This project is concerned with the unsteady aerodynamics and associated sound production mechanisms which result from flows around bluff bodies. Such systems comprise regions of fully separated turbulent flow and strong fluid-structure interaction. From an applied perspective, the motivation for studying such flows derives from clear societal needs (safety, chemical and noise pollution) and strong industrial competition, while from a fundamental point of view such flows present a real challenge to scientists working in the fields of aerodynamics and aeroacoustics: a comprehensive understanding of these kinds of flow is hampered by the difficulty of quantitatively analysing the unsteady flow field and the mechanisms by which it drives sound fields (both internal and external). Experimentally, quantitative analysis approaches suffer from the difficulty of accessing the full space-time structure of the flow, and the fact that much of the essential aeroacoustic dynamic is below the noise floor of the measurement device. Numerical approaches on the other hand, while capable of providing a more complete spatiotemporal picture, struggle to resolve the finer details of the flow in near-wall regions, and are not well suited to supplying the fully converged statistics which are required for implementation of analysis tools which can help better understand the dynamics of the flow. The principal objective of the project is thus to develop integral analysis methodologies for study of the flows and source mechanisms evoked above. The strategy which we propose to follow in order to achieve this, and which constitutes an important originality of the project, involves the association of experts from different fields (aerodynamic, aeroacoustic, numerical, experimental, theoretical). Such a multi-disciplinary initiative is necessary to obtain analysis tools adapted to the very large data bases generated by experiments and computations and is central to an understanding of the more subtle aspects of these flows. Three complementary model problems will be studied: (i) a massive two-dimensional separation generated by a thick plate [LEA-C1], (ii) a strongly three dimensional cavity flow [LIMSI-C2], (iii) a more complex three-dimensional separation involving a conical vortex interacting with a solid surface, which is of interest on account of the particular instabilities which it supports, and its capacity to act as a wave-guide for intermediate-scale perturbations [LEA – C3]. The three configurations will also be simulated by means of a number of complementary methods: Large Eddy Simulation (or DNS in C1) [LIMSI C1 + C2; PSA C3] and hybrid RANS/LES [LEA C1+C3]. Databases corresponding to C1 and C2 will be available from the project outset. The project will comprise two workpackages. The first will be dedicated to a direct analysis of the unsteady flows generated by the three configurations, and the developement of specific quantitative analysis tools. Further simulations and experiments will be performed during the course of the project, in order to complement those which currently exist, and to aid in the development of novel analysis tools. These will include Quantitative Topological Analysis, Lagrangian Coherent Structure tracking, Linear and Quadratic Stochastic Estimation, Extended Proper Orthogonal Decomposition, and Causality Correlation Analysis; and they will be largely based on synchronous sampling of pressure (in-flow, surface and farfield; experimentally obtained via arrays of unsteady pressure probes), and velocity via full-field and temporally resolved optical measurement tehniques (Stereo PIV and 3C LDV respectively). The objective will be to develop integral analysis methodologies for the extraction and tracking of flow events, important either in terms of their energy or their unsteady wall pressure signatures. The second workpackage will deal with the question of how the unsteady flow dynamic couples both with the model body and with the acoustic farfield. Our principle objective will be to understand how to pose the problem such that the source terms we generate experimentally and numerically are both amenable to physical understanding (for the wall region and the farfield), and robust enough to provide an accurate description of the most important flow/`source' events where the vehicle body and the acoustic farfield are concerned. The experimental and numerical databases generated for C1, C2 & C3 will serve to help us understand how the flow skeletons identified in workpackage 1 drive the near and farfield pressures. This ambitious project promises to be rich in fundamental and applied developments, thanks to the synergy of recent numerical, experimental and analysis techniques, and the association of experts in aerodynamics and aeroacoustics. Such a multidisciplinary fusion will ensure a dynamic research environment, necessary for and conducive to the generation of new scientific knowledge.

The Spark Ignition (SI) engine represents 70% of light duty vehicles worldwide and should still represent 50% in 2030. For this reason, reducing CO2 emissions of SI engines is of primary importance to contribute mitigating the global warming. For this purpose, the technology favored today by engine manufacturers is downsizing, which consists in reducing the displacement and increasing the specific power by using a turbocharger. The fuel saving potential of up to 20% offered by this technology is however limited in practice due to an increased occurrence of abnormal combustions (knock and super-knock) which lead to using sub-optimal spark timings. A key measure for limiting the occurrence of abnormal combustion is to increase the EGR (Exhaust Gas Recirculation) rate from presently 5% to reach values as high as 20% and even 30%. This allows substantially reducing abnormal combustions but leads to larger cycle to cycle variability and decreased heat release rates. In order to reach such high EGR rates, complex strategies have to be developed (aerodynamics, injection targeting etc…), the design and optimization of which increasingly rely on Computational Fluid Dynamics (CFD). Recent research clearly showed that existing combustion models fail for EGR rates larger than 10% in downsized SI engines. MACDIL proposes acquiring an unprecedented understanding of combustion under intermediate (15 to 25%) and high (beyond 25%) EGR rates and under pressure and temperature conditions representative of turbocharged SI engines, and to capitalize it in the form of both LES and RANS models integrated respectively in the reference LES research AVBP and the industrial RANS code CONVERGE. The major scientific challenges addressed by MACDIL to reach its ambitious objectives is the absence of sufficient knowledge on turbulent flames under such extreme conditions, the lack of adapted chemical kinetics, as well as the absence of experimental studies due to the practical difficulty with high pressure and temperature flame measurements. Concerning experiments, a unique set-up called NOSE (New One Shot Engine) will allow for the first time to study diluted combustion, both at low pressure and temperature in order to support LES model development, and at high pressure and temperature conditions for providing validation data for the developed models. This experiment will provide flame visualizations and a global comparison with CFD (in terms of pressure) but it won’t give access to the flame surface properties. For this reason, Direct Numerical Simulations (DNS) of a reduced scale NOSE configuration will be performed to provide detailed local flame statistics for orienting and supporting the LES combustion model development. The resulting LES models will first be validated against NOSE data, before being evaluated on an existing downsized research engine. Experimental measurements of the local effective flame speed at such conditions are also impossible. MACDIL thus proposes exploiting ab-initio calculations for formulating chemical schemes. The latter will be used to generate tables of planar and effective laminar flame speeds which constitute key input parameters in the combustion models to be developed. For simplicity, the experiments and CFD simulations of MACDIL will be conducted with isooctane as a fuel and N2 as a diluent. But kinetic modeling will also consider a surrogate fuel of gasoline and real EGR (including CO2 and H2O). As a final step, the combustion models developed within MACDIL will be adapted for RANS, and integrated into the industrial CFD code CONVERGE. This will allow the industrial partners to apply the MACDIL models to real downsized engines of interest to them, to evaluate their performance under high dilutions, and thus assess the benefit they offer for their future engine developments.