Ultrasonic Consolidation (UC) is a solid‑state additive manufacturing technology that enables layer‑by‑layer bonding of metallic foils at temperatures substantially below their melting point. This low-temperature processing window preserves the bulk material properties while allowing multimaterial integration and embedding of sensitive components. Despite its significant industrial potential, the underlying physical and mechanical mechanisms of UC remain only partially understood, and process optimisation is still largely guided by empirical approaches rather than theoretical models. In this context, numerical modelling has become a key tool for exploring the coupled phenomena governing UC, including frictional heating, plastic deformation, thermomechanical softening, and microstructural evolution. Nevertheless, existing modelling efforts remain fragmented in terms of their physical scope, spatial and temporal scales, and validation methodologies. This review presents a critical assessment of the numerical frameworks developed to date, categorising them into thermomechanical, crystal plasticity, energy‑based, and system‑level approaches. Particular attention is given to the representation of frictional behaviour, constitutive material response, heat generation, and softening mechanisms. Furthermore, as a major limitation, model validation remains a challenge, as most studies rely primarily on macroscopic observations, such as thermal evolution or geometric variation, whereas advanced characterisation techniques, including scanning electron microscopy, in situ strain mapping, and high-resolution microstructural analysis, are comparatively underexploited. By clarifying methodological assumptions, pinpointing mechanistic ambiguities, and outlining unresolved knowledge gaps, this review proposes a structured framework to guide next-generation UC models toward multiscale, multiphysics, and experimentally anchored predictive tools. Such integration is essential for computationally efficient, scalable simulations and for supporting digital twin and AI-driven industrial implementation of UC.
Gomes et al. (Sat,) studied this question.