Carbon Fiber / Fiberglass Bodywork

[JonW]

Looking good. Interested to hear the story, I reckon you did well considering the setbacks and environment.

[wolfeman28]

Yes it all turned out ok. In fact, even though this first tank was a test, I'm actually going to use it as my main tank. I got the top/bottom bonded a few days ago and it has passed the 24hour no fuel leak test. Not happy with the cap height (I didn't add spacers as I mentioned I would do in an earlier post) but I think I'll get over it.

I'll use it until I have a few spare days to make a new one. Kinda cool to be able to do that :)

[JonW]

Kinda cool to be able to do that :)

[Yellow and Black RZ]

Amazing work.
Would be cool to do a black and yellow paint job where the black was clear coated carbon fiber.

[evan_calgary]

Very cool!

Where do you source those cap parts?

[wolfeman28]

Very cool!

Where do you source those cap parts?

Coyote Gear. They make all sorts of useful parts for fuel tanks.

[RuZty]

Bravo! Planning and preparation make all the difference, well done.

[wolfeman28]

Ok so lets start off with some small parts I made. This pertains to the right side motor covers and gas tank mounts. Just a caveat to what I've mentioned in prior posts- I ended up deciding to cut off the stock tank mounts and make molds of them, due to the complications involved of trying to leave them on the tank and making the bottom tank mold multiple parts. By removing them, I was able to make a one part bottom tank mold which you will see later.

I started off preparing for these molds by giving the parts a very light wet sanding followed by a mist of 2 part clear. Following that drying, I wet sanded and then applied 3 coats of Partall Paste #2 wax to the parts. Waxing helps with pulling the plug from the mold. It's not perfect on its own, but helps.



I then used various polypropylene board and other flat plastics to make edges for the molds. In these pics you can see the methods I used. There are many ways to skin this cat, and a lot of different materials can be used. The key is to only use materials that won't react or break down when in contact with the resins you use. The styrene in polyester resins (those most used in tooling/mold making/etc) can ruin many different materials. I utilized wax to fill gaps and on all edges. you want to get as close to a true 90 degree angle as you can on the fillet without going 'negative' of that, or else you will never get plugs out of the molds.









After the flanges are done, I spray on some PVA. PVA is basically a glue that is water soluble (yes, it's basically the exact same thing as hairspray!). It is thin enough to spray in light coats, leaves a nearly flat surface on the part, and cleans off easily. You use the pva to ensure that the plug does not bond to the gelcoat from the mold. At least that is the goal, but sometimes things don't go as planned. Sometimes you miss a spot, sometimes it lifts from the part when laying the gelcoat, etc. That is why you always wax the part before the PVA. Two release methods is better than one. I apply it using an airbrush for small parts, but you can also use a HVLP gun or even brush it on. Brushing does tend to leave streaks which will tranfer to your gelcoat, so on molds i definitely recommend spraying. A light mist coat over the whole part, then wait 15 min or so and hit it again with a slightly thicker coat, wait another 15 min and give it another coat. You should see a light green haze over the part, and when you touch it it should not be tacky.



And here started my issues with weather, as can be seen in the next photo. The day I sprayed these parts with PVA, the humidity was 98-99%. The PVA just would not dry under these conditions. I tried multiple methods to get the PVA to lay down properly, but nothing would work quite right. My mist coat wouldn't dry even after an hour, and applying a coat after that would just cause it to run. I tried using a blow dryer on low heat to get the alcohol to evaporate quicker, but that would just cause the PVA to shrink and wrinkle. You can see the 'quality' of my best attempt that day in the following picture. Not good.




