Mountain Bike Reviews Forum banner
1 - 20 of 59 Posts

·
Registered
Joined
·
126 Posts
Discussion Starter · #1 ·
As the thread title says I am working on a carbon fiber composite frame. I have never built a frame before, so I have spent a lot of time reading, planing, trying to figure out bike specific parts, measurements etc.. I asked a few questions here on this forum about the things I didn't understand. I am confident that I have thoroughly thought this through and I have begun the work. The general idea is to make tubes by bladder molding and then to glue them together using unidirectional carbon fiber wrapping. This is not very elegant compared to more refined more-or-less monocoque techniques, but I believe it is challening enough to start with.

I made a CAD model of the frame and printed the outline on paper. The paper is glued on an MDF board to shape a model of each tube. Here is a picture of the model of top tube. I coated it with three layers of a gelcoat (clear, dark and light gray) and sanded and polished it. I used three colors so that I can see when I get close to sanding through the coating. Sanding and polishing isn't really my strong suit. Let's hope I get better with practice.

Green Parallel Composite material Rectangle Stationery

I got a scrap piece of chipboard to support the first half of the female mold. The outline of the tube is cut out to allow the model to fit in. The gaps are filled using plasticine. Additionally, small pyramids of plasticine are used to as alignment help for the mold halves. I got this idea from a vacuum infusion tutorial on YouTube. I took those pictures in the middle of the work. The plasticine has to be cleaned up before proceeding.

Composite material Plastic Plywood Aluminium Steel
Blue Electric blue Azure Cobalt blue Hardwood

I made both female mold halves using a black gelcoat layer and several layers of glass fiber mat on top. This is the first half. Unfortunately I have some irregularities in the surface because of trapped air.

Photograph Atmosphere White Space Black

I decided to fill the irregularities with a polyester filler after applying the release wax. It's not pretty, but I thought it might help in avoiding further air inclusions. I have successfully split the mold halves and removed the tube model. There is a defect in one spot, where I probably missed another air bubble, but otherwise the molds turned out alright I guess. I took a picture after washing off the PVA release agent. It was still partially wet, hence the blotchy appearance.

Space Metal Silver

I will make the preparations for the bladder molding next.

Daniel
 

·
will rant for food
Joined
·
3,996 Posts
If you had a bubble free result on your first try, you'd be a liar.

Everyone makes bubbles their first try. First ten tries.

You going to carve a neck from your existing mold in order to fit a bladder neck?

What bladder material are you planning on trying?

You mention an infusion tutorial - are you planning on doing pressurized infusion, or wet fiber layup and pressurized epoxy ejection?

I want you to know that you started a CFRP related thread WITH PICTURES - this is unusual and appreciated, you are ahead of the game, bud.
 

·
Registered
Joined
·
126 Posts
Discussion Starter · #3 ·
Thanks, Drew! The plan is to make a cutout in the mold halves where I left the plasticin blobs at each end of the tube. The larger one will be for insertion of the bladder and the smaller one to get the air out.
I still need to figure out how to apply the gelcoat. I have a UP gelcoat, which needs to be cured before the epoxy comes into play. This means I will have to apply it in the closed mold somehow, because there would be a seam if I were to let it cure in the separate halves. Maybe spinning will work.

I will try a wet layup with a room-temperature curing epoxy. In case I don't get good results this way, I can try to combine bladder molding and vacuum infusion. I have seen tubes that were made this way, so I know it can be done.

I am considering two bladder materials. PE film (e.g. garbage bag) because it does not stick to epoxy and EVA because I can get thicker film that can take a lot of stretch. I will have to make the bladder shape close to mold shape because of the near rectangular cross-section, so stretchability is probably not very important.

Daniel
 

·
will rant for food
Joined
·
3,996 Posts
I'm afraid I don't have much advice for you on gelcoats, beyond the notion they are heavily recommended to me by others more experienced with cosmetics than I.

I think the theory behind gelcoat is that it is the one place where epoxy travels the least willingly under pressure (against the hard clamp), yet conveniently is easily coated.

It's like when people keep on cutting off a sharp corner of singletrack, and the local dirt boss throws up his hands and builds a feature out of it instead.

