Evaluating the role of restricted nonlinear interactions in drag degradation in rough wall-bounded turbulence

Understanding and characterizing the role of mechanisms that erode the drag-reduction performance of riblets is critical for optimizing design. Recent studies have found spanwise rollers to be associated with the break-down of drag reduction and contribute to the subsequent drag increase in certain geometries. Secondary flows have also been linked to this performance degradation but the relative importance of these mechanisms and others has yet to be understood. In this study, we investigate the role of nonlinearity through a combination of DNS data analysis and direct investigation of relevant scale interactions using the restricted nonlinear model (RNL). The scale decomposition of the RNL provides a means to directly analyze and simulate the flow with restricted nonlinear scale interactions (which are confined to those contributing to the dynamics of the large-scale streamwise averaged mean flow). Comparisons are also made with data obtained from an augmented RNL (ARNL) simulation which includes intermediate modes that nonlinearly interact with the mean flow while still retaining order reduction, thus extending the number of scale interactions. The reduced order models allow us to isolate certain terms to understand the specific contributions to the spatio-temporal interactions. Results obtained using an amplitude-modulated triple decomposition of the data provide insight into these nonlinear contributions to the drag degradation mechanisms.