Large-Eddy Simulation and its potential
Further improving the environmental performance of powertrains requires pushing back the prediction limits of simulation tools. Over the last few decades, Large-Eddy Simulation (LES) has attracted a great deal of interest, not only in academic research, but also increasingly for specific industrial flows, for which a time-resolved, non-statistical prediction of large flow scales is essential to capture key aspects and ensure more reliable prediction. These phenomena include mixing in turbulent flows with concentration or density/temperature stratifications, strong unstable and non-periodic features, and more generally flows in which the assumption of isotropic and possibly homogeneous turbulence described by a statistical approach is not appropriate.
Initially developed within the framework of the Navier-Stokes (NS) approaches, the key principle of LES – to precisely resolve large flow scales and to model the effects of small unresolved scales – is also at the basis of other approaches for calculating fluid dynamics which are gaining interest for present and future applications such as Lattice-Boltzmann methods (LBM) or Smoothed Particle Hydrodynamics (SPH).
Scope of LES4ECE
From a fundamental point of view, although flows inside ICE and EM or other powertrain devices exhibit some fundamental differences, they share common key features that can be addressed using very similar approaches and codes, making research originally addressing ICE also relevant to address other powertrain devices.
In terms of specific features, ICE flows require addressing high-speed turbulent aerodynamics in comparatively large volumes, high-pressure fuel injection, break-up and mixing, as well as the interactions between turbulence and chemical reactions. EM are subject to turbulent non-reactive flows in small passages, and can exhibit low pressure two-phase phenomena and wall films in the case of liquid cooling.
A common feature for flows inside ICE, EM and other electrified powertrain components is the heat exchange between a turbulent, single- or two-phase flow and complex walls for cooling or heating purposes. Addressing the underlying Conjugate Heat Transfer (CHT) problem requires comparable modelling and methodological approaches. A core scientific challenge is the significant difference in time and space scales between the fast convection and conduction in the fluids, and the slower conduction in solids. This requires adapted CHT methodologies to yield reliable predictions at a reasonable computational cost which becomes a primary concern when combined with LES. Among the many physical phenomena to be taken into account in such situations are thermal boundary layers, the formation and evolution of liquid wall films, and phase changes or rheological effects in the fluids under the influence of high temperatures and intense heat fluxes.
Furthermore, past experience has shown that it is crucial in the context of LES to have appropriate experimental data to implement, validate and complement its predictions. Space- and time-resolved, quantitative experimental data is required for the validation and calibration of LES subgrid-scale models for turbulent convection and mixing, multi-phase phenomena, chemical reactions and combustion, thermal wall boundary layers, or heat radiation. Related experiments consist of academic set-ups that allow well-controlled and accessible conditions, while being representative of the real operation of practical interest.
Experimental data is also indispensable for providing local and global, time-resolved and averaged data required to support the implementation of LES methodologies to simulate the performance of a component or system under real operating conditions, and to validate and complemen t its predictions.
LES4ECE aims to include experimental techniques as in particular optical diagnostics (e.g. PIV, LIF, LIP, etc.) to characterize all aspects of fluid flows (velocities, mixing state, dispersion, liquid wall films, temperatures), as well as techniques to measure temperature and heat fkuxes at wall boundaries (thermocouples, LIP, IR sensing, etc.). It is only by combining LES and experiments that breakthrough can be achieved in terms of understanding and controlling complex flows and their interactions with the walls.
In summary, LES4ECE aims to cover this broad spectrum of high-fidelity simulations and experimental techniques required to deal with the complex flows found inside modern electrified powertrains, as they share a number of common challenges and require similar approaches. We strongly believe that the resulting cross-fertilization between these fields, and the exchanges between researchers and engineers working in them, is of high benefit to the community.