I am a senior lecturer in the Turbulence, Mixing and Flow Control (TMFC) group, in the Department of Aeronautics at Imperial College London (ICL). I hold a PhD from the University of Poitiers in France in the field of Computational Fluid Dynamics (CFD) applied to turbulence. Before becoming a lecturer in 2014, I was a Research and Teaching Assistant at the P' Institute in Poitiers and a Research Assistant in the Department of Aeronautics at ICL.
Understanding turbulent flows and how to use them in various engineering applications is the motivation behind my research. In the last ten years, I successfully developed and used powerful numerical methods based on a Cartesian mesh to work on the turbulence problem, with a specialisation in High Performance Computing (HPC). I am a specialist in Direct and Large Eddy Numerical Simulation (DNS/LES) that can only be performed on the most powerful supercomputers in Europe. At ICL, and thanks to dCSE support from NAG/HECToR, I have fully parallelised my high-order fluid flow solver Incompact3d using an innovative domain decomposition technique that makes it one of the only few codes in the world that can run on hundreds of thousands computational cores. Another important element in my work is an Immersed Boundary Method (IBM) technique which can be used as a very useful strategy for the modelling of flows with solid geometries.
At Imperial College, I am currently working on Optimization techniques based on Machine Learning, on the active control of free shear flows and on moving Immersed Boundary Methods. With my colleagues in Porto Alegre, I am investigating particle-laden gravity currents.
2017-Present: Optimization techniques based on Machine Learning for drag reduction in turbulent boundary layers;
2015-Present: Modelling of Dielectric Barrier Discharge Plasma Actuators for Direct Numerical Simulations with the P' Institute in Poitiers;
2013-Present: DNS of the interaction of a wall-attached cube with a turbulent boundary layer J.C. Vassilicos (Imperial College) and BAE Systems;
2009-Present: Numerical dissipation via the viscous term for DNS/LES, with the P' Institute in Poitiers and J.C. Vassilicos (Imperial College);
2009-Present: Acoustic predictions using an incompressible DNS database with the P' Institute in Poitiers;
2010-Present: Turbulent jet controlled by fluidic actuators with the P' Institute in Poitiers;
2008-Present: Development and use of a virtual wind tunnel based on the high-order flow solver Incompact3d (www.incompact3d.com): parallelisation with up to one million computational cores, post-processing tools using virtual probes, virtual cameras and virtual microphones (http://www.2decomp.org/mpiio.html). In collaboration with NAG and STFC Daresbusry;
2006-2014: High-fidelity simulations of multiscale generated turbulence;
2008-2010: Proper Orthogonal Decomposition (POD) technique combined with DNS for a spatially-evolving mixing-layer;
2007-2010: Improvement of Particle Image Velocimetry (PIV) filtering using a DNS database, with B. Ganapathisubramani (Southampton, UK) and O. Buxton (Imperial College);
2010-2015: Numerical study of the mixing of a passive scalar in the presence of a mean gradient for grid-generated turbulence with J.C. Vassilicos (Imperial College);
2012-2013: Superfluid turbulence, with A. Baggalew (Newcastle, UK)
Laizet S, Ioannou V, Margnat F, A diagnostic tool for jet noise using a line-source approach and Implicit Large-Eddy Simulation data, Comptes Rendus Mécanique, ISSN:1631-0721
et al., Three‐dimensional turbulence‐resolving simulations of the plunge phenomenon in a tilted channel, Journal of Geophysical Research: Oceans, ISSN:2169-9275
et al., 2018, Coarse large-eddy simulations in a transitional wake flow with flow models under location uncertainty, Computers and Fluids, Vol:168, ISSN:0045-7930, Pages:170-189
Ioannou V, Laizet S, 2018, Numerical investigation of plasma-controlled turbulent jets for mixing enhancement, International Journal of Heat and Fluid Flow, Vol:70, ISSN:0142-727X, Pages:193-205
et al., Multi-Fidelity Uncertainty Quanti cation Using RANS and DNS, CTR Stanford Summer Program