|dc.description.abstract||Interactions between droplets and liquid films are found widely in our daily life, and many interesting phenomena can be observed. Scientists study the phenomena not only because they are fascinating but also apply the knowledge to many fields such as industry, biology, oceanography, astronomy etc.
The knowledge can be used to improve the efficiency and to develop design tools for heat exchangers in the industrial LNG processes. The experimental investigation of micro-scale level droplet-film interactions is critical in order to improve the understanding in this field. The main focus of the study is to experimentally investigate the vertical impact between droplets and a deep liquid film of the same fluid. The investigation aims at improving the understanding of different phenomena in the drop-pool impacts.
A literature review showed that there was insufficient information on micron-level droplets (diameter below 1 mm) impacting with a deep pool, and thus the present work aimed at giving this part of information. An experimental setup was designed and constructed in order to carry out the experiments in a controllable manner. The setup had a special function which reduced the impinging frequency of a droplet stream, and thus the impact can be studied with a reasonable isolation from the impacts of the neighboring droplets. Besides, other components designed and used in the experiment, such as the droplet generator, light sources, safety issues etc., are described in detail in the this work. The experimental setup enables the generation of droplets with the diameter range approximately 0.1 mm--0.7 mm and the velocity range approximately 0.1 m/s--10 m/s. The uncertainty analysis showed that the relative uncertainty for diameter and velocity measurements are generally below 5%, and the relative uncertainties for the dimensionless numbers (Re, Oh, We, Fr and Ca) are generally below 10%.
Four different phenomena, coalescence, bouncing, partial coalescence and jetting were generated and observed by using different fluids including distilled water, technical ethanol, n-pentane, methanol and 1-propanol. Observations of different phenomena are presented and described thoroughly. Results are presented with the uncertainties which are evaluated specifically for this work.
Data analysis was carried out to characterize the thresholds between different phenomena, two regression methods, the least squares and the least points, were used to find the curve-fitted threshold models. The thresholds between coalescence and jetting for five fluids are characterized using an exponential model using We and Oh and a linear model using Fr and Ca, and both models give very good characterizations with few uncertain points within the diameter and velocity ranges in the present study. The literature jetting-threshold data with much larger diameter (up to 3 mm) and lower velocity (approximately 1 m/s) was compared with the models, and the comparison showed that the exponential model applies better in the millimetric range than the linear model. For predicting the thresholds for fluids other than the five experimental fluids, calculation methods for the parameters in both models are suggested.
Two thresholds between coalescence and bouncing are characterized by using the critical Weber number, at which a phenomenon transits to the other. The thresholds of bouncing-coalescence are characterized for distilled water, technical ethanol and 1-propanol, and the thresholds of coalescence-bouncing were characterized for distilled water and technical ethanol. For assessing the energy loss during bouncing, the restitution coefficient was analyzed, and the stable levels of the restitution coefficients were between 0.2-0.3 which agreed well with the literature.
Based on the observations, characterizations of thresholds and analysis of the restitution coefficients, the effects from the physical properties of the fluids were analyzed. The effects of viscosity was found very dominant. Due to the dissipation of the turbulence, viscosity reduces the perturbations for the crown formation and breaking, giving higher critical Weber number for the bouncing-coalescence threshold and higher restitution coefficient. Surface tension inhibits the formation of the crown and giving higher restitution coefficient due to better elasticity.||nb_NO