Investigation of Particle Trajectories and Flow Patterns in a Hydrocyclone by Positron Emission Particle Tracking (PEPT)
MetadataShow full item record
Cyclone separators are widely used for separation of solids or droplets from gaseous or liquid fluids in several industries. Cyclones can be used for many purposes and each cyclone can be fitted to its area of use. In the oil and gas industry there are problems with too much sand in the produced oil. Separation of the solids are critical for other separation and processing equipment and when concerning the environment effective hydrocyclones are very important. The purpose of this thesis is to investigate particle trajectories in a hydrocyclone to achieve a better understanding of the particle flow. A Positron Emission Tomography (PET) scanner and a technique called Positron Emission Particle Tracking (PEPT) are used to detect the trajectories. The PET scanner is placed at Haukeland University Hospital and is originally used to detect cancer cells in patients. In this thesis the PET scanner is used to track particle behaviour in a hydrocyclone which is done by making the particle radioactive and injecting it into the hydrocyclone. Based on cross-triangulation of lines of response (LOR) obtained from the output of the PET camera, the particle trajectory can be followed throughout the hydrocyclone. The use of this technique makes it possible to see details in the particle trajectory, and hence discover flow abnormalities and trends connected to the operating settings. The PEPT technique has been developed over twenty years and is currently used for process technology research just a few places in the world. Some numerical simulations has been performed to supplement the experimental results, rather than to perform a detailed comparison of the experiments. The detailed particle trajectories obtained in this thesis show a flow abnormality appearing in almost all the experiments. The flow phenomenon end of vortex (EoV) is also present in some of the experiments. The numerical simulations of the experiments coincide well with the experimental results. Additionally the Burgers model of flow gives good results when fitted to the tangential velocity of the particle.