Enhanced Thermal Performance of a Circular Heat Exchanger with Internal Blades through Bayesian Optimization and Experimentation

Abstract

This study presents an experimental investigation of convective heat transfer and hydrodynamic behavior in a circular water‑cooled heat exchanger equipped with internal blades under turbulent flow. A 1000 mm long, 30 mm inner‑diameter tube was tested in plain and internally bladed configurations with 2, 4, 6, and 8 blades over a Reynolds number range of 6000–14,000. Inlet and outlet temperatures and pressure drops were measured to determine the Nusselt number, friction factor, convective heat transfer coefficient, and thermal performance factor (TPF), enabling a detailed assessment of heat transfer enhancement relative to the associated hydraulic penalties. The results show that internal blades effectively disrupt the thermal boundary layer and generate secondary flow vortices, yielding a substantial increase in convective heat transfer accompanied by a rise in friction factor. To identify operating conditions that maximize thermal efficiency, Bayesian optimization was applied to the experimental database, predicting the optimal combination of blade count and Reynolds number that achieves the highest TPF. The optimized configuration delivered superior heat transfer with an acceptable pressure drop penalty, demonstrating the value of Bayesian optimization as a predictive design tool and supporting the development of compact, high‑efficiency water-cooled heat exchangers for industrial applications.