Multi-Scale Study of Reactive Engine Spray Gauss Centre for Supercomputing e.V.


Multi-Scale Study of Reactive Engine Spray

Principal Investigator:
Heinz Pitsch

Institute for Combustion Technology, RWTH Aachen University, Germany

Local Project ID:

HPC Platform used:
Hazel Hen of HLRS

Date published:

Propulsion based on liquid fuels provides a large fraction of today’s transportation energy. Since alternative technologies, such as electric mobility, are not able to substitute propulsion technology based on liquid fuels within the next decades, processes of conventional propulsion such as injection of a liquid fuel into a gaseous environment and energy conversion in a turbulent environment need to be improved in order to deal with stricter emission regulations and the finiteness of fossil fuels. Additionally, the usage of alternative fuels can potentially overcome classical limitations of conventional energy conversion machines. For example, it was found that synthetic fuels, such as oxymethylene ethers, have the potential to burn soot-free and thus eliminate the well-known soot-NOx tradeoff of classical diesel engines [Omari et al., Fuel, 2017]. Also, CO2 emissions are potentially lower than in gasoline engines and an efficient production from renewable resources seems possible leading to a closed CO2 loop. However, the optimization of these processes is very challenging and details are still not completely understood due to their complexity and the difficulty of performing experiments characterizing the flow features inside the nozzle and the atomization and chemistry processes outside the nozzle. The performance of a particular injection system/fuel combination depends on a cascade of physical processes, originating from the nozzle internal flow driven by the injector geometry and operation conditions, potential cavitation, turbulence, the mixing of the coherent liquid stream with the gaseous ambient environment during the atomization process outside of the injector orifice, and the chemical properties of the fuel.

Beside the numerical problems of such simulations descending from the discontinuities at the interface, complex boundary conditions, and ill-conditioned systems of equations, the large range of relevant scales in an injection system is a major issue preventing the realization of simulations governing both nozzle internal flow and atomization/ignition process. Coupled simulations can reduce the computational cost by simultaneously performing simulations with different resolutions and considered physics.

As successfully demonstrated by the authors [Bode et al., SAE Technical Paper, 2015], the usage of a large-eddy simulation (LES) for the nozzle internal flow, a direct numerical simulation (DNS) for the primary breakup, and a Lagrangian particle based LES (LP-LES) [Davidovic et al., Oil & Gas, 2017] describing the far-field spray evolution is a promising solution for the coupling and enables accurate simulations on current supercomputers.

The Spray A and Spray B injectors specified by the Engine Combustion Network (ECN) were used as target injectors for the coupled simulations performed on supercomputer Hazel Hen at the High-Performance Computing Center Stuttgart (HLRS) in this work. While simulations of the Spray A injector, which features only one hole, were used to develop and validate LES models for soot, the Spray B injector with three holes were used to optimized injectors for synthetic fuels. Optimizations with respect to nozzle geometry, injection conditions, and fuel properties were realized, which help to use synthetic fuels more efficiently in future propulsion systems and reducing anthropogenous CO2 emissions.

Figure 1 shows primary breakup results for an injection with oxymethylene ether (OME1). Formation of small droplets but also ligaments can be seen. Figure 2 shows the instantaneous flow field at the end of a n-dodecane injection. In the upper picture, the liquid droplets, an iso-line of stoichiometric mixture, and the soot volume fraction are shown. It can be seen that soot is formed in rich regions, which are inside of the mixture fraction iso-line. The highest soot volume fraction is observed in the spray head, where the residence times are the longest. In the lower picture of Figure 2, a 3-D temperature iso-surface of the value 1800 K is shown that corresponds to the occurrence of high temperature chemistry. Additionally, the experimentally determined time-averaged flame lift-off length (FLOL) is illustrated by the black arrow. The FLOL is the distance from the injector orifice to the location, where a stable flame establishes after ignition.

The project was awarded with the Golden Spike Award 2018 from the High-Performance Computing Center Stuttgart (HLRS).

Project Team

Mathis Bode, Marco Davidovic, Heinz Pitsch (PI)
Institute for Combustion Technology, RWTH Aachen University

Scientific Contact

Mathis Bode
Fakultät für Maschinenwesen
Institut für Technische Verbrennung (ITV)
RWTH Aachen University
Templergraben 64, D-52062 Aachen (Germany)
e-mail: m.bode [@]

HLRS project ID: GCS-mres

September 2019

Tags: HLRS RWTH Aachen Large-Scale Project Computational and Scientific Engineering