Search Results

Optimization of the engine cold start is critical for gasoline direct injection (GDI) engines to meet increasingly stringent emission regulations, since the emissions during the first 20 seconds of the cold start constitute more than 80% of the hydrocarbon (HC) emissions for the entire EPA FTP75 drive cycle. However, Direct Injection Spark Ignition (DISI) engine cold start optimization is very challenging due to the rapidly changing engine speed, cold thermal environment and low cranking fuel pressure. One approach to reduce HC emissions for DISI engines is to adopt retarded spark so that engines generate high heat fluxes for faster catalyst light-off during the cold idle. This approach typically degrades the engine combustion stability and presents additional challenges to the engine cold start. This paper describes a CFD modeling based approach to address these challenges for the Ford 3.5L V6 EcoBoost engine cold start.

This paper describes a CFD modeling based approach to address design challenges in GDI (gasoline direct injection) engine combustion system development. A Ford in-house developed CFD code MESIM (Multi-dimensional Engine Simulation) was applied to the study. Gasoline fuel is multi-component in nature and behaves very differently from the single component fuel representation under various operating conditions. A multi-component fuel model has been developed and is incorporated in MESIM code. To apply the model in engine simulations, a multi-component fuel recipe that represents the vaporization characteristics of gasoline is also developed using a numerical model that simulates the ASTM D86 fuel distillation experimental procedure. The effect of the multi-component model on the fuel air mixture preparations under different engine conditions is investigated. The modeling approach is applied to guide the GDI engine piston designs.

Recently, Ford Motor Company announced the introduction of EcoBoost engines in its Ford, Lincoln and Mercury vehicles as an affordable fuel-saving option to millions of its customers. The EcoBoost engine is planned to start production in June of 2009 in the Lincoln MKS. The EcoBoost engine integrates direct fuel injection with turbocharging to significantly improve fuel economy via engine downsizing. An application of this technology bundle into a 3.5L V6 engine delivers up to 12% better drive cycle fuel economy and 15% lower emissions with comparable torque and power as a 5.4L V8 PFI engine. Combustion system performance is key to the success of the EcoBoost engine. A systematic methodology has been employed to develop the EcoBoost engine combustion system.

The accurate prediction of fuel sprays is critical to engine combustion and emissions simulations. A fine computational mesh is often required to better resolve fuel spray dynamics and vaporization. However, computations with a fine mesh require extensive computer time. This study developed a methodology that uses a locally refined mesh in the spray region. Such adaptive mesh refinement will enable greater resolution of the liquid-gas interaction while incurring only a small increase in the total number of computational cells. The present study uses an h-refinement adaptive method. A face-based approach is used for the inter-level boundary conditions. The prolongation and restriction procedure preserves conservation of properties in performing grid refinement/coarsening. The refinement criterion is based on the mass of spray liquid and fuel vapor in each cell. The efficiency and accuracy of the present adaptive mesh refinement scheme is demonstrated.

Parallel computing schemes were developed to enhance the computational efficiency of engine spray simulations with adaptive mesh refinement (AMR). Spray simulations have been shown to be grid dependent and thus fine mesh is often used to improve solution accuracy. In this study, dynamic mesh refinement adaptive to spray region was developed and parallelized in KIVA-4. The change of cell and node numbers and the local characteristics in the dynamic mesh refinement posed difficulties in developing efficient parallel computing schemes to achieve low communication overhead and good load balance. The present strategy executed AMR on one processor with data scattering among processors following the adaptation, and performed AMR every ten computational timesteps for enhanced parallel performance. The re-initialization was required and performed at the minimized cost.

This paper proposes a film dynamics model for liquid film resulting from fuel spray impinging on a wall surface. It is based on a thin film assumption and uses numerical particles to represent the film to be compatible with the particle spray models developed previously. The Lagrangian method is adopted to govern the transport of the film particles. A new, statistical treatment was introduced of the momentum exchange between the impinging spray and the wall film to account for the directional distribution of the impinging momentum. This model together with the previously published models for outgoing droplets constitutes a complete description of the spray wall impingement dynamics. For model validation, films resulting from impinging sprays on a flat surface with different impingement angles were calculated and the results were compared with the corresponding experimental measurements.

The influence of mixture preparation on misfires at idle in a Direct Injection Spark Ignition (DISI) engine was investigated. A wall-guided DISI engine was run at idle conditions in a stratified charge mode (750 rpm / 90 kPa MAP). Images of the mixture composition at the spark plug were taken at spark timing using Planar Laser Induced Fluorescence (PLIF) for several different End-of-Injection (EOI) timings and spark timings. Cylinder pressure data were acquired simultaneously with the images to identify misfire cycles. The misfire rate was found to increase as the EOI timing was advanced from the optimal timing, defined by maximum stability and lowest ISFC. Images show that the misfire rate at a particular operating condition can be correlated to the fuel distribution and the location of the stratified charge in the engine. Cycles that showed a lower amount of stratification (overmixing) and/or high gradients in fuel concentration near the spark plug were the least stable.

