Experimental strategies for frost analysis

Abstract

An area of increasing importance in the field of refrigeration is the study of frosting and defrosting. Frosting poses a concern to many refrigeration systems, as frost growth both obstructs airflow through low temperature heat exchangers and increases heat transfer resistance. Drastic decreases in system efficiency result from the compounding of these problems, and because it is difficult to prevent the frosting process, refrigeration systems must be defrosted periodically to restore optimal operating conditions. A deeper understanding of the complex physical processes of frosting and defrosting will lead to more efficient refrigeration system designs; an idea which has driven a rise in frost growth research over recent decades. Although research has shown great progress, there remain significant challenges associated with predicting the frosting and defrosting processes accurately under wide ranges of conditions. The equations governing such behavior still remain insoluble by exact analytical methods. Numerical approaches have shown the most promising results, but are yet in an early stage of development. Most research has instead been concerned with developing correlations for frost properties and growth, though few are applicable to varying conditions. The most commonly used correlations are shown to have widely different results, perhaps owing to different experimental methods used to acquire data and a lack of deeper level analysis. A new thickness correlation is proposed which attempts to reconcile to some degree the gap between theory and application. Broader ranges of data are used for fitment which enables the application of the correlation to a wider range of conditions. To improve the consistency of results in frost research, it is suggested that new forms of data acquisition be explored. Proposed alternative methods utilize high magnification imaging equipment in combination with computer based measurements, which are shown to be capable of improving accuracy by an order of magnitude in some areas (specifically frost thickness measurement) when calibrated appropriately. In addition to improving measurement accuracy such methods make possible the rapid calculation of droplet geometry during defrosting, an area which has seen little research until recently. The influence of the experimental apparatus on results is also investigated, and a variety of different setups used in past and recent research are categorized according to capability and functionality. Pros and cons of related parameters are discussed with an emphasis on goals. Opportunities for future work include the further development of computer based measurement methods, the acquisition of data over wider ranges of conditions and improvements on the experimental apparatus required to achieve those conditions reliably.It is clear from this research that frost growth is a developing field where much progress is yet to be made. Experimental setups of types ranging from small enclosed tests to wind tunnels on industrial evaporators have provided a clearer understanding of the phenomenon in many aspects. Research presented in this thesis shows that small scale experiments are preferable at this point in time to reach deeper understanding of the frost growth process. It is shown here that many current methods of measurement for important frost growth parameters can be greatly improved upon by the use of computer based algorithms. Faster and more accurate measurement opportunities mean that larger data sets spread across wider ranges of testing conditions can be obtained, setting the stage for more advanced correlation development. Currently, most correlations are only applicable to specific conditions and are still not highly accurate. An attempt is made to show that larger collections of reliable data can be used to develop more robust correlations. To do so a new correlation is proposed which fits a wide range of conditions well. Finally it is shown that the defrosting process may be understood more fully by the use of digital analysis of visual data during defrosting.