Tungsten fractures through a specific sequence, and the sequence matters more than any individual step.
Hussein et al. trace the cascade in molecular dynamics: first, mobile dislocations exhaust themselves (dislocation starvation). Then, twins nucleate to accommodate further deformation. Then, the twin boundaries propagate until they encounter surface asperities — roughness features at the crystal's exterior. The boundaries pin. Disconnections — step-like defects on the boundary — pile up at the pinned sites. The pile-up concentrates stress beyond the local fracture threshold. The crack nucleates.
Note (2026-11-13; read against the paper's full text): this is a molecular-dynamics study of single-crystal tungsten pillars (a 300 K example at a strain rate of 10^7 per second), not a general account of how tungsten fractures. The paper's sequence is as the essay says, but the essay's 'remove any step and the cascade breaks' is not shown step by step: what the paper tests is temperature, where at 1500 K and above the pillars deform ductilely (below 1000 K they fracture; the paper puts the transition in its model between 1000 and 1500 K), which it attributes to smoother surfaces and more mobile boundaries; it suggests that polishing the surface roughness could lower the transition temperature in its model, and that large surface obstacles could cause brittle fracture even at high temperature. It describes the crack as forming at low macroscopic stress, about 1.5 GPa against a bulk flow stress of 4 GPa, and I did not find the essay's 'local fracture threshold' in it.
Remove any step and the cascade breaks. Without starvation, dislocations handle the strain. Without twins, no boundaries form to pin. Without surface roughness, boundaries propagate freely. Without disconnection pile-up, stress stays distributed. The fracture mechanism is not a cause — it is a narrative, each chapter enabling the next.
The role of surface roughness is particularly counterintuitive. Surface asperities are typically treated as minor features — cosmetic imperfections that affect surface properties but not bulk failure. Here, they are the geometric trigger that converts an internal microstructural event into catastrophic fracture. The surface pins the interior. An external feature determines when an internal process becomes fatal.
This is fracture as plot, not as event. The brittleness of tungsten at low temperatures is not one thing going wrong but a sequence of things going right — each deformation mechanism activating in order, each one setting up the conditions for the next, until the final step (the pile-up at the surface pin) crosses the threshold. The material doesn't fail because it's weak. It fails because its responses are too well-organized.