CoEC works on 11 state-of-the-art European HPC codes in combustion. The consortium is formed by the main developers of the codes and advanced users that pursue to increase the TRL of the codes for certain applications.
YALES2 aims at the solving of two-phase combustion from primary atomization to pollutant prediction on massive complex meshes. It is able to handle efficiently unstructured meshes with several billions of elements, thus enabling the Direct Numerical Simulation and Large-Eddy Simulation of laboratory and semi-industrial configurations. The recent developments are focused on the dynamic mesh adaptation of tetrahedral-based massive meshes for fronts and interfaces.
The AVIP code is devoted to the resolution of cold plasmas, as encountered in ignitors such as NRP (Nano-Repetitive Pulse Discharge). It is based on a fluid formulation taking into account the out-of-equilibrium nature of plasmas, and is coupled to a Poisson equation to solve the electric field (using libraries PETSC or MAPHYS) associated to sparking systems. The simulation of plasma requires the resolution of transport equations for electrons, ions and neutrals, including complex chemistry. AVIP is able today to compute 2D streamers and gives results in good agreement with the literature
JAGUAR is a new code, still under development, that aims to reach high accuracy on unstructured grids with high computing efficiency. The numerical approach that appears most promising for these objectives is Spectral Differences with possibility of h/p refinement. The code solves fully compressible, multi-species flow equations, reacting or non-reacting.
DISCO solves the fully compressible Navier-Stokes equations for reacting flows with detailed chemical kinetics on rectangular domains. It employs a third-order low-storage Runge-Kutta scheme for time integration and compact higher-order finite-difference schemes for the spatial derivatives. The code is parallelized using domain decomposition and MPI. The code has been used to simulate turbulent expanding flame kernels and mixing layers. Besides detailed reaction mechanisms, it can also use FGM tabulated chemistry.
Alya is a simulation code for high performance computational mechanics. Alya solves coupled multiphysics problems using high performance computing techniques for distributed and shared memory supercomputers, including GPU accelerators . Alya’s simulation capabilities include incompressible and compressible turbulent flows, non-linear solid mechanics, multi-species reacting flows, excitable media, heat transfer, n-body collisions and electromagnetic fields. The CASE Department at BSC works intensively on developing each of these modules in collaboration with industry and international partners around the world.
AVBP is a parallel CFD code that solves the three-dimensional, multi-species, multi-phase compressible flow equations on unstructured and hybrid grids and in the Large Eddy Simulation (LES) framework. Initially conceived for steady state flows of aerodynamics, today it has a wide range of application, from aerodynamics, aeroacoustics, combustion and reacting flows, heat transfer, particle-laden flows with phase change, etc. It is used for both research and industrial design.
PRECISE-UNS is a finite volume based unstructured CFD solver for turbulent multi-phase and reacting flows. It is a pressure-based code, which uses the pressure correction scheme / PISO scheme to achieve pressure velocity coupling. It is applicable to both low-Mach number and fully compressible flows. Discretisation in time and space is up to second order. The linearized equations are solved using various well-known libraries such as PETSc, HYPRE and AGMG. Several turbulence models are available: k-epsilon, k-ω-SST, RSM, SAS, LES. Different combustion models are available, ranging from the classical conserved scalar (flamelet) models and global reaction mechanism, to FGM and detailed chemistry. To model the interaction of chemistry and turbulence, EBU, Presumed PDF, ATF and Eulerian stochastic field PDF closures are available. To model emissions NOx and two equations and hmom soot models are available. In order to model liquid fuel, a Lagrangian spray model is available.
