Numerical analysis of heating aerosol carbon nanofluid flow in a power plant recupesrator with considering ash fouling: a deep learning approach

The main goal of this research is to optimize a corrugated high-temperature helical recuperator filled with aerosol-carbon-black nanofluid. This heat exchanger has inner semi-sphere-shaped corrugations with different diameters. In this study, the related effects of different geometric parameters according to the heat-hydraulic performance of the studied helical heat exchanger are analyzed. Moreover, to select the optimal model, the hydraulics-thermal index is examined and its maximum value is introduced as the optimal model. The Discrete Phase, Model (DPM) approach is also used to model multi-phase currents. In order to numerically simulate and solving the governing equations, the Ansys Fluent software, SIMPLE algorithm and the finite volume method were used. As it is realized, the helical corrugated recuperator with corrugation diameter of d=10mm and filled with nanofluid (NF) with volume concentration of phi=0.7% has the maximum thermal-hydraulic performance evaluation criteria. Then, a deep learning method is applied to data in order to obtain the average Nusselt number (Nuav), the pressure drop (Delta P) between outlet and inlet, the friction factor (f), and the PEC for all values of the volume fraction, mass flow rate, and turbulators diameter within the examined range. To this end, an Artificial Neural Network (ANN) with one hidden-layer is employed. The results indicate that the ANN can model the investigated phenomenon with acceptable level of precision.