ATELIER: SIMULATION NUMÉRIQUE POUR LES GROUPES MOTOPROPULSEURS 2 FÉVRIER 2017 SAINT-ETIENNE-DU-ROUVRAY 1
ENGINE AND VEHICLE MODELING & SIMULATION SCIENTIFIC AND TECHNICAL CHALLENGES 2
CONTEXT Main drivers for future mobility Social Demand: Air quality Global warming New standards and regulations: Measurements and control in real driving conditions Both pollutant and CO 2 emissions will be monitored Non regulated pollutants (in 2017) will be monitored Political drivers and Green incentives: Taxes Bonuses Circulation restriction measurements Sustainable mobility for everyone: Along the full vehicle line Affordable and competitive cost for customers 3 2016 I F P E N
CHALLENGES FOR FUTURE IC ENGINES Competitive against 100% electric mobility Growing hybridization: Powertrain electrification necessary to reach 2021 target (EU 95 g/km) Level and type of hybridization depend on usage Diesel engines: Keep their main advantage: High efficiency and low CO 2 emissions Improve efficiency Cope with pollutant emissions regulations at no extra cost Spark ignition engines: Improve efficiency at a reasonable cost Improve synergy with electrification 4 2016 I F P E N
NUMERICAL SIMULATION PRESENT AT ALL STAGES OF VEHICLE DEVELOPMENT Innovative concept proposal and pre-dimensioning Calibration of the engine control system: Prototype or industrial Design of all or part of the powertrain Development on test bed: Systems, energy, pollution control Optimization (combustion system, air and fuel loops, etc.) Validation: Proof of concept, robustness Complete engine manufacture 5 2016 I F P E N
ICE AND VEHICLE SIMULATION: STATE OF THE ART System Simulation Different levels of complexity: Physical models: High sensitivity to engine control variables Deep understanding of the calibration impact on combustion Reduced number of experimental data Semi-physical models: Mean-value models Limited number of calibration parameters Empirical models: Require a large set of data Fast-running Engine simulation: 0 0 20 40 60 80 100 Point number Conventional Diesel and SI engines in a large set of engine operating conditions Vehicle simulation: Driving cycles based on steady-state engine maps for hybrid powertrains After-treatment simulation: Full set of conventional after-treatment devices ( 6) 1800 1600 1400 1200 1000 800 600 400 200 NOx PPM Corrected dry NOx [ppm] 12G Expe 6 2016 I F P E N
SYSTEM SIMULATION: CHALLENGES Internal Combustion Engine High Frequency heat release models: Physical based models able to handle RD transient conditions, relying on cylinder pressure sensors On-the-fly mathematical models (learning experience) Generalized use of 1D CFD Spark-Ignition engines: Air-path: VVT, VVL, EGR, 2-stage supercharging Increased CR and VCR Combustion modes: Stratification, lean-burn, GCI Diesel engines: Full operation EGR Injection: Very high injection pressure, generalized multi-injection Pollutant formation: Soot formation and oxidation, NOx Transient data transfer from Engine to Vehicle models 7 2016 I F P E N
SYSTEM SIMULATION: CHALLENGES Vehicle Hybridization & Electrification Batteries: Evolution of current Li-ion technologies both in positive and negative (medium term) electrodes Post-Lithium-ion technologies (longer term) Electro-chemical models for battery ageing Electric motors Maximum CO 2 potential for a given solution Optimal hybrid powertrain architecture for a specific application: Optimal energy management laws and maximum CO 2 potential Fast screening of numerous configurations Automatic generation of energy management laws 8 2016 I F P E N
SYSTEM SIMULATION: CHALLENGES Exhaust gas after-treatment Diesel engine complex after-treatment systems: DOC exhaust line + SCR-F + ASC DOC + NSC + SCR-F + ASP De-Sox Spark-ignition engines: TWC TWC + GPF (including all-in-one option) Low temperature HC storage Lean burn TWC + NSC & SCR Current unregulated pollutants (future regulations?) Generalized OBD: NOx Soot particles RDE conditions ASC : Ammonia Slip Catalyst DOC : Diesel Oxidation Catalyst GPF : Gasoline Particulate Filter NSC : NO x Storage Catalyst SCR : Selective Catalytic Reduction TWC : Three Way Catalyst NO X cumulative mass flow [mg] 1600 1400 1200 1000 800 600 400 200 Bench - Inlet SCR Bench - Outlet SCR Model - Outlet SCR 0 0 200 400 600 800 1000 1200 Time [s] 9 2016 I F P E N
ICE AND VEHICLE SIMULATION: STATE OF THE ART 3D CFD Simulation Past major challenge: To reduce setup time of a 3D CFD engine simulation Today, solutions are available to move from a meshing CFD engineer to an engine CFD engineer Conventional Diesel and SI engines combustion simulation: Heat release; abnormal combustions (SI engines); first steps of CHT coupling Quantification of NOx and CO; qualification of UHC and soot vol fraction trends Fuel effects: Fuel discrimination (physical and chemical properties) is not straight forward Chemical mechanisms available; performance difficult to evaluate in engine conditions RANS and LES: RANS solvers widely available and easy to use LES solvers proposed by majors but hardly adapted (based on RANS solvers) In general, limited in-house computing resources 10 2016 I F P E N
