Numerical simulation and indoor airflow and heat transfer study for thermal comfort

An investigation of indoor thermal environment has been carried out by computational fluid dynamics approach. The study focuses on the thermal comfort evaluation, particularly the flow and heat transfer effects due to conjugate natural convection, furniture arrangement and occupant number, and flow oscillations. Key physical features of thermo-fluid such as velocity and temperature distributions, thermal sensation maps, and oscillation frequency and its energy are quantified, analysed and compared. The benchmark case study of airflow and heat transfer showed that ANSYS Fluent RNG k - 5 turbulence model with temperature boundary condition on the heated boundary calculated the bestresults, compared with available data. It also showed that air velocity increased along the boundary walls and especially hot wall which led flow direction upwards. At the centre of the flow circulation, air momentum is very weak (e.g. almost zero velocity magnitude). The increase of complex features (6. g. a box with/without heat) in the domain would lead to flow separations causing recirculations above the box and in the rear space of the domain and swirls in the front space presenting three-dimensional flow, and a thermal plume, compared with a two-dimensional clockwise flow in an empty room. The flow recirculations and thermal buoyancy enhanced velocity magnitude and turbulence level in the domain. In fact, the highest frequency was obtained in the room with an unheated box, followed by the room with a heated box. The forrhation of thermal plume from the heated box stabilised the flow in the upper part and the sides of the heated box on a spanwise plane. The frequency of velocity oscillation was consistent with temperature at the location although the energy of the fluctuation is much higher in temperature. Moreover the dominant frequency depended on the orientation of the flow circulation, for example a high energy at a lower frequency on a spanwise plane while a low energy at a higher frequency on a streamwise plane. In an empty room, it was found that there is no direct relation in an empty room (case 3.2.1) between velocity and turbulent flow in power spectral density and frequency, and each of time-history velocity oscillations is independent and random. At the mid- height of the domain, the energy of the velocity fluctuation is relatively weak. The results from the study of conjugate natural convection heat transfer in a ventilated room with localised heat source and window glazing showed that the size of heat source and window glazing, the wall thickness and wall material property are important factors to temperature change and heat loss. For example, 30 % of wall thickness reduction caused 35 % more of heat loss through the wall and 9 % of comfort temperature. From the study of furniture arrangement and occupant number in a 3-D model room with localised heat source and window glazing, it was found that the presence of furniture induced flow recirculation and higher velocity around furniture and the presence of thermal occupant formed thermal plume in the fluid domain, increasing volume-averaged temperature by maximum 15 %, compared with that of unoccupied and empty model room. Increase in the number of occupants and thermal furniture helped increase air temperature by 6.5 %, compared with that of single occupant and the averaged PPD (Predicted Percentage of Dissatisfied) value around the occupants by maximum 5.4 % for one occupant and 11.5 % for two occupants, respectively. The location of occupant was very sensitive to flow stream path, e. g. the PPD distribution was symmetrical in the spanwise position but became asymmetrical in streamwise position. Further investigation of thermal comfort level using Fanger's indices due to ventilation rate and thermal load led that desirable indoor environment might be achieved with higher ventilation flow rate (Uinlet > 0.7 m/s) rather than reducing heat generation from the heating sources for more occupants introduced to the room. The results in the thesis summarise some of the important reservations with regard to the CFD capability and reliability for indoor thermal environment and present data would be useful for the built environment thermal engineers in design and optimisation of domestic rooms.