Adaptive neuro-fuzzy modeling of convection heat transfer of turbulent supercritical carbon dioxide flow in a vertical circular tube

• Published in: International communications in heat and mass transfer Vol. 37; no. 10; pp. 1546 - 1550
• Main Authors: Mehrabi, M., Pesteei, S.M.
• Format: Journal Article
• Language: English
• Published: Kidlington Elsevier Ltd 01.12.2010; Elsevier
• Subjects: Adaptive neuro-fuzzy inference system (ANFIS); Applied sciences; Circular tubes; Computational fluid dynamics; Convection heat transfer; Empirical analysis; Energy; Energy. Thermal use of fuels; Exact sciences and technology; Fluid flow; Fuzzy sets; Heat transfer; Mathematical models; Supercritical carbon dioxide; Theoretical studies. Data and constants. Metering; Turbulence; Turbulent flow; Convection heat transfer; Supercritical carbon dioxide; Adaptive neuro-fuzzy inference system (ANFIS); Turbulent flow; Nusselt number; Supercritical flow; Carbon dioxide; Circular pipe; Neural network; Modeling; Adaptive method; Convection; Vertical tube; Fuzzy logic; Numerical simulation; Heat transfer
• ISSN: 0735-1933; 1879-0178
• DOI: 10.1016/j.icheatmasstransfer.2010.08.019

Heat transfer of supercritical fluids has been the subject of many investigations; however, since the analysis of heat transfer in these fluids established by a mathematical model based on the planning parameters is complicated, this study attempts to provide a model for convection heat transfer of turbulent supercritical carbon dioxide flow in a vertical circular tube with a hydraulic diameter of 7.8 mm in inlet bulk temperature of 15 °C and a 8 MPa constant pressure by empirical results obtained by Kim et al.[1] and adaptive neuro-fuzzy inference system (ANFIS). At first, we considered Nu x as a target parameter and q w , G, Bo* and x + as input parameters. Then, we randomly divided 123 empirical data into train and test sections in order to accomplish modeling. We instructed ANFIS network by 75% of the empirical data. Twenty-five percent of primary data which had been considered for testing the appropriateness of the modeling were entered into the ANFIS model. Results were compared by two statistical criterions (R 2 and RMSE) with empirical ones. Considering the results, it is obvious that our proposed modeling by ANFIS is efficient and valid and it can be expanded for more general states.