Generalized Heat Transfer and Entropy Generation of Stratified Air-Water Flow in Entrance of a Mini-channel

In the present study a rule-based fuzzy inference system is used to predict heat transfer and entropy generation of stratified air-water flow in horizontal mini-channel as a function of a wide range of important parameters. Numerical data of our recent study are used to develop and test the system. The GK clustering algorithm is used to cluster the data. Fuzzy rules are generated based on the Sugeno-Yasukawa algorithm by using trapezoidal membership functions. The FATI and FITA approaches are implemented in the inference engine and finally the combination of the two approaches is defuzzified. The Mamdani and logical methods with the Yager operators are used and unified in both approaches. The parametric form of the system is a feature of the present study which can be used as an effective tool to improve the accuracy of the results. The novelty of the present study is the presentation of the generalized diagrams for the developing region of the channel which seems to be useful for engineering applications. In addition, generalized diagrams of average Nusselt numbers as well as total entropy generation can identify the appropriate range of volumetric flow rate ratio and the Reynolds number.