Thursday, April 19, 2018
Dr. Ugo Piomelli
Queen's University
2:00 - 3:00pm
SEH, B1220
Abstract
Roughness is present in many applications in engineering, meteorology, and the geophysical sciences, and its effects on the fluid flow have been studied for almost a century. Early studies measured only the drag (resulting, for instance, in the well-known Moody diagram); more recently, turbulence statistics have been collected in many geometries. It is very difficult and expensive, however, to measure the flow between the roughness elements; thus, most studies concentrate on the region above the roughness crest, where similarity exists: the roughness determines the velocity scale that makes turbulent statistics collapse. Over the last decade, the development of efficient Immersed Boundary Methods has allowed the numerical simulation of flows over very complex geometries to become feasible. The increase in available computational power, furthermore, has allowed the achievement of Reynolds numbers sufficiently high that the effects of roughness are significant while the roughness elements are small enough that the global characteristics of the flow are not affected. The access to the flow inside the roughness sublayer has helped obtain new insight on how roughness affects the dynamics of turbulence. Several examples will be presented, including studies of the flow over realistic surfaces, channel flows subjected to rotation, and accelerating and decelerating boundary layers. These examples will highlight the importance of the wake field (the steady spatial disturbance of the time averaged field) in generating turbulent fluctuations, how the destabilizing effects of roughness interact with the stabilizing effects of acceleration or rotation, how separation is modified when roughness is present, and the implications of these findings for turbulence modelling.