CIAO performs DNS and LES with multiphysics effects (multiphase, combustion, soot, spark, …). It is a structured, arbitrary order, finite difference code with compressible and incompressible/low-Mach solvers. Moving meshes are supported and overset meshes can be used for local mesh refinement. Spatial and temporal staggering is used to increase the accuracy of stencils. The sub-filter model for the momentum equations is an eddy viscosity concept in form of the dynamic Smagorinsky model with Lagrangian averaging along fluid particle trajectories.The compressible solver uses a low-storage five-stage, explicit Runge-Kutta method for time integration. The low-Mach solver uses Crank-Nicolson time advancement along with an iterative predictor corrector scheme. The Poisson equation for the pressure is solved by the multi-grid HYPRE solver. Momentum equations are spatially discretized with central schemes of arbitrary order, while for scalar equations various different schemes (WENO, HOUC, QUICK, BQUICK, …) are available. Temperature and species equations are advanced by utilizing a Strang operator splitting. The chemistry operator uses a time-implicit backward difference method (CVODE).
Nek5000 is a computational fluid dynamics code that employs the spectral element method, ahigh-order weighted residual technique, for applications in a wide range of fields including fluid flow, thermal convection, conjugate heat transfer, combustion and magnetohydrodynamics. It features state-of-the-art, scalable algorithms that are fast and efficient on platforms ranging from laptops to the world’s fastest computers. Nek5000, which is actively developed and improved for more than 30 years at Argonne National Laboratory (ANL), was extended for the direct numerical simulation of low Mach number reactive flows at the Swiss Federal Institute of Technology Zurich and is been used to investigate gas-phase and catalytic combustion in a number of laboratory-scale setups of fundamental and applied interest including internal combustion engines. Nek5000 won a Gordon Bell prize for its outstanding scalability on high-performance parallel computers and the 2016 R&D 100 Award. It is part of the Center for Efficient Exascale Discretizations (CEED) co-design effort, and its user community involves hundreds of scientists and engineers in academia, laboratories and industry.
The CLIO code solves the Conditional Moment Closure transport equations for reacting scalars. These equations are PDE’s in three space directions, one or two sample space directions (mixture fraction, progress variable, two mixture fractions etc), and time. A fractional step approach is taken for the transport in real space, transport in conserved scalar space, and chemistry. Various stiff solvers have been used such as LIBSC, DVODE, VODEPK,
LSODE, CHEMEQ, and various chemical schemes can be implemented. CLIO has been interfaced with various CFD codes in RANS and LES such as openFOAM, PRECISE_uns, FLUENT, and STAR-CD
OpenFOAM is a C++ object-oriented numerical simulation platform, which has a modular code-design suitable to be extended with new functionalities through additional libraries. OpenFOAM uses the unstructured grid formulation with a collocated cell-centred variable arrangement, which allows handling arbitrarily complex geometries and it can be applied in different fluid dynamic problems. For turbulent reacting and multi-phase flows, OpenFOAM provides a wide runtime-selectable flexibility in terms of turbulence models, e.g. RANS and LES, with several turbulence closures, and combustion models, such as tabulated chemistry with presumed PDF, ATF (TU Darmstadt in-house extension), and finite-rate chemistry.
Multi-phase problems as liquid sprays can either be treated using a fully coupled Lagrangian point-particle method or within an Eulerian framework as Volume of Fluid Method. OpenFOAM can be easily coupled with external libraries, as for example the Quadrature-based Method of Moments (QMOM) library developed at TU Darmstadt to account for the soot particle size distribution.
CoEC’s Exascale Challenge Demonstrators (ECD) are proofs of concept aimed at testing codes on Exascale hardware prototypes. Targeted simulation studies will evaluate the accuracy, reliability, and performance of the codes and, eventually, measure progress in terms of TRL increases.
Large-scale DNS calculation of formation, growth and transport of particulates
Objectives: To use the (pre-)Exascale capabilities to perform highly accurate simulations and gain a deeper understanding of the fundamental processes occurring in particle matter generation. This demonstrator will provide reliable data about particle size distributions with the aim of improving soot models and generating a useful database for further research.
Prediction of soot formation in practical applications
Objectives: To demonstrate the predictive capabilities of Exascale simulations to provide accurate results of soot formation when applied to large-scale simulations. This ECD will narrow the gap between simulations and experiments obtaining state-of-the-art soot models and showing that satisfactory degrees of accuracy can be achieved for the prediction of an extremely complex process such as soot formation in engines.
