Selection of a computational fluid dynamics (CFD) model and its application to greenhouse pad-fan cooling (PFC) systems

Abstract The use of pad-fan cooling (PFC) systems in greenhouses during summer is essential for improving crop production in tropical and subtropical regions. However, computational models have not been evaluated and compared for their ability to accurately simulate the temporal-spatial cooling generated by PFC systems. In addition, PFC systems generally have high electrical costs and provide uneven micro-climatic cooling, which is not conducive to crop growth. In the current study, four computational fluid dynamic (CFD) models (SST, Stke, Rngke, and Rke) combined with the Discrete Ordinate (DO) radiation model were compared for their ability to simulate the PFC cooling effects (wind speed, humidity, and temperature) in a greenhouse. After the best CFD model was identified, the significant factors (fan height or fan power density) which mostly affected the performance of the PFC system were determined. The results showed that, among the four CFD models, the SST model had the best simulation accuracy, i.e., it had both the smallest root mean squared error (RMSE) and the highest R2 when compared with measured indoor wind speeds and temperatures. And, fan power density significantly affected (p