Great success for AVL FIRE™ in JSAE Benchmark - News and Highlights
Great success for AVL FIRE™ in JSAE Benchmark
The Japanese Society of Automotive Engineering (JSAE) benchmarked all major CFD codes offering models for internal combustion engine simulation. AVL FIRE™ was successful in both categories, gasoline and Diesel engine. While the newly developed Flame Tracking and Particle Method impressed GDI Engine developers through its knock calculation capabilities, Diesel enthusiasts applauded the excellent agreement between measured and predicted soot/NOx trade-off.
ABSTRACT: Diesel Engine Emissions Prediction and Gasoline Engine Knocking Estimation by CFD Method
Gregor KOTNIK, AVL-AST d.o.o., Slovenia
In this work an engineering approach in the field of internal combustion engine (ICE) simulations focused on reliable emissions prediction for diesel engines and efficient knock onset estimation for gasoline engines, is presented. Whole scope of investigations is thereby divided into two parts, gasoline and diesel. For the first, investigations were performed for a (full model) PFI gasoline (VVT) engine for high load operating points at 3 different engine speeds (n=1200, 2400 and 4800RPM). Scavenging flow simulation was not considered here, appropriate quantity fields (given as a result from a foreign CFD code) of mixture state and composition at the point of IVC were read-in and mapped on corresponding calculation mesh. Hence the simulation duration was from IVC till EVO for all 3 operating points. The focus here was to establish a fast but still reliable engineering method for (mean) knock onset prediction, hereby avoiding (long-lasting) multiple-cycle knocking combustion simulations. Combustion process was calculated with the Flame Tracking Particle Model (FTPM), a numerical algorithm to simulate the kinetics of the premixed and partially premixed flame represented by a surface. In order to predict the knocking trend in a comprehensive manner, a spark timing variation for each operating point was carried out. Calculated mean cylinder pressure traces are compared to available experimental data and show satisfactory accuracy, additionally predicted knocking probability and resulting knocking trends according to shifted spark ignition timing are presented.
For the second part, CFD simulations were carried out for two HSDI engines (first with simplified and second with realistic piston bowl layout). Both engines were set with (almost) identical geometrical (compression ratio, bore, stroke) and fuel injection setup and were both considered at same engine speed at 2000RPM and two different operating points (low vs. high charge/injection pressure/EGR). Additionally, an EGR ratio sweep (3 points) was performed for both operating conditions. Fuel injection was defined with nozzle flow data at the injector orifice for both operating conditions, Lagrangian spray primary and secondary break-up models were applied. Detailed chemistry approach was used for calculation of chemical kinetics (NO reactions included) in combination with an empirical emission model for Soot modelling. Diesel simulations were carried out practically without any model parameter tuning, resulting predicted cylinder pressure traces and emissions (Soot/NOx) showed good agreement with available experimental data.
The full article as well as the articles provided by the other participants of the JSAE CFD Benchmark 2016 can be obtained through the website of the Japanese Society of Automotive Engineering.