Verifying the reliability of CFD domain decomposition technique on modelling the airflow field inside a naturally ventilated cattle barn

Abstract

Conventionally, the airflow fields outside and inside the naturally ventilated livestock buildings are modelled simultaneously in one computational domain using CFD (Computational Fluid Dynamics). The presence of surrounding buildings, indoor facilities and animals for large scale cattle barns make the required computational power extremely high and even unfordable to achieve simulation results with reasonable accuracy. The Domain Decomposition Technique (DDT), dividing simulations into two separate steps, is an alternative CFD framework to provide sufficient accuracy with affordable computations at each step. The objective of this study was to verify the reliability of DDT on modelling the airflow fields inside a naturally ventilated cattle barn (NVCB) by employing wind tunnel measurements. The exterior airflow fields around the targeted NVCB, which was opened with varying opening ratios, were first simulated to obtain the airflow boundary conditions at sidewall openings by applying exterior wind conditions at the inlet of the computational domain. The interior airflow of the targeted NVCB, were secondly simulated by applying the achieved airflow boundary conditions at sidewall openings from the first step simulation. The interior airflow fields obtained by DDT were in good agreement with wind tunnel measurements. This indicates that DDT can provide an alternative for CFD application in large-scale NVCB with presence of surrounding buildings, indoor facilities and animals, though these had not been considered in this study.