Detailed chemistry DNS calculation of gas phase pollutants: NOx and CO
Objectives: NOx and CO show slow chemical rates complicating their prediction in many models. This ECD will provide further understanding about the evolution of these chemical species in the flow. It is expected that useful information will be obtained from these simulations to improve combustion models.
Prediction of pollutants and design of low-emission burners
Objectives: To optimize burner performance in terms of pollutant emissions making use of large-scale simulations. Advanced combustion and soot models will be used to pursue this objective. This ECD will make tangible the potential of HPC as an important tool to increase reliability and accuracy in numerical simulations for practical applications with a strong industrial focus.
Detailed chemistry DNS calculation of turbulent hydrogen and hydrogen-blends combustion
Objectives: It is focused on the study of thermo-diffusive instabilities in turbulent lean hydrogen flames and its effects on burning velocities, unstable combustion and noise. The effect of preferential diffusion will also be investigated due to its influence on equivalence ratio fluctuations and eventually on the local burning velocity. This work will be extended to syngas and high hydrogen content (HHC) fuels.
Use of alternative fuels, H2 and H2 blends in practical systems
Objectives: The current ECD intends to shed light in the potential of alternative fuels to substitute conventional fuels with the current engine technology. In consequence, it is expected to address open questions about the technical aspects of the application of bio-, synthetic and HHC fuels and its blends to provide further understanding on burner operability, flashback and lean-blow out.
Fuel atomization and evaporation in practical applications
Objectives: It is considered a priority to contribute to the knowledge of liquid fuel injection and atomization due to its crucial influence on the combustion process and pollutants formation. This demonstrator includes the study of primary and secondary breakup, and the influence of heat conduction and droplet heating on the evaporation rates prior combustion takes place. The final objective will be the study of reacting sprays at relevant engine conditions.
Plasma assisted combustion
Objectives: The application of plasma in combustion simulations provides an unprecedented opportunity for combustion and emission control thanks to its capability to produce heat, active radical species and modify the transport properties of the mixture. This demonstrator is focused on the study of plasma-assisted combustion by Nanosecond Repetitively Pulsed (NRP) discharges in order to control the formation of combustion instabilities and pollutant formation.
Fuel ignition with high-energy sparks
Objectives: To develop modelling capability by high-performance computing for the multi-physics, multi-scale problem of spark ignition in reciprocating engines and gas turbines, and marine engines with pre-chamber. The Ignition with various alternative low-calorific value fuels will be evaluated to provide insights about the performance of the fuels in future engine technologies.
Combustion of metal particles
Objectives: Combustion of metal particles has gained attention in the last years and the capabilities to reduce pollutants formation and it is explored in this ECD. The main objective of this demonstrator is to study metal powder combustion where the flames interact with metal particles. The deflagration process has to be sustainable and able to heat and ignite the particles without the assistance of a continuous and combustible environment.
Objectives: The impact of flame-wall interactions on the design of the combustion chamber and unburned hydrocarbons emissions point to the importance of understanding flame-wall interactions. The intrinsic and extreme difficulty to experimentally measure such interactions evidence the relevance of DNS simulations in the frame of HPC, targeted in this ECD, in order to provide new insights and assist on near-wall model development.
Near-wall reacting flow modelling
Objectives: This ECD complements ECD11 and it is focused on the development of near-wall models that can be used in LES and RANS. The model generation will be conducted using physical models, and ML and ROM. Addition of variable composition and wall-roughness will be investigated to obtain more accurate and reliable predictions of heat fluxes and temperature and species distribution along the boundary layer.
ML and ROM in combustion simulations at relevant conditions
Objectives: ML and ROM have opened new frontiers in the path to understand complex physical phenomena. This ECD is expected to apply ML and ROM models in high-fidelity simulation to obtain more accurate predictions at reduced cost in turbulent flames.