Optimization design and PIV experimental investigation on the heat transfer and flow characteristics in circular tube with oblique wavy tape based on machine learning

To induce longitudinal swirl flow and enhance heat transfer in heat exchanger tubes, a novel oblique wavy tape insert is introduced in this paper. With identical geometric parameters, the flow and heat transfer characteristics of a tube with V-shaped oblique wavy tape (VOWT) are numerically compared to those of tubes inserted with existing conventional wavy tape (CWT) and center-tapered wavy tape (CTWT). The results show that the VOWT induces larger tangential velocity and a stronger longitudinal swirl flow with two pairs of vortices, effectively reducing flow dead zones. Among the three wavy tapes, the VOWT demonstrates superior heat transfer and overall performance. To expand design space and achieve further performance improvement, multiple types of wavy tape are parameterized into a unified form for optimization. Four geometric parameters of the unified wavy tape are considered as design variables, with the performance evaluation criterion (PEC) as the optimization objective. The surrogate model based on machine learning (SML) is employed, which incorporates the Kriging model, genetic algorithm, and machine learning technology. The SML method enhances the local representation of the surrogate model and contributes to improved efficiency. The optimization results show that the optimal wavy tape (OWT) is identified as an oblique wavy tape. Further analysis through field synergy and thermalhydraulic performance evaluation reveals that the OWT promotes the synergy between velocity and temperature fields. Within the Reynolds number range of 500-1500, heat transfer in the tube with OWT is enhanced by 7.10 to 11.86 times compared to a smooth tube, and the PEC ranges from 3.21 to 4.44. This capability of achieving both high heat transfer and overall performance makes oblique wavy tapes suitable for high heat flux scenarios. Finally, the optimized flow field achieved by OWT is verified through a particle image velocimetry (PIV) experiment.