A \(\phi\)-Variance Wall Function and Spectral Scaling for Turbulent Channel Flow
Accurate near-wall modelling of second-moment turbulence statistics remains a persistent challenge in computational fluid dynamics. Classical wall functions are primarily designed for the mean velocity and are considerably less developed for quantities such as the streamwise velocity variance. In this paper, we propose a \(\phi\)-based framework that exploits the inverse hyperbolic sine transformation \(\phi = \operatorname{arcsinh}(U^+/U_{\mathrm{ref}}^+)\) and the exact quadratic relation \(\langle\phi^2\rangle = \phi_0^2 + \langle\phi'^2\rangle\), which follows directly from the Reynolds decomposition. Using this relation, we construct a wall function for the \(\phi\)-variance by reconstructing the streamwise variance \(\langle u'^2\rangle\) with a piecewise cubic Hermite interpolating polynomial. Validated against direct numerical simulation (DNS) data for turbulent channel flow at \(Re_\tau = 395\), the proposed wall function achieves a coefficient of determination \(R^2 = 0.9792\) in the near-wall region (\(y^+ \le 30\)). We further derive a spectral scaling law, \(E_\phi(k) = E_u(k)/(1+U^{+2})\), which provides a simple kinematic interpretation of how the energy distribution of the transformed variable varies across the boundary layer. For completeness, we also assess a blended \(\phi_0\) wall function for the mean velocity; it is found to be competitive with, but not superior to, classical laws. This assessment delineates the scope of the framework and shows that its primary value lies in the variance and spectral domains. The proposed wall function is algebraic, free of empirical damping functions, and easy to implement in existing Reynolds-averaged Navier--Stokes codes.
Authors
- Mahmoud F. Abdel-Sattar
Publication Details
- Journal
- Zenodo (CERN European Organization for Nuclear Research)
- Published
- 2026-11-14
- DOI
- https://doi.org/10.5281/zenodo.24220382
- Primary Topic
- Fluid Dynamics and Turbulent Flows
- Type
- preprint