When you set up ventilation simulations, the boundary conditions must represent the wind field correctly. To achieve this you need appropriate WIND settings on the inflow boundaries and NOZZLE on outflow boundaries, and correct settings for turbulence and wind profiles.
This article concerns simulations where ventilation is caused by wind. The wind is defined on the boundary using the WIND condition. The WIND condition pushes air into the domain at a given velocity (speed and direction).

# Setting boundary conditions

The boundary conditions for each boundary (XLO through ZHI) are set in the Boundary conditions tab of the Scenario Settings:

The following parameters must be given:

• Wind speed, in meters per second. This is defined at the reference height (more on this later).
• Wind direction, as a vector. E.g. moving from XLO to XHI (eastward wind) is <1,0,0>. The magnitude of the vector is not important, only the direction.
• Relative turbulence intensity, the ratio between turbulence and wind speed (see the FLACS User’s Manual for more information). This parameter should normally be between 0 and 0.1 (0.1 is typical for ventilation simulations).
• Turbulence length scale, the typical length scale of turbulent eddies. 1 m is often used for ventilation simulations in open areas. However, the length scale should not be larger than the grid cell size.
• Wind buildup time. The wind speed at the boundary starts at 0 and increases linearly in time to 100% of the given wind speed during the given buildup time. The gradual increase is used for stability reasons and the wind buildup time is typically set between 1 and 10 seconds. Note that using buildup time 0 applies the wind in the entire domain from the start of the simulation, disregarding obstructions. This is unphysical and can lead to numerical instability/crashing.

WIND conditions should be defined on all boundaries where the wind direction is into the domain, as well as on the top and the sides where the wind is parallel to the boundary. The other boundaries should use the NOZZLE definition. For our eastward wind, this means the following:

NOZZLE allows the air to flow out of the domain – without this, the inflow from WIND would build up the pressure in the domain and give unphysical results!

ZLO, the lowest grid line, is typically placed at the ground or sea level. The ground or sea should be represented by a geometry object, not a boundary condition (such as SYMMETRY).

# Setting initial conditions

The WIND boundary condition is also affected by some parameters in the Initial conditions tab:

• Reference height is the height above the ground where the wind speed equals the specified velocity. Meteorological data is typically measured at 10 meters above the ground, so that would be the reference height. Below the reference height, the wind speed will decrease to be zero at ground level. Above the reference height, it will increase above the specified speed:
• Ground roughness is a measure of the typical size of objects which affect the wind. Objects inside the domain are resolved exactly, but objects outside (such as waves, trees, or even buildings) are represented by the roughness parameter when defining the wind profile. This parameter should not be larger than the control volume height close to the surface. Typical values are as follows:
• Pasquill classes are a means of classifying the atmospheric stability. When a Pasquill class is defined, the turbulence length scale and relative turbulence intensity on the WIND boundaries is overwritten by a profile that varies by height. The classes range from A (very unstable) through F (stable). Consult the FLACS User's Manual for more details on Pasquill classes. Classes A-C may make the simulation unstable and are normally not recommended. Typically the Pasquill class is determined by the weather in the following way:
• Canopy height is the level where the wind speed is zero. Typically, this is 0 m (i.e. equal to ground height), but in the case of dense trees, buildings etc. the wind profile can start some distance up from the ground. The value is typically 2/3 of the average height of such obstacles