Methodology for the Evaluation of Natural Ventilation in ... - Cham

windows of naturally ventilated building. Though designed specifically for the prototype mainwindows, the design could be adapted for application for other window geometries and otherbuildings.There were additional factors that influenced the modeling methods along with the challengesaddressed in monitoring and evaluating the full-scale prototype building. The range of modelingmethods available, at scales ranging from full-sized to 1/250 th scale, and their characteristicswere presented. Two of the more common techniques for modeling, using either water or air,and their respective flow visualization techniques, were described. Flow visualization has beenshown to be a powerful technique to determine flow patterns within a model but there have beenconcerns raised regarding the neutral buoyancy of the material selected to introduce into the fluidstream. The use of fog, at the same temperature as the entering air, proved useful for use in thereduced-scale model airflow pattern visualization. Modeling techniques were described for arange of applications, including how each method incorporated flow visualization into theevaluation of fluid flow for the model. The requirements for tracing flow patterns included notonly neutral buoyancy, but also effective visualization and low diffusion rate. The benefits andproblems were identified for both water and air as the working fluids in modeling. Finally, themethods of flow visualization used for the reduced-scale model were presented, includingimaging techniques. Based on the above requirements, a fogging machine as the neutrallybuoyant tracer, a digital video camera to capture the flow visualization and fluorescent lamps toprovide illumination proved a useful combination.Critical to the use of scale modeling as a method for predicting fluid flow in full-scale buildingshas been the similarity of calculated or monitored parameters. Dimensional analysis andsimilitude were presented as means to ensure similarity between the prototype and scale model.The governing equations, in their full and dimensionless forms, were presented in chapter five.The resulting dimensionless parameters, the Reynolds, the Archimedes or the Grashof, and thePrandtl numbers, were identified, and determination of characteristic length was discussed. Theselection of characteristic length affected the regime used to describe the flow, and results in theuse of several characteristic lengths, depending on the field under evaluation. Comparisonsbetween the full-scale prototype, reduced-scale air, and reduced-scale water for the keyparameters including the Reynolds number, the Prandtl number, and the Grashof number werepresented. All of the models and the prototype were found to fall in the turbulent regime whenusing the hydraulic diameter of a single zone as the characteristic length.The reduced-scale air model constructed to meet the geometric similarity requirement waspresented in chapter six. Descriptions of the test chamber, control system and equipment used incarrying out the experiments were discussed, along with descriptions of the experiments. Eachof the experiments for buoyancy, wind, and combined wind-buoyancy cases was described indetail. The influence of window location on the flow pattern in the occupied space wasidentified using the single zone model. Though there was little change in the overall airflow ratethrough the building, the temperature distribution within the heated zone was affected by thelocation of the windows. The two-zone model demonstrated the effect of the stack vents on theairflow and resulting temperature distribution within the two zones. In poorly designed naturallyventilated buildings, the upper floor can become the only location for outflow from the building,causing warm, stale air to traverse the occupied space before exiting the building. The full-162