Manufacturing of CFRP Part 1

[tohck]

Hi all,

I am new to carbon fiber reinforced plastic, and would like to ask what is a suitable process to fabricate a cfrp part of hyperboloid (Link) shape?

I was told that cfrp molding can only handle simple geometry. Is that true?

Best Regards
toh

Engineers like to solve problems. If there are no problems handily available, they will create their own problems.

 

[MikeHalloran]

cfrp molding can only handle simple geometry

Well, nothing more complex than a Boeing 777 or a Blackhawk helicopter.
... and an astounding variety of trim parts for high end motorcycles and automobiles.
... and an astronomical number of cold-air inlet tubes.

That said, it's a good idea to start with fiberglass reinforcement and simple geometries until your skills and your process can get you beyond making scrap. Carbon fiber is expensive stuff to throw away.






Mike Halloran
Pembroke Pines, FL, USA

 

[tohck]

Hi Mike,

Thanks for your reply. I understand that CFRP are commonly used to create the enclosure, cover and casing in automobile and aerospace industry. However, most of them are sheets:

,
instead of enclosed structural parts

.


I am particularly interested in converting a machined part with 3d features into a cfrp molded part. For example, a hollow tube with wings:

Can this be done with any process similar to injection molding?

it's a good idea to start with fiberglass reinforcement and simple geometries until your skills and your process can get you beyond making scrap

Do you mean that most cfrp moldings depend on manual layup of carbon fiber sheets?

Thanks,
Toh

Engineers like to solve problems. If there are no problems handily available, they will create their own problems.

 

[RPstress]

Tubing like that in your picture is quite commonly available, see

http://www.strongwell.com/

and, from

http://www.tygavac.co.uk/

, Masterbar 300 and 400 and Masterbar composite tube fittings (Masterbar 300 or 400 CTF). Chopping up the tube and gluing it together with flat laminates (and the tube fittings) is one option.

While more-or-less isotropic short-fiber composite material with low mechanical properties (with fibers a couple of inches long its undamaged strength can approach 6000-series Al; about 40 ksi is possible) can be injection molded (and if very low mechanical properties are adequate then it can be additively manufactured, see the Strati—

http://www.popularmechanics.com/cars/a16726/local-motors-strati-roadster-test-drive/

).

However, to get good properties continuous fiber reinforcement is needed which means laminates (or maybe 3D woven if you're ambitious enough and rich enough), and for best results altered design from isotropic. Attaching the wings by their ends at right angles isn't very strong and making them into one wing bonded with a large shear area on one of the faces of the tapered tube would be better (or split the tube in two and glue the wing between the two tubes, etc.). (In a pinch you could use angles either side of the wing to make the joint as you show but that's still fairly weak for most loadings.)

 

[berkshire]

For the most part manufacturing with fiber reinforced plastic FRP is an additive manufacturing process, requiring that the layers of reinforcing be added in the directions that make the finished product most suitable for its intended use. The one thing the material does not like is sharp corners and abrupt changes of direction. Your hollow tube with wings is not a good candidate for this type of construction, although it could be done with a 3 piece mold.
A piece like that is better from die cast metal or 3D printing using laser sintering. Your hyperboloid shape can be made using a drape able fabric that can conform to a 3D shape.
B.E.


You are judged not by what you know, but by what you can do.

 

[MikeHalloran]

The particular part you have illustrated, a tapered rectangular tube having internal sharp corners, with external wings, would be a bitch to machine from billet in any material, because of the sharp corners. It would be somewhat less difficult to injection mold, though the sharp corners don't help the molded part's performance. As noted, injection molding and similar processes are not super well adapted to use with long fibers as reinforcement, and short fibers, even of a strong material, don't make a particularly strong composite.

Note also that even if you could mold the proposed part with long fibers, you may not be happy with the effect of drilling fastener holes in the completed part, both cosmetic and structural. If you're going to the trouble of molding a part, it makes sense to mold the part to its exact as-used configuration, without secondary operations.

There are a couple of processes that might deserve consideration, depending on what you are actually making.

- In Resin Transfer Molding, the reinforcement fibers/fabric are laid up dry, the cavity around them is closed, and a catalyzed resin is introduced at one area while air is removed elsewhere. This process has displaced hand layup in boatbuilding, because the closed cavity captures what would otherwise be VOC emissions and nasty odors, and the usual associated vacuum bag squeezes everything together, resulting in a product with less resin and more fiber, and if it's done correctly, no 'dry spots'.

- In products that would ordinarily be amenable to fabric/ roving/ tape hand layup, it can make sense to buy 'prepreg' intermediates from third party suppliers. Instead of buying bulk resin and reinforcement, you buy a roll of reinforcement that's already impregnated with resin and partially catalyzed. You just unroll the stuff from the supplied roll, peel off the separating film, and wrap the prepreg around a core for whatever you're making, then bake it to make a finished product. It's sticky enough to adhere to most things, but doesn't run off on the floor, so housekeeping costs are reduced, and you don't have to worry about your own people not mixing the resin correctly every time.

Without knowing more about how your proposed part is stressed, and exactly how it interfaces with any mating parts, we here probably can't be of much more help to you. Sorry.



Mike Halloran
Pembroke Pines, FL, USA

 

[tohck]

Thanks for your advise RPstress, berkshire and Mike.

