Improving building natural ventilation simulation accuracy: calibration of standard k − ε closure coefficients using design of experiment

Abstract

The objective of this study is to improve the accuracy of the standard k − ε (SKE) turbulence model for natural ventilation simulation, a critical component in the analysis and design of energy-efficient buildings. The methodology involves generating 17 design points for the four closure coefficients (C ɛ 1,C ɛ 2,C µ ,σ k ) for the SKE model using 2k full-factorial design of experiment (DOE) with a centre point. A sheltered single storey-building with windward and leeward openings is modelled and simulated using computational Fluid Dynamics (CFD) to obtain the Root Mean Square Error (RMSE) and Factor of two observation (FAC2) for each design point. Response models for RMSE and FAC2 are then generated, and optimization is employed to maximize FAC2, minimize RMSE, and the combination of both. The results are validated and compared with experimental and simulation data reported in existing literature. The findings indicate that the indoor velocity profiles with closure coefficients obtained by maximizing FAC2 are better aligned with the actual experimental results. Compared with the existing calibration method, the proposed calibration method can yield superior results with fewer sampling points, highlighting the efficiency of the proposed approach. The calibration framework developed in this study can be extended to other simulation scenarios, contributing to the robust and efficient modelling of natural ventilation systems in various contexts. This research presents a novel calibration method for the SKE model in natural ventilation simulation, providing a more accurate and efficient approach compared to existing methods. The results have important implications for researchers and practitioners working in the field of building energy simulation and sustainability.