NRAResearch Opportunities in Space and Earth Sciences - 2015 (ROSES-2015)

Project: Research project

Project Details


The objective of the proposed research is to develop a novel Large Eddy Simulation (LES) model capable of predicting the liquid fuel atomization process under realistic aircraft engine operating conditions, i.e., high Reynolds and Weber number flows. We propose a hybrid Eulerian/Lagrangian approach that combines a Eulerian Volume of Fluid interface capturing approach in the primary atomization region with a Lagrangian approach in the secondary atomization region. The objective of the proposed approach is a model free of adjustable tuning parameters. To that end, a novel dual scale LES approach is proposed to predict the primary atomization process close to the fuel injector. Instead of relying on the traditional LES modeling paradigm that dynamics on the unresolved sub-filter scale follow a certain universality and can be inferred solely from the dynamics on the resolved scales, the dual scale approach aims to maintain a representative fully resolved realization of the liquid/gas phase interface geometry. This not only provides direct closure models for the filtered terms in the Navier-Stokes equation but also, more importantly, allows for direct capturing of the small scale phase interface dynamics during breakup, avoiding any phenomenological ad-hoc models for the primary atomization process on the sub-filter scale or the questionable assumption of universality of the phase interface dynamics on the sub-filter scale. Instead, broken-off individual liquid structures and drops/blobs with their position, momentum, and mass are the outcome of the LES model for the primary atomization process and are interactively coupled to a Lagrangian spray model using a novel shape aware parallel drop transfer algorithm.

These liquid drops (blobs) need further accurate modeling of the drop dynamics at the subgrid scale level. A Lagrangian approach here is advantageous as it allows for explicit modeling of various aspects of the drop dynamics important for secondary atomization and droplet dispersion. A two-parameter stochastic breakup model will be employed for secondary atomization at large Reynolds and Weber numbers representative of realistic aircraft fuel injectors. In addition, near the injector, in the dense spray regime, the droplet-droplet interactions, droplet deformation and internal circulation effects on the drag and lift forces can be important. It is thus necessary to improve upon the standard point-particle, two-way coupling based subgrid scale models for spray dynamics. Accounting for the fluid mass displaced by the droplets is important and requires reformulation of the Eulerian fluid phase governing equations to include carrier phase volume fraction variations.

To make the developed models applicable to realistic aircraft injector geometries a fully unstructured flow solver will be used that has proven successful in the past in simulating the flow in aircraft engines. The to be developed models will be verified using a hierarchy of Direct Numerical Simulations of ever more complex atomizing flows, culminating in the simulation of a realistic high shear fuel injection nozzle geometry proposed by United Technologies Research Center consisting of six liquid jets injecting into a swirling inner crossflow and further atomizing with an outer swirling airflow. Both high resolution DNS type simulations and experimental measurements will be used to verify/validate the new LES model in this realistic fuel injector geometry.

Effective start/end date1/1/1612/31/20


  • NASA: Glenn Research Center: $723,160.00

Fingerprint Explore the research topics touched on by this project. These labels are generated based on the underlying awards/grants. Together they form a unique fingerprint.