Reference Artefacts and Benchmark Geometries for Additive-Manufacturing Qualification

Additive-manufacturing test artefacts compress many geometric features into one build, but a dense collection of cones, holes, walls and overhangs is not automatically a qualification method. Each feature needs a declared measurand, a measurement route, an uncertainty compatible with the decision and a reason for occupying a particular position and orientation. Otherwise the artefact is a demonstration piece whose attractive complexity can conceal ambiguous evidence. This review develops a fitness-for-purpose architecture for reference artefacts and benchmark geometries used in additive-manufacturing qualification. It distinguishes a digital nominal definition, a manufactured test artefact, a calibrated reference object and a routine witness feature. It links feature selection to geometric capability, spatial effects, surface response and minimum-feature claims; addresses file identity, specimen state, measurement access, uncertainty and repeated builds; and defines boundaries between characterisation, calibration, monitoring and acceptance. ISO/ASTM 52902 and the NIST additive-manufacturing test artefact provide established foundations, while the wider literature shows that no single geometry serves every process and decision. The central conclusion is that an artefact is useful only when every retained feature closes a traceable question. Qualification belongs to the coupled design-build-measure-decision chain, not to the geometry alone.