Determination of k-factors of HVAC system components using measurement and CFD modelling
This thesis conforms conventional and advanced experimental techniques for the measurement of and mathematical prediction of velocity pressure-loss factors (k-factors) for fittings used in heating, ventilation and air-conditioning (HVAC) systems. After an extensive study of different tracer-gas experimental techniques, the constant injection method is applied to various duct fittings on a small scale HVAC system situated in a laboratory. The results are compared with those of experiments performed using a more conventional technique using a Pitot-static tube. The basis of the experimental procedure is to achieve an accurate method of measuring the mean air velocity within a duct. This allows an accurate estimate of the velocity pressure-lossf actor to be obtained. A wide variety of duct fittings are investigated experimentally and numerically including bends, transitions, branches, inlets, outlets and obstructions such as orifice plates, wire mesh and lateral pipe obstructions. Computational fluid dynamics (CFD) is applied to each duct fitting tested in the lab. A commercially available package FLUENT is used with a high powered computer to simulate the airflow through various duct fittings. The pressure loss and velocity vectors are predicted for each particular duct fitting and therefore a prediction for k-factors is obtained. k-factor predictions are compared with experimental results and published data given in ASHRAE and CIBSE guides in order to assess the accuracy of CFD prediction. It is shown that as an accurate method for prediction of k-factors in duct fittings, CFD is a useful tool for the design and development of HVAC systems. The application of CFD allows the designer to vary any duct component with ease to observe the effect on a particular duct fitting without incurring the expense of laboratory experimentation. It is also shown that values of current published kfactors are greatly over estimated leading to oversizing of HVAC system fans. Experimentally produced k-factors obtained using the tracer-gas method and CFD predictions are approximately 20% lower than current data available to HVAC system designers. CFD may be applied to various applications in the field of heat-pumps and refrigeration systems. A detailed investigation is carried out here to compare CFD prediction and experimental results of several low pressure and high pressure ejectors commonly found in refrigerator absorption cycles. The compressible flow of refrigerants was modelled through an ejector to obtain a prediction of the entrainment ratio ( i. e. the ejector's ability to entrain a refrigerant from an evaporator using a hot main flow through a nozzle). These predictions were then compared with experimental results and this indicated that CFD could serve as a useful tool in the design of refrigeration systems. Application of CFD has also been studied in relation to the investigation of pressure loss through different types of evaporator/condensecr oils found in heat pump systems; here the design of such coils is important to the operating efficiency. The pressure loss across heat-pipes found in ducted flows is also predicted using CFD; in this case the geometry and the thermal conditions play an important role in the overall pressure loss.