Data-drivEn reduCed-Rank modEls to deSCribe flamE dyNamics anD their sOund – DECRESCENDO

In the context of the global transition to sustainable and carbon-neutral energy production paradigms, hydrogen combustion has emerged as a prominent candidate, particularly in the field of transport systems. However, the aeroacoustic ramifications of hydrogen flames, a critical aspect for the development of advanced burners, remain inadequately studied. DECRESCENDO addresses this gap through a comprehensive investigation of direct combustion noise in unconfined flames, using a multidisciplinary approach including experiment, high-fidelity simulation and simplified modelling. Building on the collaborative track record of the Principal Investigators, the research framework will establish a comprehensive database of validated high-fidelity numerical simulations. The research consortium comprises experts in high-fidelity simulation, combustion dynamics, turbulent flow analysis, acoustics, aeroacoustics, thermoacoustics, linear modelling and experimental diagnostics. The overarching objective is to unravel the intricacies of ultra-lean premixed turbulent hydrogen flames, with specific objectives structured to systematically explore combustion noise in unconfined systems. The first objective involves a detailed experimental characterization of hydrogen flames and their sound, utilizing advanced techniques such as microphone arrays in anechoic environments and Schlieren together with OH-CL flame imaging. This endeavor aims to comprehensively characterise the link between the flame organization and the far-field sound and their dependence on relevant flow parameters (Reynolds number, Damkohler number, and Lewis number). The second objective adopts an "outside-in" approach for heuristic sound-source modeling, positing that sound generation predominantly emanates from unsteady heat release associated with coherent structures. In parallel, the third objective takes an "inside-out" approach, leveraging high-fidelity data from DNS, LES, and experiments to identify acoustically relevant source features. This approach complements the "outside-in" methodology, providing a nuanced understanding of the spatiotemporal features of flame structures responsible for sound generation. The final objective entails physics-based modeling of mechanisms driving the flame dynamics responsible for sound generation, employing Linear Mean Field Analysis. The focus here is on elucidating instability mechanisms that underpin the flame dynamics, the fluctuations in heat release, and the ensuing acoustic radiation. These findings not only contribute to the development of data-driven sound-source models but also offer insight into the root causes of turbulence-driven combustion noise. Importantly, the sensitivities and optimal forcing structures derived from this approach guide future endeavors in noise mitigation strategies. DECRESCENDO's outcomes promise to advance the scientific understanding of combustion noise in unconfined hydrogen flames, thereby facilitating the development of next-generation burners and cleaner energy solutions.