Heat transfer in pipes with twisted tapes: CFD simulations and validation

Abstract

Inserts are placed inside heat exchangers to promote turbulence and maximize the heat transferred. Twisted tapes enhance heat transfer with minimal pressure drop increase for double pipe heat exchangers. Their design typically relied on experimental correlations, but nowadays CFD software is gaining interest. The choice of the turbulent model is of paramount importance and not addressed in the literature. This research aims to compare the combinations of k-ε, k-ω and RSM as well as their different wall treatments available in Ansys Fluent® and literature experimental data. Different twist ratios and Reynolds numbers are tested. Currently, no research is found in literature comparing different CFD methods for this type of units. The main objective of this research is to find the combination of RANS turbulent model and wall treatment that will most accurately reproduce the global values needed (Nusselt number and friction factor) when designing a heat exchanger with twisted tape inserts. Results show that the selection of the wall treatment is far more relevant than the turbulence model. Simulations have less discrepancy between themselves than the empirical correlations. Best performing models were k-ε Standard with ML wall treatment, which provided an average deviation from correlations ranging between 15 and 18%. K-ω SST models also provided accurate performance when estimating friction factor values with 17 to 20% deviation. Results provide clues for choosing a suitable turbulent model and are useful to minimize the error provided by the models.