A New and Interesting Composite

While working on an undergraduate project at my university, I got the opportunity to make an HPR fuselage with an interesting carbon fiber composite. Unlike most carbon fiber composites which use thermosets like epoxy as matrix (the stuff that isn't the fibers) this one used PEEK thermoplastic. I very quickly fell in love with the stuff. It manages to be as light as most professional carbon fiber composites with a density of around 1.44 g/cc (OpenRocket lists typical carbon fiber density to be 1.78 g/cc). It is also might be surprisingly strong. It was supposed to withstand 4000 N in compression but simulations indicated that it could go up to almost ten times that number. Overall, it manages to be superior to aluminum in most ways (again, this is according to simulation). To achieve this, we stacked sixteen layers of the material in four different directions (0, 90, +/- 45). This can be seen below:



Another great thing about it is that it rolls off the machine ready to go, only needing for the rough ends to be cut. The manufacturing process can sometimes take less than half a day. You can see the robot that is used to make it as well as a finished section below:





It also has very good dimensional accuracy (mostly due to the fact that it was done by a robot). This means that couplers are a piece of cake. We made one for the rocket's payload bay and its fit was pretty much perfect:



Are there any downsides to this material? Of course! The fact that it is made with a thermoplastic means that it is weaker than other composites at higher temperatures (it melts at around 343 C which is still higher than most thermoplastics).

The biggest problem, however, was the cost. For us, it was about 950 CAD per meter, which puts it out of the range of most hobbyists. This is, of course, assuming that you can get access to one of the robots that does this. It is possible that the technology will become more accessible in the future, at which point I hope that its use will become more widespread in rocketry.

Overall, working with this material was a worthwhile experience and I hope that I've imparted some of my love and interest for composite materials science to the reader.

Very interesting! Is the only change the replacement of epoxy to PEEK? I'd love to see some spec sheets on this.

Alex

Aksrockets said:

Very interesting! Is the only change the replacement of epoxy to PEEK? I'd love to see some spec sheets on this.

Click to expand...

In terms of the composition, yes. However, the machine doesn't lay down a single sheet. Rather, it uses unidirectional tape (0.25'' wide by 0.005'' thick) which it presses against the mandrel.

Cool project - fun when you have fancy tools like that to work with!

I use PEEK at work - its just about the most expensive plastic I've seen in common use (a 6" sq 1/4think piece is ~$100USD at mcmaster); you noted the high cost as well. Its designed for chemical resistance, and is quite good when compared to fluoropolymers, but with far better mechanical properties (and 2x the price than most) - I've machined multiple sample cells from PEEK, as it can handle many of the nasty solvents I use.

Are there some special design requirements that called for chemical resistance? Just curious why one would choose PEEK for composites.

How is its adhesion to the fiber reinforcement? That is usually the limitation a given matrix has on the strength of the composite.

PEEK might be a tad lighter than epoxy, but the glass transition temperature is only ~150 degC so I would expect the usual composite heating problems at high Mach. The elongation is an order of magnitude greater than epoxy, hard to figure if that's good or bad in the composite. My guess is not very good for tube stiffness and fin flutter.

How well does it take glue (like epoxy)? Will the glue joints be a weakness?

AndyC said:

Are there some special design requirements that called for chemical resistance? Just curious why one would choose PEEK for composites.

Click to expand...

We actually only had 2 choices of material for use with that machine. There was Nylon and PEEK. Between the two, PEEK was definitely the better choice as the manufacturer specifications listed it as being stronger and more temperature resistant. There is also an optional attachment for the machine which would have allowed us to wrap a tube with a semi-cured composite and then finish the cure in an autoclave, but the mandrels we had were too long to fit.

WoShuGui said:

PEEK might be a tad lighter than epoxy, but the glass transition temperature is only ~150 degC so I would expect the usual composite heating problems at high Mach. The elongation is an order of magnitude greater than epoxy, hard to figure if that's good or bad in the composite. My guess is not very good for tube stiffness and fin flutter.

Click to expand...

From what we can tell so far, the final material has excellent stiffness and is very strong. The Elastic Modulus we measured was about 75 GPA, but we have yet to determine the fracture strength of a 10 cm sample. We weren't too worried about heating, as the rocket we were building wasn't supposed to travel at high Mach.

CarVac said:

How is its adhesion to the fiber reinforcement? That is usually the limitation a given matrix has on the strength of the composite.

Click to expand...

To our knowledge, it must be pretty good. Unfortunately, that information wasn't in the package provided by the manufacturer of the tape.

CZ Brat said:

How well does it take glue (like epoxy)? Will the glue joints be a weakness?