One thing needs to be said: seamless molding is HARD. You have to get the fiber bundle into the mold while still allowing the bundle to unfurl on command when pressurized. It can be maddeningly awkward. One veteran put it to me as "trying to wrestle a pissed off baby octopus".

That's a good idea regarding EVA. I'll have to check that out. I often use platinum silicone because of its unwillingness to stick to anything. Epoxy just flakes off it. Fair warning, silicone also has high toughness blends, but they tend to run: start a small tear, and the tear will continue. I've blown up so freaking many bladders, please please please do a dry fiber pretend run first. Hearing a "pop" and then having to wait half a day for the mess to cure is a total b!tch.
 

·
Registered
Joined
·
126 Posts
Discussion Starter · #5 ·
The gelcoat has to be brushed inside the mold and cured. I will try to apply it right up to the edge. Maybe some spinning can make the coat merge. If a small seam in the gelcoat remains I will just have to do some additional finishing work with additional gelcoat after the demolding. According to the tech support of the manufacturer that is the standard procedure.

Did you use silicones that can be brushed on a male model? I considered this option to get a pre-shaped bladder, but I decided to try other materials first because the silicone is not that cheap and if I don't get a reasonably even thickness it might not inflate nicely.

I have some test data from a few EVA grades from work. I thought EVA would be a good candidate because it is much more compliant than PE. The downside is that it will stick if no additional measures are taken (I am thinking additional thin plastic wrap). The material I got is an encapsulation material for photovoltaic modules, so it is basically used as a glue.

Before I do any laminating attempts, I will try to make a Finite Element simulation of the frame to get a reasonable estimate for the layup. I have looked at steel tubes. The wall thickness is around 0.5 mm. This is the target stiffness.
 

·
Registered
Joined
·
24,083 Posts
Man, talk about really taking the bull by the horns, definitely subscribed to see how this turns out.

With regard to the gel coat, I think that if you brush it on the 2 halves and then clamp together and spin like you say it should give an even coat and minimize a seam, but even if you do get a seam then it's just some sanding and maybe filling to be done.
 

·
Registered
Joined
·
90 Posts
Great project- very bold

Couple of questions:

Why do you need a gel coat? If you do a wet lay-up, it should create a good surface on the outside of the tube. It will also add weight. I imagine that you will be joining these tubes together and then painting the whole frame?

There's no mention of what you are doing to seal and prep the mold for wet epoxy? Those surfaces have to be sealed, polished, waxed then coated with a mold release agent.

If you haven't already spent the money- I would look into Entropy Resin. Its seems to work as well as West System, but its vegetable based. The fumes aren't nearly as toxic.

Last crazy idea: I've often thought of trying a similar molding technique (not for bikes) and I always thought of buying plastic tubing so it won't stick then using an old bike tube for inflating. You could cut the tube in half and seal up the ends to the appropriate length.

Looking forward to more details and pictures!
 

·
will rant for food
Joined
·
3,996 Posts
Did you use silicones that can be brushed on a male model? I considered this option to get a pre-shaped bladder, but I decided to try other materials first because the silicone is not that cheap and if I don't get a reasonably even thickness it might not inflate nicely.

Before I do any laminating attempts, I will try to make a Finite Element simulation of the frame to get a reasonable estimate for the layup. I have looked at steel tubes. The wall thickness is around 0.5 mm. This is the target stiffness.
Silicone can be brushed or slushed or dripped, depending on the particular blend's viscosity. They run quite a range. You're exactly right about the absurd expense, and the sensitivity to wall thickness for proper inflation. Any thicker than 4mm and nothing stretches.

Really though, if you're making a straight tube, you might not need a custom bladder. I'm thinking polyethylene flat-roll tubing might suit this first task.

0.5mm carbon wall thickness? Not very many plies there. I hope you plan on extending the lug work out a ways to simulate the butted wall thickness of a steel reference tube. (That is sort of the fun with composites... don't have enough layers? Then add some layers.)

As ru-tang suggests, and for other reasons, I'd steer clear of West epoxies. Far too viscous.
 

·
Registered
Joined
·
126 Posts
Discussion Starter · #9 ·
The gelcoat is for UV protection. Additionally it can polished better than epoxy according the manufacturer.
I use wax and PVA as release agents. I am open to suggestion for something better. Maybe I do something wrong, but waxing and PVAing leaves ream patterns and the mold is not as shiny as it should be. Could be that I just need a finer polish. I bought an additional finer one recently, but haven't tried it yet. Note that the polish must be silicone free, if it's not it might interfere with the release agents.

