Our daily life is heavily influenced by almost countless electronic devices, which continuously decrease in size while becoming increasingly powerful. In order to continue this development of increasing power density, the demand for efficient techniques of heat transfer increases as well. In recent years, the evaporation process has established itself more and more as the preferred means of transferring high heat flux densities because it uses the large amount of energy connected with the phase change. Indeed, heat pipes, which are routinely built into notebooks to cool processors, are the prime example for the application of this process. However, the biggest obstacle to increase the amount of transferable heat flux lies in the limited potential of heat transfer within the evaporator. As numerous studies by various authors have shown, the area of the thin liquid film close to the wall significantly influences the heat transfer. Stephan and Busse call this area micro region and determine the percentage of heat transferred through this region to be up to 45% of the originally supplied amount of heat. Based on these results, Brandt developed the concept of an advanced capillary structure, which could be used in heat pipes to maximize the area of the micro region by enlarging the surface through microchannels. As a result, Brandt was able to increase the heat transfer coefficient by 330%.

The objective of this study is the description of the respective influence of the heat flux density, the saturation temperature, the refrigerant, the surface topography of the evaporator, and the amount of non-condensable gas inside the system on the heat transfer during natural convection. In order to conduct all necessary experiments, an experimental set-up was designed and put into operation, which allows the study of five different surfaces of evaporators while varying the aforementioned parameters. The separate study of two parameters descriptive of the evaporator surface, i.e. the size of the interface and the length of the three phase contact line, and their influence on the heat transfer as well as the definition of the temperature of the far field by means of measuring the system pressure are novelties in this thesis. The numerical part of this study covers the development of a model to examine the influence of those parameters on the heat transfer that cannot be quantified in the experiment. These are the geometry of the transition between the wall and the area of thin liquid film, the curvature of the interface, and the value of the condensation coefficient. Further, using the model renders possible a detailed investigation of the experimentally determined global heat transfer, as it allows the analysis of the relationships between the total heat flux density and the heat flux density transferred through the micro region and the interface of the macroscopic liquid meniscus.

Comparing the heat transfer coefficients, the following conclusions concerning the heat transfer can be drawn based on the results of the experiments conducted: With minimal wall superheat and minimally applied heat flux density, the heat transfer coefficient reaches its maximum, which continuously decreases, even if not linearly, until nucleate boiling occurs. When the saturation temperature is increased while the heat flux density is constant, both the heat transfer coefficient and the maximally transferable heat flux increase as well, even though both are virtually independent from the topography of the surface of the evaporator.

The heat transfer coefficient also increases when the three phase contact line is extended and / or the interface is expanded. However, this increase is not linear and more pronounced when the three phase contact line is extended. Discovering that even the occurrence of a minimal amount of non-condensable gases in the system will lead to a significant change in heat transfer coefficient is another salient result of this study. The findings of the experiment further indicate that these gases directly influence the process of evaporation, which at the time of publication has never been described in the literature before. The results of the numerical model prove that heat transfer depends on the value of the condensation coefficient. In fact, the condensation coefficient cannot be determined through the experiment, in which it drastically decreases due to inevitable contamination of a few parts per million, which in turn results in lesser heat transfer. Furthermore, using the numerical simulation, the heat flux ratio between micro zone and macroscopic interface can be quantified. Hence, the dependencies of heat transfer on the applied heat flux density, saturation temperature, and refrigerants respectively – which have been determined through the experiment – can be better understood using the numerical model. With this work it is now possible to describe the heat transport across several length scales starting from the micro region up to macroscopic interface.