A metal component is loaded cyclically — stress applied, released, applied again. A crack grows with each cycle. The naive expectation: longer rest between cycles means more time for the material to relax, less accumulated damage, slower crack growth. The observation: the opposite. Fatigue cracks grow faster when allowed more time to relax between loading cycles.

Mäkinen et al. (arXiv:2503.07280) resolve this by showing that crack growth is not a smooth process but a sequence of activated avalanches. The crack tip encounters local barriers — microstructural toughness variations, grain boundaries, inclusions. It cannot push through these barriers smoothly. Instead, it waits, accumulates stress, and then overcomes the barrier in a sudden, coordinated rupture event. The fluctuations in crack velocity are not noise; they are the mechanism.

The rest period matters because it allows the plastic zone ahead of the crack tip to evolve. During rest, the material near the crack tip reorganizes — stress relaxation, dislocation rearrangement, local aging. This reorganization changes the effective barrier landscape that the crack faces on the next loading cycle. Longer rest means more reorganization, which can lower the activation barriers for the next avalanche. The material that rested longer is easier to break — not because it weakened in a bulk sense, but because its local barrier landscape shifted.

The through-claim: the history of a crack is not just the cumulative damage but the sequence of pauses. Two components with identical loading histories but different rest schedules will fail differently because the rest periods modify the barrier landscape independently of the loading. What happens between the loads matters as much as the loads themselves. The rested crack is the faster crack because rest is not recovery — it is reorganization, and reorganization can open doors that the stress alone could not.