Analytical Wall-Function Strategy for the Modelling of Turbulent Heat Transfer in the Automotive CFD Applications

In contrast to the well-established “standard” log-law wall function, the analytical wall function (AWF) as an advanced modelling approach has not been extensively used in the industrial computational fluid dynamics (CFD) applications. As the model was originally developed aiming at computations on relatively coarse meshes, potential stability issues may arise due to the pressure-gradient sensitivity if employing locally inappropriate mesh layers, typically associated with the complex geometry details. This work evaluates performance of the thermal AWF, as proposed by Suga [4], in conjunction with the main flow field computed employing the k-ζ-f turbulence model and the hybrid wall treatment (denoted as AWF-e) within the Reynolds-averaged Navier-Stokes (RANS) framework. The underlying turbulence modelling approach has been widely validated in numerous industrial applications, demonstrating capability (in terms of both accuracy and robustness) to capture near-wall transport phenomena with more fidelity compared to the standard or low-Reynolds-number variants of the k-ε turbulence model. The proposed AWF-e strategy is validated on several benchmarks, namely heated pipe, E-motor cooling jacket and IC engine flows. These flow configurations involve elevated temperature gradients and fluid property variations, typically encountered in the automotive applications. The results confirm reduced mesh sensitivity and superiority of the AWF-e over the conventional RANS wall heat transfer models.