In the effort to improve combustion of a Light-load Stratified-Charge Direct-Injection (LSCDI) combustion system, CFD modeling, together with optical engine diagnostics and single cylinder engine testing, was applied to resolve some key technical issues. The issues associated with stratified-charge (SC) operation are combustion stability, smoke emission, and NOx emission. The challenges at homogeneous-charge operation include fuel-air mixing homogeneity at partial load operation, smoke emission and mixing homogeneity at low speed WOT, and engine knock tendency reduction at medium speed WOT operations. In SC operation, the fuel consumption is constrained with the acceptable smoke emission level and stability limit. With the optimization of piston design and injector specification, the smoke emission can be reduced. Concurrently, the combustion stability window and fuel consumption can be also significantly improved.

A new Light Stratified-Charge Direct Injection (LSC DI) spark ignition combustion system concept was developed at Ford. One of the new features of the LSC DI concept is to use a ‘light’ stratified-charge operation window ranging from the idle operation to low speed and low load. A dual independent variable cam timing (DiVCT) mechanism is used to increase the internal dilution for emissions control and to improve engine thermal efficiency. The LSC DI concept allows a large relaxation in the requirement for the lean after-treatment system, but still enables significant fuel economy gains over the PFI base design, delivering high technology value to the customer. In addition, the reduced stratified-charge window permits a simple, shallow piston bowl design that not only benefits engine wide-open throttle performance, but also reduces design compromises due to compression ratio, DiVCT range and piston bowl shape constraints.

A spontaneous Raman scattering diagnostic was used to measure the residual fraction in a single-cylinder, 4-valve optically accessible engine. The engine was operated at 1500 rpm on pre-vaporized iso-octane at several intake manifold pressures (50-90 kPa). Cam phasing was varied to determine the effect of intake valve timing and valve overlap on the residual mass fraction of the engine. A simple model based on the ideal Otto cycle and 1D gas flow through the exhaust valves was proposed to analyze the results of the Raman experiment. The model showed good agreement (R2=0.91) with the experimental results and demonstrated its potential for use as a method to estimate the residual fraction in an engine from available dynamometer data. The experimental results showed that the residual fraction was reduced at higher manifold pressures due to less backflow through the exhaust valves and varied with intake cam phasing.

Particle Image Velocimetry (PIV) experiments were performed on single-cylinder versions of a 0.375 L/cylinder and a 0.5 L/cylinder engines from the same engine class to determine the differences in swirl flow between the two engines. Two engine speeds (750 and 1500 rpm), manifold pressures (75 kPa and 90 kPa) and valve timings (maximum overlap and with the intake valve 20° retarded from the max overlap position) were examined. The swirl ratio (SR) and mean velocity (|V|) were calculated at BDC for every case in the mid-stroke plane and the fluctuation velocity (U') calculated for the 1500 rpm / 90 kPa / maximum overlap case. The in-cylinder velocities do not differ by the expected ratio of mean piston speed caused by differences in the engine stroke. The smaller engine was expected to have lower in-cylinder velocities and SRs due to a shorter stroke and lower piston speeds but instead has SR and |V| levels that are the same or higher than the larger engine.

Particle Image Velocimetry (PIV) measurements were performed on a single cylinder optically accesible version of a 3.0L 4-valve engine using a Camshaft Profile Switching (CPS) system. The flow field was investigated at two engine speeds (750 and 1500 rpm), two manifold pressures (75 and 90 kPa) and two intake cam centerlines (maximum lift at 95° and 115° aTDCi respectively). Images were taken in the swirl plane at 10 mm and 40 mm below the deck with the piston at 300° aTDC of intake (60° bTDC compression) and BDC respectively. In the tumble plane, images were taken in a plane bisecting the intake valves with the piston at BDC and 300° aTDC. The results showed that the swirl ratio was slightly lower for this system compared with a SCV system (swirl control valve in the intake port) under the same operating conditions. The swirl and tumble ratios generated were not constant over the range of engine speeds and manifold pressures (MAP) but instead increased with engine speed and MAP.

Swirl plane Particle Image Velocimetry (PIV) measurements were performed in a single-cylinder optically accessible gasoline direct injection (DISI) engine using a borescope introduced through the spark plug hole. This allowed the use of a contoured piston and the visualization of the flow field in and around the piston bowl. The manifold absolute pressure (MAP) was fixed at 90 kPa and the engine speed was varied in increments of 250 rpm from 750 rpm to 2000 rpm. Images were taken from 270° to 320° bTDC of compression at 10° intervals to study the evolution of the velocity fluctuations. Measurements were performed with and without fuel injection to study its effect on the in-cylinder flow fields. Fuel was injected at 10 MPa and 5 MPa. The 2-D spatial mean velocities of individual flow fields and their decompositions were averaged over 100 cycles and used to investigate the effects of engine speed and image timing on the flow field.