Multi-objective optimization of micro-fin helical coil tubes based on the prediction of artificial neural networks and entropy generation theory

The heat transfer, flow resistance and entropy generation characteristics of micro-fin helical coil tubes (MFHCTs) are investigated numerically. MFHCT with different fin numbers (2 <= N <= 6), coil pitches (150 mm <= P <= 450 mm), coil diameters (600 mm <= D <= 1200 mm) and Reynolds numbers (10945 <= Re <= 30845) are examined. The effects of these geometric parameters on the Nusselt number (Nu), friction factor (f) and improved entropy generation number (Ns') are discussed. The performance of MFHCT is then compared to that of a smooth helical coil tube (SHCT). The results show that MFCHT always performs better than SHCT, especially in the lower Reynolds number region. Moreover, artificial neural networks (ANNs) are established to predict Nu, f and Ns', which are trained by simulation data. This model fits the simulation results better than the multiple linear regression, and the maximum error is no greater than 8%. With the prediction of the network, the micro-fin helical coil tubes are optimized by the entropy minimization method and NSGA-III algorithm. Through optimization, the distribution of design variables is examined. The results demonstrate that a higher Reynolds number and a larger coil diameter and coil pitch lead to a better performance. Additionally, the optimal Pareto points can be utilized to guide the design and operation conditions of micro-fin helical coil tubes.