carbon part is giving me nightmares...

General Composites Discussions Composites Talk
#1

Hi there, I signed up to try to get some help in making a part that’s been giving me nightmares, I’ve worked with infused fiberglass for years so I thought I knew my stuff, apparently not. The part is a .875od carbon tube with a wall thickness of about .090, it’s part of a handlebar system for race bikes. I started trying to make the part by wrapping wet out 11oz carbon around a latex tube, then inflating it inside a clamped aluminum mold. That turned out promising, but had some voids between the tow, so I tried lighter cloth at the surface, same issue, then tried vaccum on the mold before inflating, led to a whole set of other issues, then tried infusing, building new molds that integrated a channel to seal butyl tape for 100% vacuum integrity, still had problems, like lack of resin between the tow and areas of the fiber that didn’t lay flat on the mold surface. then I tried prepreg, building a mold that had an integrated heater and controller to ramp, hold and ramp down temperature. this time the surface was perfect, but because I had to build the part in two halves with one half underlapping the other on inflation, the parts were too weak and split at the seam… seriously about to give up on the project alltogether…

posting pics of the prepreg mold and destroyed part. any help is appreciated

#2

Have tried braided carbon yet?

#3

tried braided, tried 2x2 twill with a braided top layer too, tried wrapping flat sheet on the bias and straight…

#4

I think your pre preg one should work out ok if you improve the join overlap. I dont see why it should become any significant weak point.

Stagger the plys to stick up/out on one side of the mould and also stagger back from the other side (into the mould) too. This way all the plys are the same width and no need to change the size of every layer.

Hope this makes sense?

BTW, how many layers are you using? What weight material?

You probably want about 20% of the fibres oriented to support the circular shape and then 80% length wise along your mould.

#5

tried rethinking the prepreg layup, so I cut out 10 pieces each wide enough to cover the diameter they would be in the finished part, they were cut on the bias to allow for good expansion and not lock up when the mandrel inflated. I wrapped them on a steel bar a little smaller than the finished ID. I lined up the seams 180 degrees per layer, for what I assume would be the strongest possible joint. pulled the part and again, super weak, the cosmetically flawed wet layup parts are at least 3 times stronger… is it possible I’ve got the wrong specs for the prepreg or was sold expired prepreg? it’s still tacky and flexible so I assume it’s okay, not sure why it would be so much more brittle than wet layup

#6

A few questions about your bladder setup. How many PSI are you running while inflated? How “anatomically correct” is your bladder to the inside of the mold?

#7

running 80psi, the bladder is a section of epdm tubing, so it’s able to inflate to match the mold pretty well, it appears to have enough flexibility to compact the laminate uniformly

#8

I have been working with a similar project, fortunately your part has far less detail and contours than the part I have been producing, but I’ll share some things we have found. As far as cosmetics, we found that using a single piece of plian weave 3 or 6K carbon for the first layer backed with bi-directional and additional weave carbon gives a good combination of strength and cosmetics. You may be able to use a outer layer of carbon sleeve as well, since the part is a fairly consistent diameter. Second thing I found is ramping up the pressure of the bladder slowly helps let some of the air trapped between the mold and part escape. It is also important to have somewhere for the trapped air to go, some “bleed off” channels may be needed, which will let the trapped air evacuate with the flashing.

#9

smp4616,

My experience with bladder molding similar parts to yours makes me believe you’re running into several problems. Some of these suggestions may help.

With a .090 wall, your laminate preform is going to be extremely difficult to compress against the mold. The solution is to make sure that you use a bladder that once wrapped with your reinforcements, just fits the mold cavity. If your preform is smaller than the cavity by even a fairly small amount, it’s gonna be almost impossible to expand it. Especially if you’re using braided sleeve.

The aluminum molds you have are very nice and probably seal to each other too well. In order to completely inflate your preform inside the mold, you’ve got to have a way for the air around it to escape. By not clamping the molds together too tight, you’ll have a small gap along the entire length. If not, a few pieces of shim stock (only takes a few thousandths) should maintain sufficient gap. It will affect the roundness of your parts, but for my tubes, I allowed for this during mold construction and the tubes come out within tolerance.