I'm Austrian and I got my stuff from the German company R&G. They sell pretty much everything needed for making composites.

The 0.5 mm is not for carbon, it is the steel reference. I made some preliminary simulations of just the toptube today. I compared UD layups with a 0.5 mm steel tube of the same shape, looking at bending and twisting. 10 UD layers and 2 layers of braiding with 50 volume percent of resing should give a wall thickness of nearly 2mm and a bending stiffness higher than that of the steel tube, but only about half the torsional stiffness. The ratio of bending to torsional stiffness can be changed by having less UD and more braided tube. I need to spend some more time thinking this through. In particular I need to find out what load cases to look at. So far I have only come across an ASTM standard for frame fatigue testing.

I am not concerned about shaping the bladder, I just want a reliable material. For my rectangular top tube cross-section with lightly rounded edges the tricky part is getting even contact between the wet carbon fiber and the mold wall. I am going to make a thinner PU foam core simply by foaming into the mold and cutting it to size. The foam core will be inserted into the bladder, which will then be vacuumed to make it stick to the foam core. Then the layup will be placed on the bladder (I expect I will curse a lot while doing that). The idea behind using the foam core is that it makes it possible to start with the fabric at a reasonably uniform distance to the mold. I am not sure how to explain it better, but think of inflating a circle inside a square. When the circle touches the edges it still has some way to deform before it fills the corners. Hence I want to use the foam core to start inflating a square. It will still want to inflate into a circle, but it will have to deform less to fill the corners.

Daniel
 

·
will rant for food
Joined
·
3,996 Posts
Daniel you are all kinds of on the ball with respect to shape deformation / wanting to match the end shape as closely as possible with your fiber preform.

Carbon will settle willingly a little bit -- a little. Beyond that, it's just like with any cloth, lots of distortion is bad.

I think this thread is popcorn mode for me now. Keep it coming dude!!
 

·
Registered
Joined
·
90 Posts
I guess my thinking on the gel coat is that it will be an extra step and extra weight. If you are painting the frame when it is complete, that should function just fine as a UV blocker. I usually just use the same epoxy with colorants added to form a gel coat. Simple and you won't have any compatibility issues between products.

My composites experience is coming from boat building, so the tolerances and procedures are a little different.

Popcorn is being popped . . . can't wait to see more
 

·
Registered
Joined
·
199 Posts
Goals?

I'll weight in here since it's the only carbon action ATM.

Is your goal to build a frame, or learn to make dies/molds???

Good on you for making the attempt!

As with anything, doing something many times is the best way to get good at it.
Used to say, "when you've done it a 100 times..."
Since I am only on frame #'s 85-86, maybe I still have a ways to go?

Bladder molding combined with a "wet lay-up" is not practiced by any mfg's for anything that I know of??? For good reasons.

Seems like it will take an awefuly long time to build; plugs/molds/bladder, foam insert???
If that is your goal, then go for it.

When I wanted to "build an MTB". I only had so much time and I wanted to see results quick.

Are you able to build your own assembly jig for alignment?

Tons of commercially available tubes are "mandrel wrapped".

You can build very strong, consistent tubes (including tapered/radius)
quickly.

Built the main triangle for my first MTB, (3 tubes) in 6.5 hrs. Including making the mandrels from Maple I had lying around.
The "S" shaped tube was made in 2 pieces with a tenon so both ends could be removed.
Started with the BB shell bolted to the seat post mandrel so it would be square and co-molded with something.


After (I call it "slapping and wrapping") the main tubes.
I added a biaxial carbon sleeve and used "shrink tubing"
That created a production quality finish that was ready for paint.
Using high-temp resin will prevent bubbles when using shrink-tube.

Just a thought that you might want to re-consider the "process",
if you want quick results.

If/when you get one complete. You will surely learn a lot and move on to another?

Room temp. pre-pregs can be found that do not need to be kept that cool if you are set on bladder molding.

"have a bar, (Kona handplant) that I want to pull a mold from.
I would leave a "parting line/flange big enough to use breather cloth/vacuum, to remove trapped air.

