Modelling of Diesel Vehicle Emissions under transient conditions Dr. Gavin Dober Combustion and Hydraulics Manager, Davide Del Pozzo Delphi Trainee 216-217 Advanced Injection & Combustion Center Delphi Powertrain, Luxembourg French GT Conference, Paris - 17 th May 217
Presentation Overview Background and Motivation Challenges Model Overview Engine Vehicle After-treatment Emissions After-treatment Thermal Results Thermal Modelling Emissions Engine Out SDPF U/SCR Ambient Temperature Influence Conclusions
Background and Motivation Traffic Ambient conditions Engine temperature Battery SoC Driving style Engine Speed/Load Dynamic conditions New emission regulations are challenging because they introduce an unknown variability into every test It will be impossible to test for compliance, simulation can fill the gap
Challenges Future vehicle emissions models must capture the interaction of a number of control variables: Engine out emissions considering the impacts of temperature, altitude, transient impacts and friction Road load and Driver variation After-treatment functionality considering thermal effects, conversion efficiency and emissions storage
Model Overview Model The model has been built of several parts: Engine Vehicle Driver After-treatment Thermal After-treatment Emissions
FMEP [bar] Engine and Vehicle Vehicle Engine The Vehicle subassembly includes: the vehicle body the transmission the differential the driveshaft the brakes and tires The engine is modelled using maps: BMEP, Friction, Fuel Consumption, Air Flow and Emissions Friction Correction Coolant temperature [k]
After-treatment Thermal After-treatment performance is temperature dependent Main difficulties : Predict transient temperature from steady state data Predict temperature evolution in the exhaust system Physical modelling is complex and a simple pipe model is used to correct the turbine out temperature Pipe parameters are fitted using a bisection search algorithm Note also: Sensor time lag must be modelled The exhaust pipe must be discretized to capture the temperature variations over its length
Efficiency Efficiency After-treatment Emissions After-treatment efficiency is mapped according to flow rates & temperatures. This is simple and very effective for SDPF & SCR LNT must additionally model the amount storage and release of NO x 1 SDPF.5 5 4 3 Bed T [ C] 2 1 2 8 6 4 Exhaust flow [g/s] 1 1.5 SCR 5 4 3 Bed T [ C] 2 1 2 8 6 4 Exhaust flow [g/s] 1
Results
Turbine Out Temp [ C] Engine OutTemp [ C] Thermal modelling results WLTC Engine1 Good correlation between experiment and simulation 7 6 Simulation 5 4 3 2 1 2 4 6 8 1 12 14 16 18 4 3 Experiment Simulation Thermal Tuning 2 1 2 4 6 8 1 12 14 16 18
CO2 [mg/km] NOx [mg/km] Engine Out Emissions WLTC Engine1 12 1 Simulation Experiment 8 6 4 2 12 1 2 4 6 8 1 12 14 16 18 8 Good correlation between the experiment and the simulation The NO x emissions are also well predicted, even without a specific correction for the impact of engine transients 6 4 2 2 4 6 8 1 12 14 16 18
Temperature SDPF [ C] Efficiency NOx [mg/km] SDPF WLTC Engine2 18 16 14 12 1 8 6 4 2 4 35 3 Experiment Simulation 2 4 6 8 1 12 14 16 18 1.8 A different engine with a slightly worse engine out NO x result Temperature profile in the SDF is well captured. Conversion efficiency increases over the cycle starting from around 25seconds. SDPF Efficiency 25 2 15 1 5.6.4.2 2 4 6 8 1 12 14 16 18 2 4 6 8 1 12 14 16 18
Temperature SCR [ C] Efficiency NOx [mg/km] U/SCR WLTC Engine2 18 16 14 12 1 8 6 4 2 35 3 25 Experiment Simulation 2 4 6 8 1 12 14 16 18 1.8 Underfloor SCR temperature is well captured Total conversion efficiency of both catalysts compares well to experiment Tailpipe emissions accurately predicted Total Efficiency 2 15 1 5.6.4.2 2 4 6 8 1 12 14 16 18 2 4 6 8 1 12 14 16 18
Turbine Out Temp [ C] Average Efficiency Turbine Out Temp [ C] Ambient Temperature Influence 25 2 15 1 5 25 2 15 1 Simulation 23 C Experiment 23 C 5 1 15 2 25 3 Simulation C Experiment C Validation with RTS95 1.8.6.4 Good validation of the prediction of after-treatment temperature at different ambient conditions Low ambient temperatures reduce the after-treatment conversion efficiency both early and late in the test cycle WLTC Engine2 15-18s 12-15s 9-12s 6-9s 3-6s -3s 5 Validation with RTS95 5 1 15 2 25 3.2-1 -5 5 1 15 2 25 3 35 Ambient Temperature [ C]
Conclusions GT Suite can be used to simply and effectively create a vehicle model for the prediction of tail-pipe emissions under transient conditions Simulations show good agreement with temperature measurements and vehicle emissions There are some key requirements for good vehicle simulations: Accurate engine out steady emissions and temperatures maps A precise model of the temperature evolution within the exhaust system Well characterized interaction between exhaust flow, temperature and conversion efficiency Such a simulation tool is useful for exploring the impact of different vehicle architectures and test conditions on emissions performance