers are much easier to control, and in hind-sight you probably want your internal layers to mostly hold the shape and the external layers to be structural.
As for orientation, this is where your intuition comes into play. My guess is that the bottom bracket area sees some tension from the downtube joining the BB and torsional forces from pedaling, so I went with a pretty isotropic layup comprised of 45 degree and 0/90 degree offset layers, slightly biased for 0 deg. You probably want to taper your layer widths slightly so the tube’s diameter doesn’t vary abruptly and cause stress risers.
Squeezing the layup into the tube proved pretty hard/annoying, especially after I test-inflated the tube and couldn’t fully deflate it. Below are the fibers after two multiple attempts of inserting them :/ The less external overlap you’re comfortable with, the easier this step is — but I’d rather be safe. There is some fraying due to the repeated insertion, I should have taken the picture earlier :) But the final results look great both inside and outside after pressurization via the tube and being fully cured.
Below are the fibers laid up, cling wrap over wet fibers and frame around hole (to avoid tape residue on other parts of frame), cover plate and then wrapped cover plate tightly in packing tape to hold shape. Then the internal tube was inflated to a little under 20psi. When I re-inflated it again two hours later to 22psi, the tube exploded again but the resin was fortunately already mostly hardened. It’s probably not wise to inflate over 20psi (that’s 20lbs per each square inch pushing apart the frame) as frames are really not built to withstand internal expansion. 20 psi was a guess, be careful and pick a pressure you think is right — if you over-inflate you will crack your frame and it’ll be done.
Here are the results after removing the internal tube and cover plate. There is some excess resin that was pushed out, I did not see as much resin on the inside. The excess resin had to be sanded down. Again, wear a (proper!) respirator when saning CF and work outside (or in a well-ventilated room).
As you can see the fibers follow the natural tube shape quite well. The whole area looks pretty good after sanding. You can see the repair pretty well due to the difference in carbon fabric.
That black stuff on the downtube is sticky tape residue from the downtube protector. Now if the downtube protector had done it’s job in the first place…
Since the ISCG tabs and bottom bracket are in the way and the damaged area is concave (easy to compress with tape) I decided not to vacuum bag the external layers and just compact them with electrical tape. This usually requires more cleanup but would be quicker overall.
I took few pictures of the external layup but I first broadened the sanded area to around 1 inch around the original hole, then applied three layers of CF 45/90/45 orientation, each layer growing in diameter to avoid stress risers.
Then cling wrap was applied, and then taped over and compressed with electrical tape. Small holes were pierced into the tape to drain excess resin. The tape wrapping usually leaves a wavy pattern where tape layers overlap and requires some cleanup.
Here are some shots after curing and removing the layers of tape. The wavy compression imprint I expected was not there. I usually compressed over a softer foam core as opposed to the harden carbon fiber surface here.
And now some final shots after sanding and clean-up. With a bit more work this could have looked even better but I couldn’t wait to build the frame and finally test-ride the bike!
The final result after 80, 120 and 400 grit sanding. This would require a couple more passes of finer grit to look really, really great and polished and some careful trimming around the carbon fiber edge area.
Some internal shots — apologies, it was very hard to get good pictures.
The repair is at the very bottom curved part, as seen straight on. The wall looks smooth and well compacted. This looks as well as I could hope.
If you look carefully you can see a bit of the inner tube’s texture and molding marks that were transferred into the resin when pressurized, that’s a good sign that we got some good compaction and pressure.
The bulbous part that sticks out of the wall near those metal bosses is a cable port and from factory.
And then it was time to build her up! So far I really like the bike though I’d like to try another rear shock (more on that later) for some of the rougher local trails where the Fox Float feels overwhelmed.
