In the previous parts of this blog series we went through making the molds, and by this point we had everything we needed to lay up the first carbon fiber part for our prototype eboard.
There are lots of different methods for making composites, but some of the most common ones are:
- Wet layup (without vacuum)
- Wet layup (with vacuum)
- Vacuum infusion
- Out of autoclave pre-preg
- Autoclave pre-preg.
We quickly ruled out wet layup with no vacuum because our parts are fairly complicated and need pressure to conform to the mold. We also ruled out pre-preg due to not having an autoclave (expensive highly pressurized oven), and out of autoclave pre-preg since it can be difficult to achieve a good surface with it, and our molds were not rated to the temperatures required for curing pre-preg. We also didn’t want to over complicate things with vacuum infusion, which requires more equipment and consumables.
So we decided to go with wet layup + vacuum. Next we chose the carbon fiber we would be working with, which comes in different weaves and weights. We initially bought some thin twill weave for the surface cosmetic layer since twill is stretchy and great at conforming to different shapes, and some thicker UD (uni-directional) weave for the backing layers.
UD weave has all of the carbon fiber strands running in just one direction. This would allow us to create a bias of strength in one direction, since carbon fiber is only strong in the direction of the strands.
However we quickly learned during our first layup that unidirectional sucks and carbon fiber is much harder than it looks in the videos. We found that the UD weave just does not conform to the part. You’d push it down it’d pop back up again due to the stiffness of the strands. The resin (epoxy) also has very little tack to it and doesn’t really help the carbon stick, if anything the resin is more of a lubricant keeping everything slippery.
Nevertheless we placed a few layers on top of each other, painting resin on with a brush for each layer, and instead we worked the layup into the corners of the mold after we had vacuum bagged it, hoping that would be enough to push the carbon where it needed to go.
It wasn’t.
We also were able to tear some of the UD strands away from the back side of the part, with the carbon filaments feeling dry like they had no resin in them. From this we learned that carbon fiber does not absorb resin as readily as fiberglass, and would require a better technique to wet it out fully.
After that we ditched the UD weave and opted for a heavier twill instead for the backing layers. For the next layup we used 1 layer of 300gsm twill for the surface, and 2 layers of 600gsm for reinforcement. The lighter 300gsm weave is more flexible and conforms better to the tight corners of the mold,, but using a heavier weave is more cost effective to build up the total thickness required to achieve the desired strength of the part.
We previously placed the dry fabric onto the part and wetted it out with a brush, but this time we laid the fabric onto a large flat surface and used a squeegee to work a large amount of resin into the fabric and repeated for the other side before placing the already wetted fabric onto the mold.
This yielded significantly better results. Far from perfect, but enough to work with for initial prototypes.
The next challenge was the enclosure. Man, what a nightmare.
We used 9 layers for the first layup, with some layers of the weave running at 0/90 degrees and others at 45 degrees for torsional stiffness.
We found that pre-cutting reliefs to help the weave conform to complex shapes just makes the carbon a pain to work with, because it likes to fray apart. So we decided to cut the reliefs for the enclosure after we had laid all the fully wetted fabric down, and then had to fold 9 layers of heavy carbon back onto itself, which technically makes 18 layers where they overlap…
Then we had to put the lid of the mold on top without getting any carbon pinched in the flange, and install 18 bolts while the resin was starting to set (~45 minute working time), shove the release film into the sticky mold with sticky hands making sure its pushed all the way into the corners that fingers can barely reach, then do the same for the breather cloth, having to use tape to keep it in place. The breather cloth also has to wrap around the entire mold so the sharp edges and things like the bolts didn’t pierce the vacuum bag. Then we had to slide the vacuum bag tube over the mold and also stuff that into the inner corners. We’d then pull a vacuum, and really try to work the carbon into the corners of the mold by pressing hard with our fingers. Its very easy to have the vacuum bag or release film bridge across a corner, which would cause the carbon to pull away from the mold and leave bubbles or voids.
The results were actually a lot better than anticipated (total disaster was anticipated).
Not perfect, but usable.
The next stage was to trim the excess material from the parts. We used a band saw as well as a Dremel for the enclosure.
For the deck we waxed the finished carbon pieces and laid up a thick fiberglass template onto them, then removed the templates after curing and had it waterjet cut to the exact shape of our deck. This way we had fiberglass templates that would perfectly key onto our parts for tracing the deck shape and drilling the mounting holes accurately.
The next step was to glue the two halves of the deck back to back with more epoxy. A urethane bumper was also glued in at this step, which was trimmed afterwards.
We had great difficulty with the final step of shaping the deck after it was bonded. We tried lots of things, such as different ways of using routers to shape the edge the same way wooden skateboards are done.
We had the most success with stone grinding, however on first attempts we were too aggressive, causing the urethane to tear, and the curvature was made freehand – very difficult.
This still resulted in a finished totally usable deck, weighing just 1.1kg.
We refined the grinding process on the next deck we made. Using a shaping stone to make a curved grinding wheel, with a grit better suited to soft urethane, and improved our technique. We’re pretty happy with the result.
In the next instalment we’ll make the new suspension system, put some wheels on this thing and try it out for the first time.