Another trick is to put precisely cut pieces in both halves before closing the mold/bladder. That will keep your entrapped air hidden ;-)

Keep at it, we need more carbon builders to piss off the "steel is real" crowd :)

JM
 

·
will rant for food
Joined
·
3,996 Posts
Eh I don't think I want to piss off any steel folks (I've ridden many of those, they're good), but yeah this place is painfully underrepresented regarding carbon and needs a shot in the arm.

You guys ever see this? Entry level filament winding. Got one in my basement. It still requires some operator skill and knowledge, but it's pretty neat. X-Winder Store

I like to wrap polypropylene mandrels with 200F prepreg, cover it in heat shrink film, cure it... after cure, the cooling back to room temp causes the polypro to downsize significantly (about a mm across a 30mm diameter rod, so it's not for making tubes with precise tolerances because the mandrel varies so much, making final size calculation tricky) and the mandrel can then be pulled out easily.

No one right way to do this stuff, it's great.
 

·
Registered
Joined
·
126 Posts
Discussion Starter · #15 · (Edited)
Wow, thanks everyone for your input and encouragement!

ru-tang, I'd like to paint the frame only partially. My road bike is a Look 481 SL with the Jalabert signature. You can find some pictures with Google. I really like the transition from white to visible plain weave carbon to red on this frame. I haven't fully decided yet - still a long way to go - but I might go for a similar finish.

Johnny, that bike looks very nice! There is a video from Felt giving an idea of how they do the bladder molding. I think it would be called a wet lay-up since they are using pre-pregs. It's certainly better in terms of fiber volume content than what I will manage. I calculated the weight of the top tube with 10 UD layers and 2 layers of the braided tube. For a fiber volume content between 40% and 50% the weight is between 245g and 290g. The same tube in 0.6mm steel weighs 395g (how do you guys measure such weight? Ounces?). Taking into account the additional wrapping and glueing to join the parts, the final weight should be about the same. I would have expected a larger weight. Is the average steel tube really this thin?

I have seen a setup for bladder molding with vacuum infusion. The dry fibers are expanded and then the resin is infused by applying vacuum. I think this is a very elegant method and perhaps I'll try it at some point. It should help towards achieving a higher fiber volume content.

I will most likely use a mandrel and wrap it to make the seat tube to make sure I get the inner diameter right. I don't think I'd manage that with bladder molding. I'll do the same procedure for the chainstays and seatstays because I'll have to fit them around the wheels, avoid the crank and the heels and also attach the dropouts. This is probably going to be complicated.

Anyway, this is what I want the front triangle to look like. The orange parts of the jig are aluminium profiles that I already have. The yellow parts are various pieces like cones for fixing the head tube. Should be a simple matter of getting them machined.

Brown Yellow Bicycle frame Orange Line

There is some waiting time involved in the preparation for the bladder molding, e.g. 24 hours waiting between waxing and applying the PVA and another night for the curing of the gelcoat. It'll be a couple of days before I can make the first bladder molding attempt.

Daniel
 

·
Registered
Joined
·
36 Posts
First off: great project! As an aerospace engineer I find this highly interesting.
Just wanna give my 2 cents too.

The wet layup on your styrofoam core seems like it's asking for problems. Fiber misalignment, ondulations etc. I think it is mostly done with prepreg fibers, because you don't have to impregnate them first.
Working with prepregs at home however is of course not really easy not to mention the cost.

I don't know about bicycle tubes specifically but I have been building a model airplane body with a friend using the same process for the mold as you did.
The difference is that the layup is done in each half of the mold and the bladder is inserted before you close them. When you inflate the bladder the two halfs connect.
The advantage is that the fibers are already molded to the contours.
It is crucial to get a good seal, but lots of people use this process to build planes professionally so it should work. I will try to find a video of the process.

