CFD Benchmark Tests for Indoor Environmental Problems: Part 1 Isothermal/Non-Isothermal Flow in 2D and 3D Room Model

Abstract

Commercially available Computational Fluid Dynamics (CFD) software have been applied in indoor environmental design in recent years, but the prediction accuracy depends on an understanding of fluid dynamics fundamentals, in setting appropriate boundary and numerical conditions. This study aims to provide practical modelling information related to prediction accuracy and problematic areas in CFD applications in air conditioning and ventilation, through a series of benchmark tests and reported the results. Six commercial CFD codes were evaluated while two benchmark test cases were performed on isothermal/non-isothermal flow in 2D and 3D room models. The influence of mesh design, and turbulence models showed that using a standard k-ε model on a coarse mesh could provide sufficiently accurate results for practical purposes, by reducing the relaxation coefficient. Evaluation using different CFD programs on a non-isothermal room airflow showed different performances in predicting temperature distributions. The OpenFOAM code showed the closest matching results between three codes tests.