ICE AND VEHICLE SIMULATION: CONVENTIONAL DIESEL COMBUSTION STATE OF THE ART Full 3D Diesel simulation BowlA BowlB BowlA BowlB NOx Soot 11 2016 I F P E N BowlA BowlB BowlA BowlB
ICE AND VEHICLE SIMULATION: CONVENTIONAL SI COMBUSTION STATE OF THE ART 1200 rpm / 6 bar 3000 rpm / 6 bar Cycle to Cycle Variations Eq. Ratio Eq. Ratio 1200 rpm / 20 bar 4000 rpm / 19 bar Abnormal Combustions RANS Simulations AVA Ref+7 AVA Ref LES Simulations 12 2016 I F P E N Knock Intensity
3D CFD: CHALLENGES RANS Simulations Liquid fuel injection and wall interaction: Predictive liquid injection and wall interaction models based on Eulerian/Lagrangian approaches In-nozzle simulations essential to provide appropriate boundary conditions Combustion and heat release: Validated fuel chemistry to take fuel effects into account (reduction as needed or tabulated chemistry) Low temperature, lean burn, chemical additives Appropriate turbulence/chemistry interaction models for all combustion modes (very diluted premixed flames, dual-fuel, homogeneous ) Models for future spark plug technologies (high dilution, lean burn ) Generalized CHT Fast and robust numerical methods, new materials Pollutant emissions: Soot formation and oxidation models including fuel effects, soot mass, number and size distribution Non regulated pollutants, including NOx and UHC speciation 13 2016 I F P E N
3D CFD: CHALLENGES LES Simulations Physical models (injection and combustion): Same as in RANS But adapted to LES Turbulent structures available up to a filter cut-off length Engine aerodynamics: Wider simulation domains with proper boundary conditions Appropriate boundary layer description, coupled with 1D CFD simulations Numerical methods: Higher precision, less dissipative approaches Adapted to new machine architectures Computing resources: LES requires large computing resources (CPU, memory and storage) But still, the purpose of engine LES simulation needs clarification: Engine LES to answer which questions? With which post-processing? Benefit/Cost still unclear 14 2016 I F P E N
ONE POSSIBLE SOLUTION: LES ON THE CLOUD! ACCESS-LES Available HPC resources An attractive and modern simulation platform Make LES available for new users and propose answers 15 2016 I F P E N
A NEW PARADIGM: TOWARDS CONNECTED MOBILITY IFPEN -Cyrille DUPONT IFPEN - Objectif Images / Clémentine BÉJAT 16 2016 I F P E N
PRESENT SOCIAL EXPECTATIONS Redefine Mobility Will to participate but with only slight change of habits: Without constraints and over-costs But awareness that things will be different Promote change by being sensitive to individual acceptation and transform it into business opportunities: Incentives Regulatory measures To follow and accompany new mobility solutions: Without jeopardizing current business Even strengthening it, if possible But moving inside a constrained field: Reduced CO 2 and pollutant emissions Constant / improved services to clients 17 2016 I F P E N
CURRENT TECHNOLOGICAL TREND The smartphone and its key role It is no longer exclusively an on-board infotainment object: New smartphone apps interact with the electronics The smartphone can manage vehicle data Electronics inside the car becomes a data source Smartphone technology goes under the hood: Use of microprocessors derived from cell phones Multi-thread technology Continuous access to cloud data 18 2016 I F P E N
TWO MAJOR SOFTWARE FAMILIES Automotive engineering Software 3D CFD: Fluid mechanics, combustion, heat transfer 1D CFD: Air path, acoustics System simulation: Fuel injection, engine, vehicle Finite element analysis: Solid structures, crash, NVH Control and calibration Connected mobility Apps 19 2016 I F P E N
TWO MAJOR SOFTWARE FAMILIES Automotive engineering Software Eco-driving / Optimal driving Remote Diagnostics Web Services Drive aids, Cartography Battery charging systems Connected mobility Apps 20 2016 I F P E N
FUTURE CHALLENGE: 1 + 1 = 3! Automotive engineering Software (c) IFPEN -GECO An answer to major mobility issues may include fully connected engineering software! Predictive multi-physics tools (fluid mechanics, heat transfer, structural analysis ) System simulation tools integrated on-board (OBD, ECU ) On-the-fly calibration for Real Driving conditions Cloud infrastructure for data storage and HPC capabilities Fully connected vehicles and dedicated Apps Connected mobility Apps 21 2016 I F P E N
MOBILITY n.0: UTOPIC VIEW? NOT REALLY! Welleman at nl.wikipedia 22 2016 I F P E N
23 2016 I F P E N
Find us on: www.ifpenergiesnouvelles.com @IFPENinnovation 24 2016 I F P E N