So what was the solution? I waited until about 3am and resprayed the PVA. Humidity at that point was near 70% and it went on in 3 coats just fine. Goes to show how long these processes can take, even just for small parts. At that point I needed to get some sleep, so waited til the morning to attempt applying the gelcoat. Well, next day was only about 60 Deg F, and about 90% humidity. Not at all good conditions for gelcoat and lamination. Due to my time constraints, I decided to attempt anyway. I wasn't all that concerned with these small parts anyway. For gelcoat I'm using Composite Envisions polyester tooling gelcoat. This was my first time using this product, and I can tell you that not all tooling gelcoats are created equal. I'm brushing on the gelcoat as I have no air compressor in this garage that can meet the needs of an HVLP gun, and since it is a borrowed space I can't have over spray coating everything in the garage. Brushing on gelcoat is fine and is done all the time, but all variables regarding weather and the gelcoat itself need to be perfect for the task. I found out quickly that this gelcoat is extremely thin for brushing, I can tell that it would be perfect for spraying. I tried to let the resin gel in the cup for a bit to get it to lay down thicker, but that didn't work too well. My other option was to get it on as thick as i could, and then attempt a second coat a few hours after the first had kicked. This becomes a problem since even after hardening, thin spots of gelcoat will lift or 'alligator' from the plug surface due to the styrene from the second coat affecting those thin spots. This problem is amplified from the lower temps slowing cure times, and high humidity making the styrene hang around in an invisible mist for longer than it should. Best to have a fan close by to try and move the vapours away from the item. As you can see from the pictures, this stuff would NOT go on thick, no matter how hard I tried. The pic of the gas tank mounts shows how thick an application should be, this was after a second coat. I just had to pray that I didn't get too much alligatoring on all the parts.




For some reason I didn't take any pictures of laying down the laminates, so bear with me here. I waited around 8 hours for the gelcoat layer to come to a tacky state, then mixed up some laminating resin. It is always good to use a laminating resin that is designed for tooling, since it has a lower tendency to shrink while adding layer upon layer, which would mess up the tool obviously. I start with one layer of fibreglass veil, it's an extremely thin fibreglass with a binder that goes on super easy and helps with eliminating print through of the thicker chopped strand layers. I let this layer cure to a tack (about another 8 hours) and then continue adding another 4 layers of chopped strand mat (1.5 oz think I used?). After 12 hours I pulled the parts from the molds, which released with ease. I did get some alligatoring as I expected, but luckily it was all on the flange areas. I trimmed down the excess edges, cleaned the molds, and went ahead and made up a filler to flatten some of the ridged areas from the alligatoring that would cause stuck parts during layup. You can see a bit of the filler (white) in the following pic. It's simply tooling resin with cabosil as a thickening agent, and then sanded flat.



After that, they all get another cleaning, wet-sanding, and many coats of wax. That's it, set them aside and onto the next parts.

[wolfeman28]

Next up was the front fender mold. I ran into almost the same issues as with the prior smaller parts, which you will see in a bit.

Same as before- clean, wax wax wax wax, flanges. This part had to be done as a 2 piece mold, as the indentions where the large triangular holes are, and the mounting holes, would make for an impossible to remove part. It is possible to make it a one piece mold(factory mold was almost certainly 1 piece for all parts on the bike) but you would have to do some major filling in of all the detent areas, and then you would be left with no guide lines in the mold on where to cut. Just an easier/better choice to go with a multipart mold. Just takes a day longer to do, as you need to fully complete one side before starting on the other.

Here you can see the split line flange and bottom flange. Also put barriers inside, and filled all gaps with wax. I wanted to have all mounting holes visible on the finished parts as to make drilling easier, so I made the wax concave in these areas.



And how I held up the barrier on the other side. Once again, one of those things where there are many ways to accomplish this. just think outside the box and use whatever is easiest for you. I used some foam blocks and double sided tape.



At this point I sprayed on some PVA. Bad humidity and temps made me run into the same issues as before. The PVA finanly dried but looked horrible, but I decided to carry on anyway (running out of days here). That was a mistake. While brushing the gelcoat, the pva ALL lifted from the surface. This was a 100% humid day so I attribute this to a layer of water be stuck under the PVA, and my attempts to heat dry it cured the outside but not against the part, trapping the water in. I tugged along anyway, thinking it would be saveable. The gelcoat layer took FOREVER to get even slightly tacky, and when I put the second coat on, within seconds I could see the whole part shrivelling and alligatoring up. It looked REALLY bad, and I called it quits at that point. I knew that since I saw the PVA lift, the part was almost certainly going to be bonded to the gelgoat layer. It was, and took off alot of decal and clearcoat from the fender. This is why you always wax, as if I hadn't, the fender and gelcoat most likely would have been permanatly bonded together. As it turned out, I just needed to give the fender a few rounds of sanding and all was back to normal.

First gelcoat layer, you can see how thin the gelcoat is laying down. At the recommended 30 mils, you should not be able to see through to the part.