More questions from me:

isotropic short-fiber composite material with low mechanical properties (with fibers a couple of inches long its undamaged strength can approach 6000-series Al; about 40 ksi is possible) can be injection molded

In the case of injection molding of short-fiber composite material, can the direction of fiber aligned in the same direction? or will it be always random (and give the strength of 6000-series AL)?

a tapered rectangular tube having internal sharp corners

It is my bad on this one, there will be internal radius inside the tube.

Without knowing more about how your proposed part is stressed, and exactly how it interfaces with any mating parts, we here probably can't be of much more help to you.

I am from the semiconductor equipment industry where the components require high rigidity and experience high acceleration. We are currently looking to convert a small machined parts full of features into carbon fiber parts for weight reduction purpose. It is hard for me to visualize what kind of features on a machined part that cannot be achieved via cfrp. For the resin transfer molding and prepreg material you mentioned, any design guidelines to suit such processes?

Besides that, is it possible to add metal parts/inserts during the layup/molding process (such that the metal part will be bonded to the carbon fiber part)?


Engineers like to solve problems. If there are no problems handily available, they will create their own problems.

 

[RPstress]

Injection molding will tend to align longer fibers in the direction of the injection and the material will be a bit non-isotropic because of that (it should go up a little bit in the direction of alignment and down a bit across it).

Quantifying how much is usually a matter of testing several as-manufactured parts with obvious problems of needing to make conservative assumptions to avoid remaking molds or at least re-machining them too much. Really short short fibers (length just a few times their diameter) will usually not suffer much from this non-isotropy but will also be weaker and less stiff than longer ones. Consulting the supplier of your injection molding equipment and/or material supplier will help avoid bad mistakes but it's never a straightforward foolproof process.

In that respect continuous fiber designs are more predictable as well as being higher performance, although the part-to-part variation will be usually be worse than molded. I assume your parts will have to be produced in large numbers which pushes you towards molding rather than laying up continuous fibers, even with AFP or similar. The people at Victrex (suppliers of PEEK thermoplastic) have a lot of experience and a lot of properties data (

https://www.victrex.com/en/datasheets

) and reckon that if you're making more than 4000 parts a year then injection molding will pay for itself. There are a number of variations where you can make a critical high performance bit using continuous fibers and then overmold short fiber-reinforced material for the bulk of the part.

A continuous fiber part (at 60% fiber volume (maybe 50% in glass)) will usually be lightest, and redesign to take advantage is often a matter of thinking outside the box compared with past experience. Just bearing in mind simply that along the fibers is strong and stiff and matching that to your loads/accelerations can be the key. (And acknowledging that through thickness properties of laminates (especially at radii) are terrible; big increases in thickness are often needed.)

 

[tohck]

Thank you RPstress, Victrex's material looks promising.
Do you have any recommendation to fabricate cfrp with embedded customized metal parts?

Engineers like to solve problems. If there are no problems handily available, they will create their own problems.

 

[MikeHalloran]

the components require high rigidity and experience high acceleration

That puts a possibly different light on the situation.

Your first examples appear to be trim pieces for something,
not severely stressed, and maybe a few tens of cm in major dimension,
so the discussion about processes was oriented toward parts like that.

If your parts are smaller, say approximately the size of sewing machine internal parts,
and you need extreme stiffness,
then plastics may not be an appropriate substitute for metal.

I can conjecture sound commercial reasons for not revealing much more here,
so I suggest you hire a temp with serious plastics skills to help you
figure out what might be feasible in your application.





Mike Halloran
Pembroke Pines, FL, USA

 

[RPstress]

Embedding metal in carbon is possible although we've usually shrunk in things like bushes after making the composite. Because the composite doesn't change shape much with temperature heating it is pointless and very close control of the metal's dimension is needed to shrink it in successfully. Usually titanium alloy (which has low CTE making it even more difficult) or stainless steel is used to avoid corrosion from the carbon (if stainless something like A286 with passable specific strength provided it's compatible with any bolts mating with the part (avoid galling so if the bolts are A286 don't use it for a bush)).

Embedding from cure or molding is always problematic because locating the metal part is tricky even if location pins or similar tooling is used because predicting the spring of the composite is hard (a simple radius will tend to spring in). If the metal needs to mate with another part it's usual to allow for machining it a bit after making the composite (this will be too expensive for high volume so making some parts by trial and error is needed). Other problems are often related to the metal shrinking on cool-down from cure (usually 180° for epoxy or sometimes 120°) or molding if it's PEEK (allow for processing at 400°C) or a similar thermoplastic. Fokker have recently used some PEK which they seem to like better than PEEK (they started out with PPS); note that they like to reinforce the thermoplastic with glass for the leading edge parts they've made for Airbus (

http://www.compositesworld.com/articles/fokker-aerostructures-hoogeveen-the-netherlands

).

A carbon/epoxy laminate will usually have a CTE of about 2 µε/K if it's more-or-less quasi-isotropic. A short fiber composite will be more (note that Victrex's CTE data is somewhat inconsistent but says it will be at least 5 µε/K and the CTE "along flow" and "average" indicates quite a lot of alignment of fibers with flow).

 

[tohck]

Hi RPstress,
Thank you so much for the tips!

Hi Mike,
Sure, will talk to vendor after figuring out the directions with info from you guys!

Engineers like to solve problems. If there are no problems handily available, they will create their own problems.