RIACO · RIsk Aversion (turbulence) COntrol

The ultimate goal of turbulence control is to reduce the big energy losses in human and fluid transportation due to turbulent montions, that currently account for more than 10% of the total. However, turbulence control is extremely challenging. First, one needs to determine the best moment/location (flow state) where to actuate. Second, one needs to identify the best way to actuate. This is limited by the capacity to measure a (or to actuate on) a flow, and also by the inherent uncertainty of turbulent flows: the higher the uncertainties are, the more unpredictable the control action becomes. Thus, there is a need to study the effect uncertainties have on the controllability of turbulent flows, and, in general, to design a tool capable of finding optimal robust actuations. The RIsk Aversion COntrol project aims at this. The project proposes a cost-efficient numerical methodology based on non-linear adjoint optimization and implemented in fast GPU codes; that allows for the thorough exploration of a vast range of robust actuations. In particular, the actuations are optimized so they are able to maximize/minimize a desired functional, for an ensemble of turbulent trajectories simultaneously. Two different applications will be numerically and experimentally studied: minimization of wall friction, and maximization of heat transfer in wall-bounded turbulent flows. As a result, the identification of potentially new robust actuations is expected. Additionally, for the first time, the limits of controllability of turbulent states: corresponding to the maximum magnitude of the uncertainties where effective robust controls can be identified, will be investigated. Overall RIACO will set-up the stage for the analysis of the link between uncertainty and control; motivate the identification of flow structures that are fundamentally related with controllability; and inspire the design of more efficient and robust actuators for turbulent flows.