Session A2: Reacting Flows: Spray, Atomization, and Droplets

Validity of a LES/flamelet approach to a transcritical O2/H2 jet flame

A large-eddy simulation (LES) employing a flamelet/progress-variable approach is applied to a transcritical oxygen (O2)/hydrogen (H2) jet flame, and the combustion mechanism is investigated. The LES is performed using an in-house thermal flow analysis code FK3. The flamelet library is created using the one-dimensional analysis version of FK3, i.e. FK3/1d, employing a chemical reaction mechanism of 8 species and 21 reactions at the pressure of 10 MPa. For both the calculations of LES and flamelet library, the SRK equation of state and the TRAPP method are used to take the real gas effects into account. The results show that in the present transcritical combustion, oxygen and hydrogen, which are issued into the chamber in the liquid and supercritical states, respectively, react in the gas state. Here, the oxygen experiences the supercritical state before reaching the gas state. It is also observed that the present LES can well capture the liquid oxygen core breakup.   

Diffuse interfacelets in transcritical flows of propellants into high-pressure combustors

Rocket engines and new generations of high-power jet engines and diesel engines oftentimes involve the injection of one or more reactants at subcritical temperatures into combustor environments at high pressures, and more particularly, at pressures higher than those corresponding to the critical points of the individual components of the mixture, which typically range from 13 to 50 bars for most propellants. This class of trajectories in the thermodynamic space has been traditionally referred to as transcritical. Under particular conditions often found in hydrocarbon-fueled chemical propulsion systems, and despite the prevailing high pressures, the flow in the combustor may contain regions close to the injector where a diffuse interface is formed in between the fuel and oxidizer streams that is sustained by surface-tension forces as a result of the elevation of the critical pressure of the mixture. This talk describes progress towards modeling these effects in the conservation equations.   

Modelling flash-boiling atomisation with the homogeneous relaxation model implemented in a fully compressible solver

Sudden depressurisation of superheated liquids through nozzles is a major challenge. This pressure drop together with the rapid phase change of the liquid are important characteristics of flashing. The resulting jet usually emerges to the low-pressure region with a high velocity and fragments to large blobs and ligaments and then droplets due to both mechanical and thermodynamic effects. The present study presents a numerical approach for simulating the atomisation of flashing liquids accounting for the distinct stages, from primary atomisation to secondary break-up to small droplets using the Eulerian-Lagrangian-Spray-Atomisation model coupled with the homogeneous relaxation model. The proposed approach has the advantage of avoiding the unrealistic common assumption of pure liquid at the nozzle exit. It models the change in the regime inside the nozzle treating flashing in a unified approach simulating the metastable jet both inside and outside the nozzle. Important mechanisms such as thermal non-equilibrium, aerodynamic break-up, droplet collisions and evaporation are modelled in a novel atomisation model. Results for turbulent flows for both subcooled and superheated liquids are presented showing that the proposed approach can accurately simulate the primary atomisation.   

Multiphysics control of a two-fluid coaxial atomizer supported by electric-charge on the liquid jet

We present an experimental setup to investigate multiphysics control strategies on atomization of a laminar fluid stream by a coaxial turbulent jet. Spray control (i.e. driving the droplet size distribution and the spatio-temporal location of the droplets towards a desired objective) has many potential engineering applications, but requires a mechanistic understanding of the processes that control droplet formation and transport (primary and secondary instabilities, turbulent transport, hydrodynamic and electric forces on the droplets, \textellipsis ). We characterize experimentally the break-up dynamics in a canonical coaxial atomizer, and the spray structure (droplet size, location, and velocity as a function of time) in a series of open loop conditions with harmonic forcing of the gas swirl ratio, liquid injection rate, the electric field strength at the nozzle and along the spray development region. The effect of these actuators are characterized for different gas Reynolds numbers ranging from 10$^{\mathrm{4}}$-10$^{\mathrm{6}}$. This open-loop characterization of the injector will be used to develop reduced order models for feedback control, as well as to validate assumptions underlying an adjoint-based computational control strategy. This work is part of a large-scale project funded by an ONR MURI to provide fundamental understanding of the mechanisms for feedback control of sprays.   

