Eur. Phys. J. Plus (2022) 137: 382
https://doi.org/10.1140/epjp/s13360-022-02578-x

Regular Article

Determination of heat transfer rates of heavy-duty radiators for trucks having flattened and double-U grooved pipes with louvered fins by ANN method: an experimental study

¹ Department of Mechatronics Engineering, Faculty of Mechanical Engineering, Yildiz Technical University, Yildiz, Besiktas, 34349, Istanbul, Turkey
² Ford Trucks, Sancaktepe R&D Center, Sancaktepe, 34885, Istanbul, Turkey
³ Department of Mechanical Engineering, Faculty of Mechanical Engineering, Yildiz Technical University, Yildiz, Besiktas, 34349, Istanbul, Turkey
⁴ Department of Mechanical Engineering, Engineering Faculty, Nigde Omer Halisdemir University, 51240, Nigde, Turkey

^c andacbaturcolak@hotmail.com

Received: 2 December 2021
Accepted: 7 March 2022
Published online: 22 March 2022

Abstract

In this study, radiators equipped with conventional pipes and newly proposed double-U grooved and brazed pipes are compared with each other under similar operating conditions. The radiators’ heat transfer performances have been evaluated experimentally and validation is done using the analytical Number of Transfer Unit Method. The cooling fluid has a fixed composition with 50% water and 50% ethylene glycol. Coolant flow rate changed between 2.5 and 7 kg/s, air velocity changed between 1.5 and 12 m/s and both inlet pressure and temperature values have been kept constant for air and coolant fluid sides throughout the experiments. Heat transfer rate, exit coolant temperature, friction coefficient, pressure drop inside the radiator and the total heat transfer coefficient have been assessed experimentally and compared under changing operation conditions. Using the obtained experimental data, an Artificial Neural Network model has been generated to determine the heat transfer rate for each radiator. The data taken from the numerical model have been compared with the practical one and analyzed extensively. It is observed that heat transfer rate and pressure drop are the highest for the Double U-grooved pipe radiator. The prediction values acquired from the reformed neural network are in good compatibility with the practical data.