Optimising the airflow performance of ventilators for natural ventilation in buildings.

Ventilation is essential if an adequate environment in buildings is to be provided. Natural ventilation is a sustainable method commonly used to achieve this. Natural ventilation can occur from random flow of air through unintentional (adventitious) openings, commonly referred to as infiltration, or can be facilitated through purpose-provided openings - commonly called controlled natural ventilation. Stringent national and international regulations have led to building structures becoming more and more air-tight so that ventilation is confined to airflow through intentionally provided openings only. These purpose-provided openings typically consist of one or a combination of basic components such as louvers, insect-screens or noise baffles. An extensive literature review indicated that the airflow performance of combinations of such components has neither been thoroughly determined nor fully understood, making it difficult to accurately predict their building performance. This investigation set out to answer some of these questions by employing experimental and computational fluid dynamics (CFD) parametric studies. For the experimental study, a test rig was designed and constructed in accordance with European Standard BS EN 13141-1: 2004. The text rig was designed with sufficient flexibility to enable individual components and combinations of various components to be investigated. Components tested ranged from ordinary slots fabricated from pine wood to commercial ventilators and mesh-screens as are commonly found in natural ventilation applications. The results of this investigation indicate that the overall airflow properties of a ventilator are influenced by the combination of constituent components, the manner in which the components are incorporated into a ventilator and also the direction of airflow through it. A CFD study utilising three-dimensional models of some of the components tested during the experimental phase, and employing the k - ? turbulence representation was facilitated by a commercial software package (FLOVENT). Comparisons between CFD results and experimental data were reasonable and acceptable. A new mathematical approach, based up on experimental results, to analyse the airflow properties of combinations of ventilator components is introduced in this thesis. The proposed equations enable relative assessments of the impacts of each component on the overall airflow performance of a ventilator to be made. Comparisons between results from the proposed equations and those obtained experimentally showed good agreement. Although more work could be done to understand the physical meaning of the equations proposed, the author believes that this investigation has set a foundation for a generic representation of ventilator airflow performance. As such, refinements to the proposed equations would inevitably result if further research into this task is undertaken.