Performance prediction and evaluation of heat pipe with hexagonal perforated twisted tape inserts

Abstract

Efficient heat transfer technologies are critical in a wide range of industrial applications, including air conditioning, chemical reactors, and heat exchangers. One method for improving heat transfer performance is to use twisted tape inserts in heat exchanger tubes. Heat transmission is aided by the disturbance of fluid flow caused by these inserts, although research is still ongoing to establish the specific design components that maximize their efficacy. The research focuses on heat transfer optimization in practical applications by exploring hexagonal perforated twisted tape inserts with varied cut orientations (horizontal, vertical, and alternate) and a pitch ratio of 4. The problem becomes more complex without a complete numerical prediction model. The study seeks to construct a hybrid deep neural network based on a gannet optimization algorithm (DNN-GOA) model in order to estimate heat transfer performance accurately. According to the experimental results, the TTA’s specific design with alternate cuts produces a thinner thermal boundary layer and a higher convective heat transfer coefficient for Nusselt number (Nu), friction factor (f), and thermal performance factor (TPF). The Hybrid DNN-GOA model has the best predictive performance, with a high R² indicating a tight match between anticipated and real Nu, f, and TPF values. It also exhibits the lowest Root Mean Square Error (RMSE), Mean Absolute Error (MAE), Mean Absolute Percentage Error (MAPE), and Mean Squared Error (MSE), confirming its exceptional accuracy.