=manufacturing =materials =composites
Glass fiber is generally $2 to
$15 / kg.
Carbon fiber is generally $20 to $100 / kg.
Aircraft
list prices
aren't always accurate, but aircraft are, in general, over $1000/kg dry
mass.
Materials are generally a small part of overall costs, even
in
construction, but that doesn't mean they're not important: you can't
build a composite aircraft without composite fibers. New materials can
enable new capabilities, and while manufacturing costs eclipse material
costs, understanding materials is central to designing manufacturing
methods.
Long stuff with a constant
cross-section can be made cheaply with
pultrusion. So,
fiberglass beams are competitive with steel ones for construction.
Fiber winding
is fairly cheap. It's common for cylindrical pressure vessels to be made of
metal and then have fiber wound around them to take the radial force.
Such existing techniques work well for some applications. The cost of
constructing things like aircraft is, I think, an indication of the
incompleteness of the set of tools available for composite manufacturing.
Things are cheap when there are good methods for doing them, and expensive
when there aren't.
A more recent manufacturing
technique is automated tape placement. Here's a
video of a big robot
doing this at Airbus:
That big robot is doing something like making
large structures out of packing tape. Clearly that's why aircraft are so
expensive: if you had US workers making stuff out of packing tape by hand,
the labor costs would be...$2/kg? Hmm.
There's a trend towards usage of
prepreg instead of raw
fiber and liquid resin.
Pultrusion, fiber winding, and tape placement
can all, in general, be done with thermoplastics instead of thermosets. With
fiber winding and tape placement, you just get some thermoplastic prepreg,
and heat it up where where where it's added. Needing to heat things up is a
disadvantage, but making parts with epoxy prepreg usually involves
refrigerating it until use and then baking the resulting part in an oven,
and needing an oven is worse than heating up tape as it's placed. So there's
a lot of interest in thermoplastic prepreg, and it should eventually replace
thermoset prepreg, but industry is conservative.
The main advantage
of thermoset resins is lower viscosity, which isn't relevant for prepreg.
So, the current dominance of thermoset prepreg over thermoplastics is
related to the past dominance of other composite manufacturing techniques.
The most significant of those were probably wet layup,
foam-core vacuum
bagging of woven glass cloth, and
spraying
loose glass fiber.
Composites and aircraft manufacturing are
considered high-tech industries, but composite aircraft construction has
involved a lot of low-paid labor doing hand layup. Not only is that
time-consuming, it's also unpleasant work. Unsaturated polyester has
been the cheapest kind of resin, but the styrene in unsaturated polyester
resin is volatile, making that especially noxious; at least respirators are
used for that more now, but of course they're still uncomfortable. Also,
while glass fiber isn't as bad as asbestos, pieces of it are a serious
inhalation hazard. Carbon fiber is also an inhalation hazard; if anything,
it's worse than glass, but loose carbon fiber is also much less common.
As I said, US workers making stuff out of packing tape by hand would be
perhaps $2/kg, which is actually fairly cheap. So, I spent a bit of time
thinking about possible manual tools for placement of thermoplastic prepreg.
That approach seems better than hand layup.
Considering theoretical
costs rather than current prices, semiaromatic polyamides such as nylon 12T
seem like a decent option for thermoplastic prepreg matrix, so I spent a bit
of time thinking about renewable routes to suitable diamines.
Making things from composites is
generally more expensive than making things from steel. Steel products can't
be made by layering tape or fabric, but steel can be stamped and welded, and
overall, the available techniques for steel are better.
Fibers are
chosen because they're strong, and mechanical strength is resistance to
change, so strong materials tend to resist being changed. Obviously,
resistance to change tends to make manufacturing harder. It's possible to
use fibers but trade strength for easier manipulation: polypropylene is easy
to make into fibers, and the fibers can just be melted together; the
resulting "nonwoven
fabric" is now widely used.
Carbon fiber is expensive largely
because a single layer of very thin fiber is
processed through
a bunch of ovens.
Glass fiber is much cheaper, because you can just
extrude it, but S-glass (which is somewhat stronger) is much more expensive
than E-glass. Glass fiber is extruded through bushings made of rare metals.
The stronger kinds require higher temperatures, which decreases bushing
lifetime. If steel was to be completely replaced by glass fiber, perhaps the
supply of platinum for bushings would become a major problem.
Obviously, it's possible to make polymer fibers, typically
with melt spinning or gel spinning, and melt spinning polymers doesn't
require such high temperatures. But polymer fibers tend to have much lower
compressive strength than glass fiber and carbon fiber, because the polymer
chains are less strongly connected to each other, so they can buckle more
easily; even M5 fiber has lower compressive strength than the cheap kinds of
glass fiber.
Thermoplastics reinforced with
short pieces of glass or carbon fiber are also important, since they can be
processed like (particularly viscous) thermoplastics. Shorter fibers make
strong interaction between the matrix and fiber more important, so the
advantage of polar polymers like polyamides over nonpolar ones like
polyethylene is larger, so 40% glass fiber reinforced nylon 66 is a
relatively common choice where "a thermoplastic but stronger and less
flexible" is what's desired.
But the theme of this post is
manufacturing techniques specific to composites, and materials that are
basically processed the same way as thermoplastics don't fit that. The
general topic of "composites of fiber in a matrix" is far too broad for a
blog post. For example, wood is a fiber composite with cellulose fibers in a
lignin matrix (well, the cellulose, hemicellulose, and lignin are all
covalently bonded) - and of course, many books have been written about wood.
For that matter, even "homogeneous" thermoplastics with good tensile
strength tend to be "molecular composites" to some extent, with crystals
containing aligned polymer chains inside an amorphous matrix of the same
polymer. (True molecular composites, with rigid-rod polymers in a
thermoplastic matrix, tend to be too viscous to process like thermoplastics,
the same way thermoplastics reinforced with long fibers are.)