computational fluid dynamics modeling and multi-objective optimization of flat tubes partially filled with a porous layer using anfis, gmdh, and nsga ii approaches

In this work, the fluid flow in flat tubes armed with a porous layer is modeled and multi-objectively optimized utilizing computational fluid dynamics methods, adaptive-network-based fuzzy inference system, grouped technique of data handling type artificial neural network, and non-dominated sorting genetic algorithm ii. the variables design includes the tubes’ two geometrical parameters, porous layer thickness ratio, tube flattening, porosity, entrance flow rate, and wall heat flux. the purposes are to minimize the pressure drop and to maximize the convection heat transfer coefficient. initially, utilizing computational fluid dynamics methods the problem is solved numerically in different flat tubes to calculate two objective parameters in tubes. using numerical results of the preceding step, and are modeled through adaptive-network-based fuzzy inference system and grouped technique for handling the data. then, pareto-based multi-objective optimizing will be performed employing grouped technique of data handling model and non-dominated sorting genetic algorithm ii. the results revealed that a better predicting is obtained by the adaptive-network-based fuzzy inference system model compared to the other approaches and the significant design information is included in the attained pareto solution on flow parameters in flat tubes partly with porous insert. based on the findings, the best configuring for the highest heat transfer and the least pressure loss is hp= 0.75 and h=4mm.