fibre reinforcement
The oldest trick for making a brittle material tough is to lace it with fibres, and it goes back to straw in mud brick. Do the same in a ceramic, embedding strong ceramic fibres such as silicon carbide or carbon in a ceramic matrix, and something remarkable happens: the composite stops behaving like glass, which shatters, and starts behaving like wood, which frays and splinters but hangs together. Fibre reinforcement is the route to ceramics that fail gracefully instead of catastrophically, and it is what allows ceramics into places where sudden failure would be unthinkable.
The secret is counter-intuitive: the bond between fibre and matrix must be weak, not strong. Each fibre is given a thin coating, often carbon or boron nitride, that makes a slippery, weak interface. When a crack runs through the brittle matrix and reaches a fibre, it does not snap the fibre; instead the weak interface debonds and the crack is deflected to run along the fibre, which stays intact and bridges the gap. As the crack opens, the fibre pulls out against friction, and the frictional sliding of thousands of fibres dissipates an enormous amount of energy. The result is not so much a higher K_IC as a hugely greater work of fracture, often hundreds of times that of the plain matrix, so the material tears rather than snaps. Silicon-carbide-fibre composites (SiC/SiC) and carbon-fibre composites (C/SiC) built this way now serve in jet-engine hot sections, rocket nozzles, and aircraft brakes.
Fibre reinforcement matters because it is the only route to ceramics that are genuinely damage-tolerant, tough enough to trust in flight-critical hot structures where a monolithic ceramic would be too risky. But the honesty is sharp and specific. A strong fibre-matrix bond ruins everything: if the fibre is well bonded to the matrix, the crack runs straight through both and the composite is as brittle as the monolith, so the whole benefit hangs on keeping that interface deliberately weak. And the coating that makes the interface weak is vulnerable; at high temperature in air the carbon or boron-nitride coating can oxidise away, the interface then strengthens, and the composite embrittles in service, which is one of the central challenges in using these materials hot.
A SiC/SiC turbine shroud pulled in tension does not snap at first crack. The matrix cracks, but the coated fibres bridge and pull out, so the stress-strain curve bends over and keeps carrying load, showing frayed, wood-like fracture instead of a clean glassy break, exactly the graceful failure a jet engine needs.
A deliberately weak fibre-matrix interface lets fibres debond and pull out, so the composite frays like wood instead of shattering.
Counter-intuitively, the fibre-matrix interface must be weak. If fibre and matrix bond too strongly, cracks run straight through and the composite is as brittle as the monolith; the toughness comes from debonding and pull-out, not from strong fibres alone. Oxidation of the coating can re-embrittle a hot composite.