It is well known that the natural ventilation of a building is driven by the combined forces of wind and thermal buoyancy. However, when opening areas are large enough, and except for high volumes, the natural ventilation is mainly driven by the wind forces.
So, for the natural ventilation assessment in warm climates, the coupling between the thermal forces and the wind driven forces is generally assumed as negligible. The pressure field on the building’s envelope generates flows through openings allowing the indoor space refreshment as shown on the following pictures.
Assess the air natural ventilation in urban area with thermal dynamic software
The assessment of the cross-ventilation efficiency is carried by considering the flow-rates through the openings. All these characteristics are fundamentally depending on the external wind pressure at the openings. Macroscopic approach gives air change rate from the pressure field characteristics.
Many numerical thermal softwares used to design natural ventilated buildings consider the values given in the “AIVC applications guide” for low rise buildings (Liddament, 1987) as universal values.
Unfortunately for designers, the pressure value on outside walls of buildings could not be assess with this method in urban places for three reasons:
- The pressure coefficients depends strongly on the buildings shape
- The pressure coefficient on buildings depends strongly on the influence of neighboring buildings
- The referent wind velocity needed to convert the pressure coefficient to pressure cannot be defined easily because of the inhomogeneity of wind in urban places
When buildings have no-standard shapes or when the flow is perturbed by close obstacles, these external wind pressures have to be evaluated either from wind tunnel tests or computational fluid dynamics (CFD) methods such as Urbawind.
Typically CFD computation domains cover an area of almost 500 m x 500 m, allowing taking into account the effects of neighbouring buildings in order to deliver efficient pressure coefficient.