Next thing is your bladder material. It doesn’t have to be a heavy wall material. I’ve found that actually the thinner the better. Also, you need to confirm that at 80psi, your bladder doesn’t leak. If you’re losing pressure through the cure cycle, your bladder leaks. I ran into this a while back and it gave me fits until I solved it. Even the most insignificant leak originating from under the laminate will create voids and hideous surface imperfections.

#10

The tube in this photo is 1.375" OD with a 0.060" wall. Wet layup with braided sleeve outer layer and inner layer with unidirectional carbon layers between. Bladder molded with latex bladder, ramp to 60psi, room temp cure.

#11

that looks perfect, what epoxy are you using? how fast do you ramp? what does the mold look like??? I have a different mold that has a .005" vent around the perimeter of the cavity that goes into a .250" diameter channel to bleed excess resin, but those parts always looked like they were resin starved. I definitely had the issue of either too small a preform not expanding enough and too large a preform getting pinched at the parting line… found out that if the bladder is even remotely constricted, it won’t inflate, but once it inflates a little bit, then it can exert force.

#12

The epoxy used for that tube is West Systems 105/206. I’ve had good results with others as well. The process has more influence over the result than the materials.

I typically ramp 10psi/minute. Starting at 10psi and ending at 60psi. It doesn’t take long, but the thickness of your laminate, viscosity of your resin and the vent area of the mold will greatly influence how fast you can proceed.

About venting the mold, having a catch area on the flange works as long as the trap is vented to the outside of the mold. Otherwise you are basically just pressurizing the resin trap. My tube molds just let the excess resin flow out the flanges and I catch it with some strips of wax paper or foil. Messy, but I didn’t think far enough ahead to mold in resin traps. My molds are due for replacement, so there will be some upgrades next time.

Definitely be careful not to pinch your laminate in the parting line. I’ve had to deal with this too. It’s especially tricky with a wet layup.

I’ll post some photos of the mold and related items shortly. It’s nothing fancy, but works extremely well. I’m sure with a few mods to your process, you’ll be able to produce very satisfactory parts. The key is then to repeat the process exactly every time.

#13

Here are a few pics.

The molds are built from tooling epoxy surface coat/glass/carbon/laminating resin/wood/aluminum/etc. It would be nice to have some machined molds, but I couldn’t justify the cost for my current needs. Special care has to be used to minimize deflection of the mold at high bladder pressures. A well designed metal mold would be less hassle in this area.

The mold was built with a tapered neck at the open end to capture the bladder hardware. A custom machined fitting serves to mate with the perforated metal form that sits under the bladder and to seal the open end of the latex bladder. Sizing of the perforated form and bladder are critical to ensure that the laminate preform just fits in the mold cavity.

#14

I’ll try the ramping of the pressure, that’s the one thing I really haven’t played with yet. how to you manage it, just a tank near the mold and a needle valve or something computerized or just coming by every minute and twisting the regulator?

#15

yeah nothing fancy, just twist the regulator every minute.

#16

Sent from my iPhone using Tapatalk

#17

tried a new tweak on the mold, using an inner channel to scavenge any leaks from the outer seal, I had a problem with it scavenging resin too, so I added another section of o ring between the resin feed and the scavenging channel. so far it’s kept the resin absolutely air free, but I’m still running into issues with the surface being too dry, I’ll experiment with lower bladder pressure next

#18

an update on the process… turns out latex, epdm, basically any flexible membrane is permeable to gases, liquids, etc, just at varying rates. the rate of permeability is also proportional to the thickness and pressure, so the more the membrane is stretched to inflate the composite stack, and the higher the inflation pressure, the faster the fluid permeates - directly into the laminate stack. experiments show even water permeates. so will probably look closely at how much the bladder is allowed to expand and the thickness it ends up being once inflated. I think most of my problems are related to the small OD of the part and the relatively large wall thickness, which leads to a bladder that needs to be very small to start and stretch a lot to compress the laminate :frowning: