Project

Discipline

internal combustion engine; direct numerical simulation; turbulent boundary layer; autoignition

Lead
Lay summary
Internal combustion engine flows are turbulent, unsteady and exhibit high cycle-to-cycle variations. Turbulence generation mechanisms are multiple and their effects overlap in time and space, rendering the simple turbulence models currently used in industry inappropriate for the in-depth understanding of underlying mechanisms as well as for predictive purposes. Even currently emerging LES-models for engine flows are subject to uncertainties with respect to sub-grid scale models, particularly for processes in the proximity of walls, which are prominent in engine combustion chambers.    Experimentally, optical methods have provided valuable insight into turbulent engine flows during the last 30 years. Despite recent progress, inherent limitations with respect to the efficient scanning of the whole domain with sufficient spatial resolution down to the smallest flow scales and in the proximity of walls still persist. Features of the high frequency part of the turbulence kinetic energy and dissipation spectrum and the associated interactions with the larger scales can therefore be uncovered only to a limited extent through measurements, and with heavy experimental effort.   This project aims at performing very large scale direct numerical simulations, first of non-reacting and then of a few autoigniting cases with gaseous mixtures, in engine-like geometries employing a highly scalable and effiecient parallel, spectral element low Mach number code. We intend to investigate the following processes: • Evolution of turbulent flow parameters as a function of the initial conditions after intake valve closing (including swirl and tumble) and combustion chamber geometry (compression ratio, clearance height, bowl-in-piston depth and diameter) in gaseous, non-reactive cases • Evolution of the unsteady hydrodynamic and thermal boundary layer structure during compression and expansion for the above mentioned parameter variation • Autoignition and combustion of gaseous mixtures employing sufficiently detailed chemistry and to study the multiple interactions between flow and thermo chemistry including near-wall phenomena.    The results of these simulations can be used for multiple purposes. Firstly, they will shed light into the complex interactions between flow and thermochemistry in unsteady engine flows at time and space scales. Secondly, they can be used to validate, or identify the shortcomings of models used in conventional RANS or LES approaches for describing turbulent engine flows. Finally, they can be used to guide future experiments for mutual (cross-) validation.

Direct link to Lay Summary

Last update: 21.02.2013

Active participation

Title	Year

Number	Title	Start	Funding scheme

Internal combustion engine flows are turbulent, unsteady and exhibit high cycle-to-cycle variations. Turbulence generating mechanisms are multiple and their effects overlap in time and space, rendering simple the turbulence models currently used in industry inappropriate for both in-depth understanding of underlying mechanisms and predictive purposes. Even currently emerging LES-models for engine flows are subject to uncertainties with respect to sub-grid scale models, particularly for processes in the proximity of walls, which are prominent in engine combustion chambers.Experimentally, optical methods have provided valuable insight into turbulent engine flows during the last 30 years. Despite recent progress regarding the development of planar or 3-D PIV techniques with high temporal resolution, inherent limitations as to efficient scanning of whole fields with sufficient spatial resolution down to the smallest flow scales and in the proximity of walls still persist. Features of the high frequency part of the turbulence kinetic energy and dissipation spectrum and the associated interactions with the larger scales can therefore be uncovered only to a limited extent through measurements, and with heavy experimental effort.The work proposed here aims at performing very large scale direct numerical simulations, first of non-reacting and then of a few autoigniting cases with gaseous mixtures, in engine-like geometries. The work will employ a highly scalable and effiecient parallel, spectral element low Mach number code that has been developed to a large extent in our group. During the last few years the code has been successfully used in several reactive flow problems, including the very large scale simulations of turbulent autoigniting hydrogen-air jets. At the current stage, the code is capable of peta-scale simulations on the appropriate hardware.We intend to investigate the following processes:•Evolution of turbulent flow parameters as a function of the initial conditions after intake valve closing (including swirl and tumble) and combustion chamber geometry (compression ratio, clearance height, bowl-in-piston depth and diameter) in gaseous, non-reactive cases,•evolution of the unsteady hydrodynamic and thermal boundary layer structure during compression and expansion for the above mentioned parameter variation,•autoignition and combustion of “homogeneous” gaseous mixtures employing sufficiently detailed chemistry and to study the multiple interactions between flow and thermo chemistry including near-wall phenomena.The results of these simulations can be used for multiple purposes. Firstly, they will shed light into the complex interactions between flow and thermochemistry in unsteady engine flows at time and space scales. Secondly, they can be used to validate, or identify the shortcomings of models used in conventional RANS or LES (currently emerging) approaches for describing turbulent engine flows. Finally, they can be used to guide future experiments for mutual (cross-) validation.