Numerical investigatinon of pool boiling on flat and patterned surface

Boiling is named the phase change between liquid and solid surface when liquid is the boiling point a and the surface's heat is over the liquid's temperature. There are two main type of boilings: pool boiling and flow boiling. Pool boiling takes place in a constant volume, while flow boiling occurs in a moving system with forced convection.During the boiling process, heat transfer rate increases greatly due to the phase change. Therefore, pool boiling and flow boiling has a important role at many enginnering applications. Especially, it affects directly in electronic cooling, in energy conversion systems and in advanced engineering systems that produce steam. For this reasons, many researches have done for boiling to be understood the boiling mechanism. Thanks to many researches, variables was found that ınfluences on boiling mechanisms. These variables have three groups that are macroscopic, microscopic and statistical. Experiments was made to get information about the interaction of variables with each other and interactions within themself. By using experimental results, many correlations were develoed for variables. These correlations were improved by observing the effect of several parametres acting on the variables. Therefore, a relation involving all boiling variables has not yet been found. In this study, pool boiling was investigated. Pool boiling has four different boiling phases. These stages are natural convection, nucleate boiling, transition boiling and film boiling, respectively. The content of this work was done for the nucleate boiling. In this area, macroscopic and statistical variables affecting boiling were examined. These variables are bubble departure diameter, bubble frequency and active nucleation site density. The aim of this study is that the correlations of the variables obtained with experimental data was modeled by using CFD method for pool boiling on smooth surfaces. Initially, numerical modeling of the nucleate boiling zone of the pool boiling was performed using Ansys Fluent 15 software, a commercial CFD program. The 2D smooth surface geometry for the numerical model was created in Ansys Geo. This 2D geometry has two different zones. The first zone is the flow zone and the second zone is the solid zone. The model boundary conditions of the smooth surface are defined as follows: The side walls of the flow zone are adiabatic, the upper surface of the flow zone is the atmospheric outlet, and the wall where the flow zone contacts the solid surface is named the hot surface. For working fluid the water was choose and the smooth surface material was stainless steel. In order to model the nucleate pool boiling on the smooth surface, the water phase must pass into the vapor phase. For this, the mass transfer between water and vapor was modeled. Bubble departure diameter and bubble frequency correlations have been used for modeling the mass transfer. In the mass transfer model, the bubble departure diameter was primarily calculated. After calculating the bubble diameter, the bubble frequency was calculated by using the bubble departure frequency correlation. The calculated bubble frequency was substituted in the Lee model so that the mass of water evaporated was calculated. The mass transfer modelling was done by using UDF (User Defined Function) ın Ansys Fluent. The latent heat was computed from the mass of water evaporated. Thus, the water-vapor phase changewas modeled by the bubble departure frequency applied to the mass transfer model. In Ansys Fluent, nucleate pool boiling was assessed by using three different correlations and one default model at the VOF (Volume of Fluid) phase model. Analyzes were performed at constant surface temperature. The surface temperature was 130 ° C and the initial temperature was 27 ° C The Kutateladze and Gogonin correlations (bubble departure diameter correlation) were found to be appropriate in terms of the results obtained by modeling the mass transfer model with three different correlations. By using the correlation of Kutateladze and Gogonin, the analyzes were repeated for the surface temperature that was 110 ° C and 120 ° C. As a result, for the mass transfer, with rising superheat (difference between surface temperature and saturated temperature) the bubble departure diameter was observed to increase but the calculated bubble frequency decreased. On the other hand, the heat flux and heat transfer coefficient values were calculated by CFD method for the smooth surface that were found to be below the boiling curve with increased superheat. In view of these results, we can see that with the heat transfer and conservation equations were not sufficient for modeling pool boiling on the smooth surface that were calculated with CFD method. Because the wal heat flux that was calculated by CFD did not include the effect of the physical phenomena occurring during boiling such as natural convection and the effect of bubbles. This shows that an additional application is required to calculate the wall heat flux. Therefore, the boiling relations have been utilized to model the heat transfer. In this study, Nusselt number developed from boiling relations was used to model heat flux. The heat transfer coefficient was calculated by using the developed Nusselt number correlation for the nucleate boiling region. The calculated heat transfer coefficient was used to model the wall heat flux. Modeling of the wall heat flux was performed by using UDF (User Defined Function) ın Ansys Fluent.The model was realized at two different phase models at different fixed surface temperatures. The surface temperatures were 120℃ and 130℃. The solutions were also evaluated according to different initial temperature values, depending on whether the fluid is saturated or subcooled. The initial temperature values were 97℃, 77℃ and 27℃. In the mass transfer model, Kutateladze and Gogonin correlations were used for the bubble separation diameter and Zuber correlation was used for the bubble separation frequency. As a result, heat flux results were calculated by using VOF and Eulerian phase model at different surface temperatures and fitted the boiling curve. It has been seen that the heat flux results changes according to whether the fluid is saturated or subcooled. For this, it is predicted that the heat flux model should have a coefficient factor that varies with the flow temperature. The mass transfer model and the heat flow model obtained from the smooth surface were used to model the nucleate pool boiling for the patterned surface. Three different analyses have been solved for a single geometric structure. In Ansys Fluent VOF phase model is used to model. The patterned surface was taken as 120℃ at the constant temperature and the fluid starting temperature was taken as 77℃. The first analysis for the patterned surface were performed without changing the mass transfer and heat flux model that were applied to the smooth surface. On the order hand, ın the second and third analysis models, the mass tranfer model coefficient is changed. In the third model, the heat flux model coefficient was increased to 4 times. Although as a result of the calculations made with CFD, the heat fluxes of the 1st and 3rd model results were very low and the surface was covered with vapor layer, ın Model 2 the heat flux was increased and bubbles were observed on the surface. In addition, by using Model 2 analyses were repeated for the surface temperature 110℃ and 115℃. As a result, experimental data and analyses results were compared. It can be seen that there is a significant difference between the heat flux and the heat transfer coefficient obtained from the experimental results of the patterned surface with different superheats and the heat flux and heat transfer coefficient calculated from the analyses. Because of the lack of variables related to surface properties and the complexity of interaction between surface physical properties and bubble diameter and frequency, the results of the analyses must be validated with experimental results. In this study, numerical modeling of the pool boiling on the patterned surface was made by using the mass transfer model and heat flow model coefficients. Finally, in this study, the mass transfer and the wall heat flux were modeled by using the relations obtained from the boiling variables. The mass transfer model was performed using bubble separation diameter and bubble separation frequency correlation. The heat flux model was also made using the Nusselt number correlation obtained from the boiling relations. Modeling is done by writing UDF (User Defined Function) in Ansys Fluent. For the smooth surface, the analyzes were performed at the constant surface temperature and 1 atm pressure. In addition, the modeling was carried out at different fixed surface temperatures and starting temperatures. Hence the mass transfer and the wall heat flux were calculated in Ansys Fluent. Accordingly, bubble formation was observed on the hot surface by the mass transfer model. Further, bubble formation, heat flux and heat transfer coefficient changed according to surface temperature and initial temperature. The results obtained from the analyses were compared with the boiling curve and experimental data. The wall heat flow model provides an overlap with the boiling curve. The wall heat flux obtained as a result of repeated analyses for the patterned surface was found below the experimental datas. This is because of the complexity of the boiling phenomenon and the lack of information on the surface geometry of the correlations. As a result of this work, the nucleate boiling phase of the pool boiling is modeled using CFD for the smooth and patterned surface.