CFD numerical setting combinations for greenhouse natural ventilation across diurnal time points using multi-criteria analysis

Natural ventilation is crucial for regulating the microclimate that influences crop growth in greenhouses, and computational fluid dynamics (CFD) is commonly employed to evaluate greenhouse ventilation performance. In CFD-based greenhouse simulations, selecting and combining CFD numerical settings—such as the turbulence model, inlet wind profile formulation, and pressure–velocity coupling scheme—are critical design considerations. Therefore, this study aims to identify an optimal combination of CFD numerical settings for a CFD-based greenhouse natural ventilation model. Scenario-based simulations were conducted using different combinations of setting, and the CFD results were validated against field monitoring data collected during three time points (08:00, 12:00, and 17:00). Simulated air temperature and relative humidity were compared with measured values, and a multi-criteria decision-making approach was applied to evaluate the accuracy of each scenario. The results show that time-varying environmental conditions and their interaction with numerical settings significantly affect accuracy. For air temperature, the combination of the SST k-ω turbulence model, power law, and a coupled scheme demonstrated the highest accuracy across all time points (mean CVRMSE = 1.55%). In contrast, relative humidity exhibited higher errors than air temperature, primarily due to the additional influence of crop transpiration. And the optimal numerical setting combinations varied depending on the time points. These findings provide practical guidance for choosing CFD numerical settings, which can be applied to the design and operational strategy development of greenhouse natural ventilation systems.