Unsteady flamelet modelling of spray flames using deep artificial neural networks

We investigate the applicability of the tabulated, multidimensional unsteady flamelet model and artificial neural networks (TFM-ANN) to lifted diesel spray flame simulations. The tabulated flamelet model (TFM), based on the widely known flamelet assumption, eliminates the use of a progress variable and has been shown to successfully model global diesel spray flame characteristics in previous studies. While the TFM has shown speed-up compared to other models and predictive capabilities across a range of ambient conditions, it involves the storage of multidimensional tables, requiring large memory and multidimensional interpolation schemes. This work discusses the implementation of deep artificial neural networks (ANN) to replace the use of large tables and multidimensional interpolation. The proposed framework is validated by applying it to an n-dodecane spray flame (ECN Spray A) at different conditions using a 4 dimensional flamelet library. The validations are then extended for the simulations using a 5-dimensional flamelet table applied to the combustion of methyl decanoate in a compression ignition engine. Different ANN topologies, optimization algorithms and speed-up techniques are explored and details of computational resources required for TFM-ANN and the TFM are also presented. The overall tools and algorithms used in this study can be directly extended to other multidimensional tabulated models.   

Assessment of swirl spray interaction in lab scale combustor using time-resolved measurements

Liquid fuel injection in highly turbulent swirling flows becomes common practice in gas turbine combustors to improve the flame stabilization. It is well known that the vortex bubble breakdown (VBB) phenomenon in strong swirling jets exhibits complicated flow structures in the spatial domain. In this study, the interaction of hollow cone liquid sheet with such coaxial swirling flow field has been studied experimentally using time -resolved measurements. In particular, much attention is focused towards the near field breakup mechanism (i.e. primary atomization) of liquid sheet. The detailed swirling gas flow field characterization is carried out using time -resolved PIV (\textasciitilde 3.5 kHz). Furthermore, the complicated breakup mechanisms and interaction of the liquid sheet are imaged with the help of high-speed shadow imaging system. Subsequently, proper orthogonal decomposition (POD) and dynamic mode decomposition (DMD) is implemented over the instantaneous data sets to retrieve the modal information associated with the interaction dynamics. This helps to delineate more quantitative nature of interaction process between the liquid sheet and swirling gas phase flow field.   

Effects of Nanoparticulate Additives on Acoustically Coupled Fuel Droplet Combustion

The present study investigates interactions between applied acoustic perturbations and burning ethanol droplets containing nano particulate additives. Reactive nanoscale aluminum (nAl) as well as inert silica (nSiO2), each with an 80 nm average diameter. Continuously-fed fuel droplet combustion experiments were conducted in the vicinity of a pressure node created in a closed acoustic waveguide, with a range of applied forcing frequencies, pressure or velocity perturbation amplitudes, and particle loading concentrations. Simultaneous phase-locked OH* chemiluminescence and high-speed visible imaging enabled quantification of the influences of nanoparticle concentration on burning rate constant K and combustion-acoustic coupling. Results indicated that nAl particles in ethanol yielded measurable increases in K with increasing applied perturbation amplitudes, as compared to pure ethanol in the presence of acoustic excitation. Droplets with nAl exposed to moderate acoustic excitation exhibited sustained combustion for much longer periods of time than for unforced conditions. Post analysis of particulate matter collected from residue via electron microscopy aids in interpreting these trends and findings.   

Effects of Energetic and Inert Nano Particles on Burning Liquid Ethanol Droplets

This study explores the effects of nano particulate additives on ethanol fuel droplet combustion in a quiescent environment. Two different types of droplet combustion experiments were performed: one involving the classic single droplet suspended from a quartz fiber and the other involving a burning droplet that has continual fuel delivery via a quartz capillary. Two alternative nano particles were explored here to demonstrate the effect of energetic additives: reactive nano aluminum (nAl) and inert nano silicon dioxide (nSiO2), each with average diameter 80 nm. Simultaneous high speed visible and OH* chemiluminescence images were taken to determine burning rate constants (K) and to study flame and droplet characteristics with varying particulate concentrations. Particle/vapor ejections were seen in continuously fed droplet experiments, while rod-suspended burning droplets showed limited particle ejection. The nSiO2-laden, rod-suspended droplets formed a porous, shell-like structure resembling the shape of a droplet at higher nSiO2 concentrations, in contrast to smaller residue structures for nAl-laden droplets. Changes in K depended on concentrations of nAl and nSiO2 as well as the method of droplet formation, and TEM images of particle residue revealed additional insights.