People spend most of their time indoors. A comfortable indoor environment is thus essential for the occupants’ good health and productivity. Buildings are responsible for about half of a modern society’s total energy consumption. HVAC (Heating, Ventilation and Air-Conditioning) which is often used to provide thermal comfort to the occupants, in turn accounts for a major proportion of this energy demand. Minimising HVAC energy consumption will thus result in great economic benefits. It also contributes beneficially to the issue of sustainable future and climate change, by reducing fuel burning. Natural ventilation can be used to help reduce significantly HVAC energy demand. Solar chimney (thermal chimney) is a device which absorbs solar radiation to heat the air. The heated air, becoming buoyant, rises through the chimney’s passage and induces further air currents. When fitted to a building, solar chimney can thus induce fresh outside air to flow through it for ventilation. As a very useful ventilation device, solar chimney has been the subject of many studies. However, due to the complex non-laminar, non-isothermal flow and heat transfer involved, there are still many factors affecting a solar chimney’s performance (measured by the induced flow rate of air, for instance) not yet considered, especially regarding 3-dimensional computational modelling in real-sized building settings. This work thus investigates computationally natural ventilation induced by a roof-mounted solar chimney through a real-sized 3-dimensional room, using a commercial CFD (Computational Fluid Dynamics) software package which employs the Finite Volume Method. Chien’s turbulence model of low-Reynolds-number K-ε is used in a Reynolds Averaged Navier-Stokes (RANS) formulation. Thus, the full set of Reynolds-Averaged governing equations pertaining to non-isothermal, buoyancy induced, incompressible, steady, turbulent flow of air near standard conditions at sea level, coupled with equations describing the Chien’s turbulence model, are solved, with appropriate boundary conditions. No further simplifying assumptions are made. Grid convergence tests are conducted to make sure that the grid patterns used are appropriate. Adequate numerical convergence is allowed; this often requires that relative changes in the successive iterated solutions be less than 0.0001. Accumulation errors resulting from massive or lengthy computation are also carefully monitored and minimized. 64-bits precision is used throughout. It is found that entrance geometry to the chimney’s channel affects significantly the ventilation rate, especially at higher solar heat flux, with rounded entrance resulting in higher rate. But these entrance-geometry effects also vary significantly with location of the room’s inlet-opening which in its turn affects the flow path before the chimney’s entrance.

This content is only available via PDF.
You do not currently have access to this content.