Title:Revisiting the Lie-group symmetry method for turbulent channel flow with wall transpiration

Abstract:The Lie-group-based symmetry analysis, as first proposed in Avsarkisov et al. (2014) and then later modified in Oberlack et al. (2015), to generate invariant solutions in order to predict the scaling behavior of a channel flow with uniform wall transpiration, is revisited. By focusing first on the results obtained in Avsarkisov et al. (2014), we failed to reproduce two key results: (i) For different transpiration rates at a constant Reynolds number, the mean velocity profiles (in deficit form) do not universally collapse onto a single curve as claimed. (ii) The universally proposed logarithmic scaling law in the center of the channel does not match the direct numerical simulation (DNS) data for the presented parameter values. In fact, no universal scaling behavior in the center of the channel can be detected from their DNS data, as it is misleadingly claimed in Avsarkisov et al. (2014). Moreover, we will demonstrate that the assumption of a Reynolds-number independent symmetry analysis is not justified for the flow conditions considered therein. Only when including also the viscous terms, an overall consistent symmetry analysis can be provided. This has been attempted in their subsequent study Oberlack et al. (2015). But, also the (viscous) Lie-group-based scaling theory proposed therein is inconsistent, apart from the additional fact that this study of Oberlack et al. (2015) is also technically flawed. The reason for this permanent inconsistency is that their symmetry analysis constantly involves several unphysical statistical symmetries that are incompatible to the underlying deterministic description of Navier-Stokes turbulence.