Hall–Petch relationship

For many polycrystalline metals over a suitable range of grain sizes, yield stress σᵧ is described by σᵧ = σ₀ + k d⁻¹ᐟ². Here d is a characteristic grain size, σ₀ is a material-dependent friction-stress term, and k is the Hall–Petch slope. The equation is empirical: its constants depend on the material and its condition. It describes a trend, not a universal formula with one set of constants for every material or process.

A polycrystalline solid consists of differently oriented grains separated by grain boundaries. These boundaries obstruct the motion of dislocations, which are defects involved in plastic deformation. Reducing grain size increases the density of boundaries and, in many ordinary cases, raises the stress needed to initiate yielding. Models of how boundaries and dislocations produce the observed scaling have been proposed; the relation is not itself a single detailed microscopic mechanism.

The classical relation works well for many metals within particular microstructural and grain-size ranges, but it should not be extrapolated indefinitely. At nanometre-scale grain sizes, experiments and interpretations are sensitive to material, processing and measurement; some materials exhibit a transition to different deformation processes, and reports of inverse Hall–Petch softening require careful evaluation. The relation for initially undeformed metals also cannot automatically be applied unchanged to heavily deformed metals, whose microstructures contain multiple kinds of boundaries and dislocations.

The relationship is named for E. O. Hall and N. J. Petch, whose early-1950s work connected grain size with yielding or fracture behaviour in polycrystalline metals. Hall's 1951 paper discussed the variation of the lower yield point of mild steel with grain size. Later research examined the relationship across materials, size ranges and microstructures, including where its simple form needs qualification.