Experimental investigations on modified thermosyphons using R134a/Al₂O₃ and comparative machine learning analysis

Thermosyphon operating with refrigerants are usual choices for low temperature heat transfer applications. The present study focuses on the performance evaluation of thermosyphons employing refrigerant R134a and Al₂O₃/R134 as working fluids, with novel designs to enhance the heat transfer. Two designs are proposed in the study in which eight internal axial fins are incorporated in one design termed as finned thermosyphon 1, and in addition to the internal fins, a uniform radial cut or gap is provided in the evaporator section in another design termed as finned thermosyphon 2. The performance of the thermosyphon shows that the thermal resistance of finned thermosyphon 1 and finned thermosyphon 2 decreases by 26.8% and 44.8%, respectively than conventional thermosyphon. Experimental investigations were also carried out using nanoparticles Al₂O₃ mixed with refrigerant R134a at a concentration of 1.0% concentration by weight on the modified design was employed in finned thermosyphon 2, as its performance was better when R134a was utilized. The thermal resistance of Al₂O₃/R134 based finned thermosyphon 2 decreases by about 55.3%, and 38.9% when compared to the conventional thermosyphon and finned thermosyphon 1, respectively. At the optimum condition, the heat transfer coefficient and surface area are determined for the conventional, finned type thermosyphons 1 and 2. The experimental analysis results are validated with a machine learning algorithm to estimate the deviation between the experimental and predicted results. The predicted and experimental results were found to be in good agreement with each other.