Predictions Of Room Air Distribution: Solved And Unsolved Problems
|Forfattere||Qingyan “Yan” Chen|
|Institusjon||Vincent P. Reilly Professor of Mechanical Engineering Purdue University, West Lafayette, IN 47907, USA|
|Redaktør||Vojislav Novakovic, Sten Olaf Hanssen, Hans Martin Mathisen|
AbstraktPredictions of room air distribution by numerical methods have played an important role in designing indoor air quality and thermal comfort. The numerical methods include computational fluid dynamics (CFD), multizone models, and zonal models. After being developed for more than 30 years, the numerical methods have been used to solve many important problems in room air distribution. The solved problems include the prediction of air velocity, air temperature, species concentrations, air pressure, etc. for rooms with relatively simple geometry and thermo-fluid boundary conditions. However, the numerical methods still have many uncertainties. For example, zonal models need to handle special zones in order to obtain correct results, but it is not a trivial job to develop a suitable submodel for a special zone. Multizone models do not solve the momentum equation so it may give unrealistic results for rooms in which the air movement momentum cannot be neglected. CFD is the most accurate method, but there is no universal turbulence model for room air distribution. The CFD method can be too demanding in computing time and capacity for rooms with complex situations. To reduce computing time and to use coarse grids in CFD simulations are very challenging, if the accuracy should be maintained. CFD cannot be fully trusted in describing a complex air diffuser; in considering moving objects; and in calculating particle dispersion, deposition and re-suspension in room air. Although the numerical methods still need to be further developed, they are already very useful in engineering applications.