After pulling off the gelcoat to try again. You can see some areas where the gelcoat did make permanent bond with the part. Those where the main areas where the PVA pulled away.



Second attempt on that side. As you can see, it looks much thicker. What I did was add some cabosil into the gelcoat to thicken it up. It worked.



Pics got slim at this point due to rushing and all the issues, so sorry. Again I laid down viel, 4 layers of CSM. Next day I pulled off the split flange barrier, cleaned off the wax, sprayed PVA, gelcoated, and glassed the other side. I key note here, is that on the day I did side 2, it was perfect conditions, about 72 deg F and 20% humidity. Next day after it was all cured, I drilled holes in the split flange for bolts to hold the mold together.



Might be hard to tell from the next picture, but the quality of the 'perfect day' sides finish is pristine. I had to rework some alligatoring on the first half using filler and sanding. All things regarding prep, gelcoat, laminating, etc where the same on both days, just weather differences. Amazing.



Cleaned up the mold, waxed, and on to the next parts.

[wolfeman28]

Here are a few small molds I made for the upper fairing mounts as well as the solo seat 'o-ring' hooks. Pretty straight forward on these- wax, flanges/barriers, PVA, gelcoat, veil, glass. You can see that I cose to use tape as a flange on some parts. resins don't stick to packaging or painters tape so they release easy, and you can apply them to curves easily. Also you may notice that on the upper I used clay to make a perimeter so that the gelcoat couldn't get to unwanted (no PVA) areas.











Finished molds before cleaning-

[wolfeman28]

Now on to the tank!

I started off with the bottom. Note- I did this mold on the same day as the failed fender half, so same crappy conditions and thin gelcoat apply. As mentioned before, by removing the tank mounts I was able to do this mold in one piece. I still needed to do some careful placement of clay fill-ins, if I wouldn't have done this the part would not have released from the mold surface. You can see the areas in the following-






The PVA went all wrong just as with the fender, gelcoat went on thin and ugly, second coat caused massive alligatoring. I decided to attempt to save this part though, as my time and tooling gelcoat was getting scarce, and I wanted to be able to get the upper parts of the tank done as well.




You can see in the next 2 images how bad the alligatoring was after the second coat of gel. If there was a brush mark or ridge on the first gelcoat layer, it lifted up on the second coat.




As stated before, I carried on with it. Applied the glass layers and pulled the part. Looked crap, but I filled in all bad areas on the part and sanded best I could. Ended up with an ugly, but useable, bottom tank mold.

[wolfeman28]

Move onto a few days later. The weather was better, and I now had a good idea of how to work within the limitations of my chosen tooling gelcoat, so I tackled the upper tank mold parts.

Did this in three pieces. Took one day for the center, one day for the left, and one day for the right. Same process as before, only difference being I've now been adding cabosil to the tooling resin to make it thicker. Allows me to brush on a more uniform coat, and only a single coat is now needed. Barriers were cut best i could from signboard, then tapped into place and held up with foam blocks and double sided tape. Lots of wax to fill in gaps.







Next day I removed the barriers, PVA'd one side, and layed down more gel. Although I thought my gelcoat problems were solved, I ran into yet another issue on the left side of the tank. Since I now had thicker gelcoat, it was possible to place it too thick. This causes issues of exothermic runaway, causing the gelcoat to cure very rapidly under high heat. This lifted up an area around the top of the 'knee curve'. I could see it happening after like 15 minutes of cure time, so I tried to do damage control with some weights during the gel cure and the laminating layer cure. It helped, but I knew there would be some issues afterwards. Will get to that in a bit.





Right side of tank went on without a hitch.





Added holes/bolts and trimmed.







After removing the part from the mold, I could visibly see the lift on the left side, about a millimetre higher than the center half of the mold. So I mixed up some filler and blocked it out best I could. It matches the opposite side perfect so I'm ok with the outcome. You can see the problem area in this next pic, as well as a few other areas I touched up. I thought these areas on the right and back were also slight lifting of the gelcoat, but after inspecting my OEM tank, these were actually defects in the 'real' tank that i just couldn't see at a glance.

[wolfeman28]

Now for the carbon test parts I've done so far.

