The goal of this study is to evaluate the computational fluid dynamic (CFD) predictions of friction factor and Nusselt number from six different low Reynolds number k–ε (LRKE) models namely Chang–Hsieh–Chen (CHC), Launder–Sharma (LS), Abid, Lam–Bremhorst (LB), Yang–Shih (YS), and Abe–Kondoh–Nagano (AKN) for various heat transfer enhancement applications. Standard and realizable k–ε (RKE) models with enhanced wall treatment (EWT) were also studied. CFD predictions of Nusselt number, Stanton number, and friction factor were compared with experimental data from literature. Various parameters such as effect of type of mesh element and grid resolution were also studied. It is recommended that a model, which predicts reasonably accurate values for both friction factor and Nusselt number, should be chosen over disparate models, which may predict either of these quantities more accurately. This is based on the performance evaluation criterion developed by Webb and Kim (2006, Principles of Enhanced Heat Transfer, 2nd ed., Taylor and Francis Group, pp. 1–72) for heat transfer enhancement. It was found that all LRKE models failed to predict friction factor and Nusselt number accurately (within 30%) for transverse rectangular ribs, whereas standard and RKE with EWT predicted friction factor and Nusselt number within 25%. Conversely, for transverse grooves, AKN, AKN/CHC, and LS (with modified constants) models accurately predicted (within 30%) both friction factor and Nusselt number for rectangular, circular, and trapezoidal grooves, respectively. In these cases, standard and RKE predictions were inaccurate and inconsistent. For longitudinal fins, Standard/RKE model, AKN, LS and Abid LRKE models gave the friction factor and Nusselt number predictions within 25%, with the AKN model being the most accurate.
Convective heat transfer enhancement in low Reynolds number flows with wavy walls
