Study and Numerical and laboratory simulation of effective heat transfer parameters using the new structure of a novel insert in heat exchangers

The present study investigates the effect of different insert inside the tubular heat exchanger on important thermal parameters and overall efficiency. In this regard, methods such as Design of Experiments (DOE) software and response surface methodology (RSM) Central composite design (CCD) were used to design experiments. Numerical methods and neural network modeling were also used to further strengthen the analysis and then the mentioned simulation methods were compared. In all cases, the results of this study indicate an increase in heat transfer coefficient and, consequently, an improvement in the Nusselt number due to the use of novel inserts. So that unlike the previous inserts of these inserts, in addition to rotating the current, it also has the ability to rotate. Finally, the average percentage of absolute error (MAPE) in the experimental design software modeling method and neural network is 10.77 and 0.94%, respectively, which indicates the high accuracy of the neural network method to determine the friction coefficient parameter.