Data Assimilation for Reacting Flows: PDE Constraints and Efficient Optimization

Speaker: Jonathan MacArt, University of Notre Dame

Abstract: Accurate prediction of turbulent flow remains a significant challenge in engineering and applied science. Reynolds-Averaged Navier–Stokes (RANS) simulations and Large-Eddy Simulation (LES) are accurate for many flows, though non-Boussinesq turbulence and unresolved multiphysical couplings often preclude predictive accuracy. In turbulent combustion, flame–turbulence interactions can lead to inverse-cascade energy transfer, which violates the assumptions of many RANS and LES closures. We survey the regime dependence of these effects using a series of high-resolution Direct Numerical Simulations (DNS) of turbulent jet flames, from which an intermediate regime of heat-release effects is apparent and associated with the hypothesis of an “active cascade.” Implications for physics-based turbulence closures are discussed. We introduce an adjoint-based data assimilation method to augment the RANS and LES equations using trusted (not necessarily high-fidelity) data. A Python-native flow solver is developed, leveraging differentiable-programming techniques, to enable automatic construction of the adjoint Navier–Stokes equations. Applications to canonical turbulent flows, turbulent combustion, laser-induced ignition, aerodynamics, and flow control are discussed.

Bio: Jonathan F. MacArt is an assistant professor in the Department of Aerospace and Mechanical Engineering at the University of Notre Dame. He received his undergraduate degree in Aerospace Engineering from Notre Dame and his doctoral degree in Mechanical and Aerospace Engineering from Princeton University. His research interests lie at the intersection of multiphysical turbulent flows, optimization theory, and numerical methods, and focus on closure models for turbulent reacting flows.