Particle- and Fiber-Reinforced Composites

Two ways to reinforce: scattered specks or long threads

The last guide handed us the core bargain of a composite: take a soft, tough, cheap matrix that binds everything, transfers load, and protects, then bury inside it a stiff, strong reinforcement that actually carries the load — and the pair beats a weakness that neither could beat alone, like straw stiffening mud or steel rebar bracing concrete. The matrix and the reinforcement each cover for the other. Now we ask the next practical question: what shape is that reinforcement? Because the geometry of the reinforcement, more than its chemistry, decides how the composite behaves.

There are really only two answers, and they define the two great families. In a particle-reinforced composite the reinforcement is roughly ball-shaped — equiaxed, about as wide as it is tall, like gravel or grains of sand. In a fiber-reinforced composite the reinforcement is long and thin, a thread with a huge aspect ratio (length divided by diameter can reach thousands). That difference is not cosmetic. A long thread is spectacularly good at one job: carrying tension along its length, the way a rope holds a load while a pile of sand cannot. So fibers give you the biggest strength and stiffness gains, and they give them mostly in one direction. Particles give you more modest, more even improvement in every direction.

One more useful axis cuts across both families: how big the reinforcement is, and how it does its job. Sometimes the reinforcement is large enough that both phases genuinely share the load — you can watch them work together with the naked eye, as in concrete. Sometimes the reinforcement is a cloud of specks so tiny you need an electron microscope to see them, and their job is not to carry load at all but to jam up the tiny defects that let metals deform. Keep both pictures in mind — load-sharing versus defect-jamming — because they explain why two composites that look similar can be strengthened for completely different reasons.

Particle-reinforced: concrete, carbide teeth, and invisible specks

Start with the large-particle kind, where you can see the reinforcement. The classic is concrete: a matrix of cement paste glueing together sand and gravel. The stones are cheap, hard, and stiff; the paste alone would be weak and crack-prone, but packed around a skeleton of stone it becomes a structural material that has held up cities for two thousand years. Another is the cermet or cemented carbide in a drill bit or cutting tool — extremely hard tungsten-carbide particles cemented together by a little tough, ductile cobalt metal. The carbide gives the cutting hardness; the cobalt keeps the whole thing from shattering. In both cases the particles restrain the softer matrix around them, both phases carry load together, and the stiffness lands somewhere between the two components — a first taste of the rule of mixtures that guide three makes exact.

Now the sneaky cousin: dispersion-strengthened composites. Here the particles are tiny — ten to a hundred nanometres, thousands of times smaller than gravel — and there are not many of them, only a few percent by volume. They carry almost no load. Instead they sit in the path of gliding dislocations, the little rucks in the crystal whose motion lets a metal bend, and they pin them in place. If you remember the paperclip that gets harder to bend once its tangles jam, this is the same idea deliberately engineered in. It looks a lot like precipitation strengthening from the metals rung, with one crucial difference: the dispersed particles are a chemically inert, insoluble second phase — fine oxide grains like aluminium oxide in aluminium (SAP) or thoria in nickel (TD-nickel) — so they do not dissolve back into the metal or coarsen away when it gets hot.

Fibers: length, orientation, and how much

Fiber composites are where the biggest performance lives, and their behaviour turns on three knobs. The first is volume fraction — simply how much of the material is fiber. More fiber means more of the stiff, strong phase carrying load, so stiffness and strength climb roughly in proportion. A fiber-reinforced composite with 60 percent glass or carbon fiber is a very different animal from one with 20 percent. There is a ceiling: pack too many fibers and there is not enough matrix left to wet them and glue them together, so real parts top out around 60 to 70 percent by volume.

The second knob is orientation, and it is where composites earn their reputation for being tricky and wonderful. Line all the fibers up one way and the material becomes fiercely anisotropic — magnificent along the fibers, feeble across them. Put numbers on it: glass fiber has a stiffness near 72 GPa and epoxy near 3 GPa, so a 60-percent aligned lay-up stretched along the fibers has a modulus of about 0.6 times 72 plus 0.4 times 3, which is roughly 44 GPa. Pull the same material sideways, across the fibers, and the soft epoxy now carries the strain in series, dropping the modulus to only about 7 GPa. Same material, a sixfold difference by direction. Scatter the fibers randomly instead and you trade that peak away for a modest, even stiffness in the plane — the deliberate design choice we untangle in guide five. This is exactly why fiber orientation is a design variable, not an accident.