Regarding the bladder, we use Qualatex or Sempertex balloons made for professional balloon artists. They have a good size and shouldn't have to stretch too much to fill the whole tube. The question is if they are strong enough to compress 2mm of carbon fiber.
Just make sure they don't pop, like Drew said. Don't ask me how I know ;)

I'm also curious what kind of fibers you will use and what weights? I'm familiar with the R&G offerings and think some samurai cloth would make a great top layer :)
 

·
Registered
Joined
·
126 Posts
Discussion Starter · #17 ·
I made some first attempts at making a bladder. I tried to make bladders from PE films using a soldering iron as demonstrated in a video I found during my research on bladder molding. I made five bladders until I managed to make one that was airtight. I wasn't happy with the strength of the films I had though. I need to try a film that can take some stretch. I also tried a road bike tube, but it didn't take the required stretch to fill the tapered mold. So I am still looking for a good material. I will look at the ones you mentioned! Perhaps I can also get a suitable latex grade from a polymer chemist I know. Haven't given up yet.

I cannot do the layup in each half because I want to use braided sleeves. I got the 60mm diameter one from R&G, it's 6k. I also have 3k UD tape and plain weave fabric. I looked into load cases for bike frames and what I have seen so far is that the primary loads are torsion and sideways bending. In light of these loads, the plain weave fabric is not that good. I haven't completed the FEA, but I expect that I will be using the sleeves for more than just a nice looking top layer because they are my best option of improving torsional stiffness.

I had no trouble laying the fabric on the foam mandrel, which is why I believe it will be fine to use this for the bladder molding as well.
I am considering the possibility that I will not be successful in bladder molding the top tube. As an alternative, I made a PU foam mandrel simply by foaming the mold. I put a layer of the fabric on one side and vacuum bagged it. I was curious about the surface quality and it turned out nice. There is a number of small pinholes, but I think it will look nice with a gelcoat or a two-component PU finish. I applied polyester filler to get a sandable surface. I'll post pictures of the mandrel when I am done sanding.

You're right about the samurai cloth. It would definetly make a nice top layer! Maybe I will get some when I have more practice. For now I will stick to cheaper materials.
 

·
will rant for food
Joined
·
3,996 Posts
I tried the welded polyethlyene sheet with a soldering iron thing as well. I had something like a 15% success rate at making airtight bladders. One of them sorta did the job for a little while, but it quickly became obvious that I didn't have the skill to make bladders that way.

One easy low-expense way of casting a bladder from a mold is to use mask maker's latex, very very thin out of the bucket. Have to rotate it otherwise you'll get thick/thin spots. Downsides are that the resulting balloon will stick to epoxy when left to its own devices, the application is noxious and requires a warm fresh air environment, and many applications are needed to build up wall thickness without drooping.
 

·
Registered
Joined
·
126 Posts
Discussion Starter · #19 ·
This is the foam core, roughly sanded with 100 grit sand paper. Nothing special about it.

Wood Hardwood Tool Wood stain Plywood

I am going to use it to test the handling of the UD tape and the braided sleeve. I am hoping to get a nice top tube out of it because it's a fairly expensive material. If it works fine I can buy a larger quantity at a less painful price.
 

·
Registered
Joined
·
126 Posts
Discussion Starter · #20 ·
These are braided sleeves and some UD tape prepared for laminating.

Mesh Natural material

I mixed the resin by weight to get a fiber volument content of about 40%. I put the UD tapes on the side to reinforce against sideways bending. The braided yarns are aligned at angle to the tube axis which is good for torsional stiffness. The braided sleeves are like chinese finger traps to work with.
I vacuum bagged it and waited for the epoxy to cure. Overall I am fine with the first attempt. There are some flaws that are not only cosmetic but also structural that I need to avoid in the future.

The cured tube with a resin seam from the vacuum bag

Line Black Beige Silver Shadow

Here is a closer view of the side. There are some small holes between the yarns here and there, which I expect to disappear when I put on a top coat. The braided yarns don't look as nice as a plain weave fabric because of the spacing between the yarns.

Line Carbon Grey Metal Composite material

There are two structural flaws in the following picture. There is a strip where resin is missing on the surface, no idea why. I don't think it is a big deal. There is one spot where the tube is indented. I suppose there was a void in the foam core that collapsed when I vacuumed. I didn't notice it until today. It is on the bottom of the top tube in the section close to the head tube.

Grey Carbon Synthetic rubber Metal Composite material

Now I have to decide what to do with it. I could finish it with filler and paint and use it. Or I could do a three-point bending test to test the bending stiffness and strength.
 
1 - 20 of 59 Posts
Top