Application of coarse-grid computational fluid dynamics on indoor environment modeling: Optimizing the trade-off between grid resolution and simulation accuracy

Abstract

Computational fluid dynamics has been playing an important role in building design and indoor environment study for decades. However, the computing cost of grid-independent computational fluid dynamics prevents its application from real-time simulation in many areas, such as building emergency evacuation and hourly-based energy simulation. Although coarse-grid computational fluid dynamics has the potential of being as fast as or faster than real time and can provide valuable information for decision making, the credibility of coarse-grid computational fluid dynamics results is questioned due to the unknown error scale it brings. This study investigates the grid-induced error, evaluates the potential computing cost saving by using a coarse grid, and provides a guideline for optimizing the trade-off between grid resolution and computing cost. The numerical error caused by coarse grid can be minimized by appropriately adapting the distribution of grid size. Following the guideline of coarse-grid specifications, coarse-grid computational fluid dynamics can provide informative prediction that is comparable to a grid-independent result on building environment modeling. The computing cost of computational fluid dynamics with an optimized coarse grid is usually orders of magnitude less than that with uniform fine grid.