July 18, 2016 — July 22, 2016
- Sergio Pirozzoli (Sapienza University of Rome, Roma, Italy)
Despite signifcant progress achieved in recent years, fluid turbulence is still escaping our complete understanding, mainly because of its complex nonlinear behavior. Within the broad subject of turbulent flow, turbulence over solid walls is of special conceptual and practical importance, and its prediction is crucial for the accurate design of aircraft, turbo-machines and ships. Understanding the physics of wall turbulence may lead to effective techniques for the reduction of wall friction, with incurred benefits in terms of reduced power expenditure.
The present course is aimed at presenting the state-of-the-art of wall turbulence and highlighting avenues for future research.
The emphasis will be mainly on canonical ows over flat surfaces including boundary layers, pipes, and channels, in the case of both smooth and rough walls.
Wall-bounded turbulence has been tackled over the years along different fronts, which include theoretical analysis and experimental and numerical investigations. Regarding theory, it appears that the best established features of wall turbulence, including the presence of a logarithmic layer in the mean velocity profile and in the wall-parallel velocity variances can be explained within relatively well-established conceptual models. Important recent theoretical findings include the discovery that linear processes of transient growth may be responsible for the onset of self-sustained global modes in the wall layer.
Experimental techniques have also undergone major development in recent years, mainly with the introduction of high-resolution anemometry probes down to the nano-scale. This has allowed to shed light on such long-debated subjects as the presence of an outer peak in the streamwise velocity variance.
Computational experiments based on Direct Numerical Simulations (DNS) have lately become of widespread use to get insight into the physics of wall turbulence, because of the potential to access any ow property of interest. Also given the exponential growth of available computer power, DNS has reached Reynolds numbers comparable to those attained in experiments.
The course is manly addressed to doctoral students in mechanical and aerospace engineering and related subjects, but post-Doc fellows and young researchers are also warmly encouraged to attend. The course is intended to provide the audience with all fundamental notions about the structure of wall-bounded turbulent flows, but most classes will be devoted to advanced topics covering the freshest developments in the discipline, and to highlight paths for future investigation. Theoretical, experimental, and numerical issues will be covered. The course will be complemented by a short tutorial on DNS of turbulent flows. Attendants will be introduced to modern techniques of parallel computing, and will be made to exercise on sample channel and pipe codes to get familiar with important practical issues as mesh generation and data analysis.