MICHAEL BAUERHEIM (ISAE Supaero) Combining Artificial Intelligence and CFD: recent progress and challenges Artificial Intelligence (AI) has been widely employed in various fields such as medicine, engineering or physics. Yet, its application to fluid mechanics is still recent, and its full potential for fluid problems is still to be explored. This presentation will present the challenges to exploit deep learning techniques when combining with CFD. A particular focus is made on unsteady flow problems and optimization. Results show that the classical AI frameworks, which are freely available, must be adapted to the specificies of fluid problems (unstructured data, long term predictions, low-data regime etc.). The benefits of combining AI with CFD is finally shown through several applications, such as discovering optimal kinematics for low-Reynolds flapping wings, or the fluid-structure dynamics of a High Altitude Long Endurance (HALE) drone in a perturbed flow.

DOMINIQUE DEROME (Université de Sherbrooke) UNDERSTANDING THE LOCAL URBAN CLIMATE FOR MITIGATION AND ADAPTATION WITH EXPLICIT CFD-BASED MODELING The impact of climate change on the urban climate can be investigated at local scale using explicit modeling of heat and moisture exchanges in the air domain, in the porous materials lining the urban environment and in vegetation. In this keynote lecture, a full physical modeling approach will be presented to take into account the urban climate at local scale. The modeling suite, based on CFD and implemented in OpenFoam is presented. Studies on the understanding and analysis of the urban climate at city scale are presented, with focus the co-occurrence of urban heat islands and heatwaves and examples show the impacts of urban morphology on surface temperature and of buoyancy on urban ventilation. Using vegetation at the center of mitigation strategies, studies for Montreal streets and urban parks illustrate the benefits and opportunities offered by greening.