Computational fluid dynamics for active solar drying of tuber crops: a critical narrative review of modelling, validation and optimisation evidence and its transferability

Background Cassava ( Manihot esculenta ), yam ( Dioscorea spp.) and sweet potato ( Ipomoea batatas ) are highly perishable after harvest, and active (forced-convection) solar drying can reduce losses with little external energy. Dryer performance is governed by coupled airflow, heat transfer and moisture transport, which computational fluid dynamics (CFD) can resolve in principle. This critical narrative review appraises how far CFD and complementary models currently support the design, validation and optimisation of active solar dryers for tuber crops. Methods A documented, non-exhaustive search of Scopus, Web of Science Core Collection, ScienceDirect and Google Scholar (to 30 September 2026), supplemented by citation chaining, was used. Evidence was classified into four tiers: direct tuber CFD; tuber-specific non-CFD numerical models; tuber experimental and kinetic studies; and transferable CFD or multiphysics studies of solar dryers for other products. Studies were compared on governing equations, turbulence and radiation treatment, product and moisture representation, numerical verification and validation error. The review is not systematic and does not report a PRISMA flow diagram or pooled estimates. Results Of the 41 sources cited, 14 are primary studies. Only one study applied CFD with a tuber product explicitly represented: a sweet-potato dryer in which the loaded racks were modelled as porous thermal media without moisture transport. In that study, CFD and a lumped-capacitance model both predicted a steady chamber temperature rise of about 40 °C, and fitting the rack heat-transfer coefficients reduced the root-mean-square deviation between the two models to 1.8 °C (lower rack) and 0.9 °C (upper rack); this was a model-to-model comparison, not experimental validation. No CFD study of cassava or yam drying was identified. The only validated cassava heat-and-mass-transfer model was a non-CFD, natural-convection model, with moisture-content differences below 0.2 kg/kg. Guidance on turbulence closure, porous-bed resistance, vapour transport, mesh verification and optimisation was therefore transferred from paddy, fruit, herb and empty-chamber studies, where reported validation errors ranged from 4.3 °C in chamber temperature to below 7.33% in CFD outputs. Conclusion CFD for tuber drying remains largely an inferred rather than demonstrated capability. A decision-driven multiscale framework is proposed that couples loaded-chamber CFD two ways with tuber-specific heat and moisture models, gated by verification, validation and uncertainty analysis; it has not yet been demonstrated on a tuber dryer. Research should prioritise moisture-coupled, experimentally validated tuber CFD and openly shared benchmark data.