A detailed plant model in CFD that resolves the microclimate around individual leaves

Abstract

Plant factories require effective ventilation to promote proper plant growth. Computational Fluid Dynamics (CFD) is commonly used to evaluate ventilation strategies in these environments. Traditionally, porous models have been employed to study ventilation in plant factories. However, this study proposes an alternative approach using actual plant geometry, consisting of leaves and stems, which reduces the need for fitting parameters typically used in porous models. The study focuses on basil, with plant geometry based on experimental data to ensure accurate representation. This new plant model accounts for the heat and mass balance of each leaf, assigning individual temperature and humidity values. Radiative heat exchange was also included in the plant model by using the solar ray tracing algorithm to solve for shortwave radiation and the surface to surface radiation model for longwave thermal radiation. Validation was conducted in a small plant factory-like environment (1500 mm × 420 mm x 800 mm) under night-like conditions without shortwave radiation and day-like conditions, with shortwave radiation. Key variables such as transpiration rate and leaf temperature were measured and simulated. The coefficient of variation between measured and simulated transpiration rates ranged from 10 % to 15 % for night-time and 15 % for day-time. Root mean square deviations for leaf temperature were 0.4–0.6 °C at night and 0.5–1.6 °C during the day. A different test case, with air supplied from the bottom instead of the side, demonstrated the new model's capabilities. Overall, the new plant model visualises airflow around and through the canopy, and shows promise for improving ventilation strategies in vertical farming systems.

Plant phenotyping relevance

個葉形状を用いたCFD植物モデルを開発し、葉温度と蒸散速度を実測値と比較検証しているため、植物状態の取得・推定手法が研究の中心である。

Code and data availability

公開状態または取得可能な本文経路を確認できませんでした。

No evidence-backed public reproduction asset is currently recorded.