Computation fluid dynamics (CFD) modeling provides the ability to accurately model air flow and heat transfer within a building, enabling heating and cooling systems to be optimized in the form of a software prototype without the expense and time involved in making changes to the actual building.

The greatest challenge in CFD is often modeling the diffusers because they have a major effect on the airflow in the building yet they are difficult to model with conventional CFD software. The difficulty with modeling diffusers arises from the fact that they are small compared to the size of a building and have very high flow velocities. To model detailed diffuser geometry would require millions of grid cells which would require large amounts of computer capacity. On the other hand, to model the diffuser at the same grid density as the room would ignore details that could introduce errors into the numerical simulation.

What needs to be known for an accurate CFD simulation is generally the boundary conditions consisting of the velocity and temperature cross-sectional profile at the diffuser outlet. Manufacturers of diffusers typically perform tests on their products, installing the diffuser in a room and measuring airflow in the room.

Until recently, the most accurate way for CFD users to model diffusers has been to build a model that duplicates the tests performed by the diffuser manufacturer, then adjust the diffuser boundary conditions until the airflow in the room matches the test results. There are several weaknesses to this approach. One is that a considerable amount of time is required to create the room model and the other is that the boundary conditions determined by this method are only completely accurate for rooms that match the room used in the diffuser manufacturer’s test.

From: “Diffuser Models For Airflow Simulation”