Keep in mind these are just tests, I use them to gauge any problem areas such as corners that don't wet out or lay nicely, amount of material to be used(carbon/resin/bagging consumables), etc. So the parts will not come out as 'production' runs would.

About my vacuum setup- cheap China vacuum pump, vacuum reserve system, and vacuum cycleing system. The reserve tanks just make it easier to draw down a large bag quickly. The electronics shut off the pump at 27inHg and kick it back on in case of slow leaks at 23inHg. This keeps the pump from having to run constantly, which would overheat one of these cheapos in no time. Ideally you'd want to ensure your bagging technique provides you with a complete seal with no leaks, then the pump won't run at all after pulling initial vacuum.



Bagging materials is Stretchlon 800. I find it to be enough stretch to not be so worried with bridging, but it is robust enough not to puncture on sharp parts of the mold.

Econostitch for the peel-ply, poly breather layer. Mastic/tacky/butyl tape for the bag edge.

I don't have many pictures of my carbon layup and bagging unfortunately, since it's a time sensitive and messy part of the composite game. My steps-
1. Wax/buff mold 2 times
2. Spray PVA on mold, 3 coats. If PVA has any defects/drips/runs, strip PVA with soapy water and start over. Any defect will transfer to final part.
3. Brush on thin layer of epoxy into mold surface. This is not a necessary step. You can either spray a layer of epoxy compatible gelcoat, brush on a layer of epoxy, or use none at all. I like to do it since it ensures a perfect mold surface transfer. Wait some time for this layer to 'kick'. This time all depends on your epoxy, and differs between all systems. You want it to be hard enough to not be deformable, yet soft and tacky enough on the top to be able to still bond chemically with the next steps. For my general purpose parts resin, this is about 6 hours. For my tank resin, this is 4 hours.
4. Lay in carbon cloth. Should stick to tacky epoxy layer. Press firmly into layer, noting that once you press hard, it's not coming off(if you try, you can pull up the soft epoxy layer from the PVA). Wet it out with resin. Repeat until all layers are in mold.
5. Add peel ply followed by bleeder
6. Put in bag, seal it up, pull vacuum. Ensure no bridged corners, leaks, etc.
7. Let cure for 8 hours at room temp, then apply a few hours of heat. You can do this with heat lamps or an oven. All epoxies will have a different recommended time and temp for post cure. You don't need to post cure at all, but post curing enhances the properties of most resins (stronger, better bond, better chemical resistance).
8. Remove from bag, demold part, enjoy.







Finished misc. parts-


Finished tank-



Test fit of fender. Haven't sanded it down fully yet, hence the raised epoxy bond line. Also pay no attention to the carbon patter, i just used scrap pieces.



Petcock mounts on inside bonded with resin/cabosil/chopped carbon mix-



Tank side panel mount/petcock-



Side panel fits as normal-

[wolfeman28]

That's it for now. Back to work and I also have a large military function I need to attend for the next 4 weeks out of town. On the weekends I'm going to try and get 4 more tanks built and put up for sale. I'm also going to get the rest of the molds done one of these days, I'm going to be spraying the gelcoat this time in a booth so hopefully i can get pics of that. Want to at least get some black primer on my finished tank so I can ride the bike as well. Will look funny but I don't care.

Did I mention that this tank is about a third of the weight of the stock one? Only a few pounds. Oh I almost forgot to mention the tank bonding of top/bottom. I trimmed down all excess carbon, and made sure to sand the bonded areas a bit with 100 grit. I then coated the whole inside of the tank with another layer of the ethanol resistant resin, which sticks well to the peelply rough surface. I then mixed up some more epoxy with a tiny amount of cabosil to keep it from running too much, and applied it to the top and bottom bond lines. Stuck the 2 pieces together, and clamped good all around the perimeter. After about 4 hours I mixed up another 200 grams of resin and poured it into the tank, rotating it around to get another coat on all sides and let it get into the bond line area. Came out perfectly sealed, it's going on 4 or 5 days now I think with fuel in it, no leaks. I did see some fuel at the petcock, ended up being my 1984 fiber washers on the mounting bolts. New fibre washers fixed this problem. The old ones were solid as a rock.

[JonW]

Really nice write up Wolfey, felt like we were part of the process. The parts came out really well, just shows what can be achieved with determination and hard work. Well done.