Convective heat and mass exchange at surfaces of horticultural products: A microscale CFD modelling approach

Knowledge of the convective heat and mass exchange of horticultural products and other plant organs (e.g., leaves) with the environment is of interest for several pre- and postharvest technologies and for plant microclimate analysis. Although these exchange processes often occur at the microscale (e.g. at lenticels, cracks, droplets or stomata, 10(-4) to 10(-2) m), they are still often assessed on a macroscale level. In this study, the convective passive scalar (i.e., heat) transfer from a single spherical product is investigated with computational fluid dynamics (CFD). For this purpose, the product is placed in a virtual wind-tunnel environment, thus not in its natural environment, e.g. on a tree. A sphere is used as reference system; an apple fruit is considered as model system. CFD validation is performed for a sphere, indicating a very good performance of the Reynolds-averaged Navier-Stokes shear stress transport k-omega turbulence model for the drag coefficient. Nusselt number, separation angle and recirculation length over a large Reynolds number range (10-3 x 10(4)). These quantities compare well for sphere and apple. The impact of discretely distributed microscopic scalar sources on the product surface on the convective transfer is analysed by explicitly modelling them. Such sources are representative for moisture loss through lenticels and cracks in the product's cuticle and evaporation of water or pesticide droplets at the surface. As scalar transfer is assumed only to occur locally at these sources, the active surface area is reduced. In this study, it is found that even at rather low surface coverage ratios, relatively large convective flow rates are found, indicating a non-linear dependency on the active surface area. This dependency is a function of coverage ratio, the Reynolds number and the source size. The proposed computational methodology for investigating the effects of these microscopic scalar sources shows a good performance, by which it is a viable alternative for experiments, which are very challenging at such small scales. (C) 2012 Elsevier B.V. All rights reserved.