The third knob is length, and it hides the subtlest idea in the whole subject: the critical fiber length. A fiber cannot grab load from thin air — the matrix has to shear the load into it through the fiber's side surface, so the stress inside a fiber builds up from each end and only reaches full value somewhere in the middle. If the fiber is too short, the stress never climbs to the fiber's breaking strength before you run out of fiber, and instead of snapping and doing its full job the fiber simply slides out of the matrix — it pulls out, wasted. The break-even length is lc = (sigma_f* times d) / (2 times tau_c), where sigma_f* is the fiber's strength, d its diameter, and tau_c the shear strength of the fiber-matrix bond. Only fibers much longer than lc behave as efficient, continuous reinforcement.

LOAD TRANSFER INTO A FIBER   (tension pulls the composite along its length)

  the matrix shears load into the fiber through its curved side,
  so the stress in the fiber builds from each END toward the MIDDLE

  fiber
  stress
    ^
sf* |         __________________         long fiber (l >> lc): plateaus at
    |        /                  \        sf*, the fiber does its full job
    |       /                    \
    |      /                      \
    |     / \                    / \      short fiber (l < lc): never reaches
    |    /   \                  /   \     sf*, so it just PULLS OUT, wasted
    +---+-----+----------------+-----+--> position along the fiber
       end                        end

  critical length:   lc = (sf* x d) / (2 x tau_c)

  carbon fiber:   sf* = 3500 MPa,  d = 0.007 mm,  tau_c = 25 MPa
                  lc = 3500 x 0.007 / (2 x 25) = 0.49 mm   (about 0.5 mm)

  a 15 mm chopped fiber is ~30x lc  ->  behaves essentially 'continuous'

Choosing a matrix, and the three great fibers

Wrap those fibers in different matrices and you get three families. By far the most common is the polymer-matrix composite, where a light polymer (usually a thermoset epoxy) holds the fibers — cheap, easy to mould, but limited to a few hundred degrees. Two nicknames rule this world: GFRP, glass-fiber-reinforced polymer, the fiberglass of boat hulls and surfboards, stiff enough and gloriously cheap; and CFRP, carbon-fiber-reinforced polymer, the black weave of bike frames and airliner wings, far stiffer and lighter. The advantage is best read as specific stiffness — stiffness per unit weight — because that is what matters when you must move the structure, and it is where fibers demolish plain metals.

Swap the polymer for a metal and you get a metal-matrix composite — say silicon-carbide particles or fibers in aluminium. It costs and weighs more but tolerates far higher temperatures and adds stiffness where a bare alloy would be too soft. In every one of these families there is a hidden third party doing quiet, decisive work: the fiber-matrix interface, the thin skin of contact where load actually crosses from matrix to fiber. Bond it too weakly and the fiber pulls out under light load; bond it too strongly and a crack runs straight through fiber and matrix alike, making the whole thing brittle. Tuning that handshake is a craft in itself — the subject of guide four.

1. Pin down the service temperature first. Below a couple of hundred degrees a light, cheap polymer matrix wins; hotter and you must climb to a metal, then a ceramic matrix.
2. Decide what you are buying. If you need light stiffness and strength, carry load with fibers (polymer or metal matrix); if you need a brittle material not to shatter, buy toughness (ceramic matrix).
3. Pick the fiber by the same trade-off: glass for cheap-and-stiff-enough, carbon for maximum specific stiffness and strength, aramid (Kevlar) for toughness and impact at low density.
4. Only now set length, volume fraction, and orientation — make fibers far longer than the critical length, pack as much as the matrix can wet, and aim them along the loads you actually expect.

Stacks, sandwiches, and what nature knew first

A single sheet of aligned fibers is strong one way and weak across — useless on its own for a real part that gets pushed in many directions. The fix is the laminate: stack thin sheets (plies) at different angles and bond them, exactly the trick in ordinary plywood, where each wood veneer is glued with its grain crossing the last. A common lay-up alternates 0, 90, and plus-or-minus-45 degrees so the stack is stiff in every in-plane direction — you deliberately spread the anisotropy around instead of fighting it. Designing that stack is the heart of guide five.

There is a second structural trick worth meeting now: the sandwich panel. Glue two thin, stiff faces (often a laminate) onto a thick, light core of foam or honeycomb, and you get a panel that is astonishingly stiff for its weight. It works exactly like an I-beam. In bending, the faces take the tension and compression while the lightweight core simply holds them apart and carries the shear between them. Pushing the strong faces far apart with almost no weight is what buys the huge specific stiffness — it is why aircraft floors, skis, and doors are built this way.

None of this is a modern invention. The best composites were grown, not manufactured. Wood is stiff cellulose fibers set in a softer lignin matrix — a natural fiber composite, strong along the grain and split-prone across it, anisotropy you can feel with an axe. Bone is tough collagen fibers threaded through hard, brittle mineral crystals, each covering the other's weakness so the whole survives both a knock and a bend. Once you see the matrix-plus-reinforcement pattern, you see it everywhere living things bear load — which is the quiet lesson under this entire rung.