Click to expand...

The payload section you see in the original post was done with epoxy and it has held up. However, we were bonding over a large area in that case so that's probably why to bond is so strong. For bonds with smaller surface areas, we were thinking of reinforcing with fasteners like U-type screws or rivets.

Another interesting possibility would be using heat to weld pieces of this composite together. The tube was created by softening the material with a jet of super-heated gas and then pressing it against the mandrel, so it might be possible that if it were reheated, it may bond to other pieces. However, this has to be done without deforming the original parts. Anyways, that's just me speculating. It would be a fun experiment to try.

PEEK is also radiolucent. It's used extensively in medical implants because of this, as well as it's not rejected by the body.

While it's typically machined to shape, I have done some injected molded parts with this. The process window is pretty tight, though.

Aksrockets said:

Very interesting! Is the only change the replacement of epoxy to PEEK? I'd love to see some spec sheets on this.

Alex

Click to expand...

No. This is not a hobby process. It used a thermoset ribbon composite and is machine wound over a mandrel and heated. If you have a few tens of thousands you can set up a machine to make it.

Bob

bobkrech said:

No. This is not a hobby process. It used a thermoset ribbon composite and is machine wound over a mandrel and heated. If you have a few tens of thousands you can set up a machine to make it

Click to expand...

This is certainly not a hobby process but I think there may be companies out and about that can provide these kinds of services. For us, mandrels and other equipment were 2000 CAD and material and labor was 1400 CAD, but this was a pretty large rocket. For smaller parts (and if you're making many of them) it might be worth it to do on a hobby scale. I know that there are a few companies here in Canada that could have done this for us. As for the documentation, I'd be happy to send out whatever data I have available (within reasonable limits, of course!). I would like to help others learn about this or even let others contribute.

The company that made the winding machine in your photograph is Automated Dynamics of Schenectady, NY. They've been around for 30 years and have sold 75 machines in 17 countries so the machine is not very common and most likely quite expensive, well over the hobbyist range. (Hobbyists complain about $125 per gallon high temperature laminating epoxies.....).

The company does do prototype and production jobs so for an individual that would be the best approach, however I doubt that it would be inexpensive. You stated that your tube cost $950CAD per meter and you had access to the machine. That ~$1050USD per meter, and I doubt any hobbyist has access to one of these machines so the price is most likely significantly higher if you have to employ their services.

A hobbyist can make a functionally equivalent one-off tube using unidirectional tape and laminating epoxy in a hand-layup for less than 20% of the ~$1000 cost you quoted since their time is free. I know as I've done it, and this is the way the majority of the aerospace industry does it today for one-off and limited production runs. For standard production tube runs, filament winding machines with wet layup are the industry standard and is the least expensive way to go.

PEEK is a wonderful thermoset resin but it is more expensive than high strength high temperature laminating phenolic epoxy resin or thermoset phenolic resin. When you begin your industrial career you will find that cost is the driver for all products and unless you need a unique quality of a PEEK carbon composite, you will choose a more conventional carbon phenolic composite because it does the job and costs less to manufacturer.

Bob

That is, of course, a good point. However, most of the cost came from startup costs for buying the mandrels. Without that, it comes out to less than 300 dollars a meter. As well, our relative inexperience with the equipment meant there was some waste in terms of time and material. I don't think the technology is yet at the point where a maker workshop can have one of these machines, but it might be possible for an already established company to manufacture airframes for hobbyists sometime in the future. Either way, this is still an interesting possibility.

A serious competitor to the technique is pultrusion. https://www.plasan-na.com/capabilities/ shows an infinitely long pultruded wing airfoil.

what kind of layering can you get with that technique? From here, it looks the layers might go in several different directions. If that's the case, this would be a very good method for making tubes.

Another disadvantage of the PEEK composite (which I am currently experiencing): loose fibers left over from the process can get embedded in your skin and it itches like crazy!

Neilw said:

what kind of layering can you get with that technique? From here, it looks the layers might go in several different directions. If that's the case, this would be a very good method for making tubes.

Click to expand...

Not an expert there. The process is used to make inexpensive (relative term) deployable wings for dropped devices.

Pultrusion is the composite analog for metal pipe extrusion. It is a continuous process and uses reasonably priced materials and makes an economic pathway for mass production of relatively simple composite shapes. The a fair number of companies in the business. Google pultrusion.

Bob

bobkrech said:

The company that made the winding machine in your photograph is Automated Dynamics of Schenectady, NY.

Click to expand...

Today I learned that I have driven by this place A LOT over the past 15 years...