Eulerian Modeling and Simulation of Moderately Dense Spray Flows: Application to Solid Rocket Motors
Speaker: François Doisneau, Sandia National Labs
Presentation
Seminar video is unavailable due to technical difficulties.
Abstract
Among two-phase flows, moderately dense sprays are suspensions which have a small enough volume fraction to remain disperse but still a high mass loading. This dense phase strongly conditions momentum and temperature of the carrier phase, at a rate which depends on droplet sizes. Moreover, droplets may collide and coalesce or break-up, bringing in a wide variety of sizes and dynamics, so-called polydispersity. Moderately dense sprays occur in mechanical engineering: in solid rocket motors, nozzle efficiency and instability occurrence rely on the aluminum oxide cloud that results from propellant combustion. Accurately solving for unsteady polydisperse moderately dense flows is a challenge for both modeling and scientific computing: (1) very small droplets, e.g. resulting from the combustion of nanoparticles of aluminum fuel, encounter Brownian motion and coalescence, (2) small droplets have their velocity conditioned by size so they coalesce when having different sizes, (3) bigger droplets have an inertial behavior and may cross each other's trajectory, and (4) all droplets interact in a two-way coupled manner with the carrier phase. In the prospect of high performance computing simulations and as an alternative to Lagrangian approaches, Eulerian models can describe the disperse phase at a moderate cost, with an easy coupling to the carrier phase and with massively parallel codes. For instance, the Multi-Fluid model allows the detailed description of polydispersity, size/velocity correlations and coalescence by separately solving fluids" of size-sorted droplets, the so-called sections: describing the size variable as a continuum is eefficient to deal with coalescence and break-up. I will present some aspects of the modeling and numerical strategy to perform industrial computations of moderately dense spray flows. From kinetic theory tools, we derive Eulerian models that are adapted to several droplet inertia regimes and to complex interactions such as coalescence and break-up. These models are reduced with high order moment methods to remain eefficient and accurate and dedicated schemes are developed to cope with the numerical peculiarities of such approaches, with the constant care of robustness and realizability. A time integration strategy is also designed for the overall set of equations to guarantee two-way coupling. Finally the methods are implemented either in research codes or in the industrial code CEDRE, developed at ONERA, to prove the feasibility and the predictability of the strategy.
Biography
François Doisneau is a postdoctoral fellow at Sandia National Labs specialized in numerical simulation of strongly coupled two-phase flows. After his Engineering and Master's degrees in Aerospace at École Centrale Paris, he received his Ph.D. in Applied Maths and Energy in 2013 from the same institution for his dissertation on "Eulerian modeling and simulation of nanometric to inertial particles in Solid Rocket Motors" under the supervision of Professor Marc Massot. As a visiting scholar at Stanford's Multiphase Flow Group and CTR Summer Program, Doisneau has developed high order size and velocity moment methods for coalescence of inertial drops. He has then worked one year as a research engineer in the Liquid Rocket Engine unit at Onera, to apply his developments to reacting liquid oxygen sprays. He is now developing models and methods and conducting massively parallel Large Eddy Simulations of high pressure fuel injection and combustion at Sandia's Combustion Research Facility.