3.5L V6 EcoBoost combustion system development flow chart. Source: Ford. Click to enlarge.

In developing the system, Ford engineers focused on fundamental physics, with an emphasis including injector spray pattern, piston geometry, intake port design, and innovation of operating strategies. The methodology relied on 3D CFD (computational fluid dynamics) modeling, together with experiments that used optical, single-cylinder, and multi-cylinder engines. The methodology reduced the number of hardware iterations required.

To meet the specifications of the D35 EcoBoost, the combustion system had to be able to meet the following attributes:

1. Operate at 17 bar BMEP over a engine speed range between 1,500 to 5,500 rpm.

2. Achieve fuel-air mixing homogeneity equivalent to its naturally-aspirated (NA) counterpart at part-load conditions to ensure fuel economy targets.

3. Deliver PZEV engine-out emissions.

4. Minimize smoke emissions and fuel liner impingement for low oil dilution, to ensure engine durability.

Development of a direct-injection spark-ignition (DISI) engine for optimal performance is technologically challenging. One reasons is that the functional requirements for fuel-air mixture formation are fundamentally different between stratified charge and homogeneous charge operation.

In the EcoBoost engine, homogeneous charge mode is used almost exclusively, except at the cold-start operation where stratified charge operation is found to be very beneficial. Local stratification at cold-start condition provides reliable ignition and stable combustion with reduced emissions. The mixture formation is affected by many complex interactions of fuel spray, in-cylinder flow motion, and piston geometry.

—Yi et al. (2009)

Optimization of intake port. Two major factors drive intake port optimization in turbocharged DISI engines:

• Port flow capacity to improve engine power; and
• In-cylinder flow motion that affects fuel droplet trajectory and fuel-air mixing, as well as the turbulent intensity, that affects burn rate, stability, heat transfer and engine knock tendency.

Unfortunately, the Ford engineers noted, the two requirements are generally competing. Higher flow capacity generally results in lower in-cylinder flow motion, while an intake port with higher flow motion generally yields lower flow capacity. Port optimization in the EcoBoost engine concentrated on adequate flow capacity and high tumble flow motion leading to good fuel-air mixing homogeneity and good knock resistance.

Through their work, the engineers identified that the geometry on the short side of the port and the top nose area (‘a’ and ‘d’ in the diagram below, left) are key to affecting the tumble flow motion and port flow capacity.

Port design parameters ‘a’ and ‘d’ shown on Port 5 (blue) vs. GTDI iVCT XO (red). Source: Ford. Click to enlarge.	Near-valve flow structure comparison between high flow port (left) and high tumble port (right). The high flow port has a more evenly distributed velocity flow field around the valve seat area, while the high tumble port has a more biased flow structure with flow separation on the short side of the port; 5000 rpm, 19 bar BMEP. Source: Ford. Click to enlarge.

Ford used CFD modeling to evaluate 20 intake port designs; only two were ever made in hardware before deciding on the final design.

Optimization of injector spray pattern. Five major types of injectors are considered for DISI applications: swirl, air-assisted, fan, outward-opening (i.e., piezo), and multi-hole. Ford selected the multi-hole injector design. The multi-hold injector is an extension of the gasoline PFI and diesel injectors, but operates in the range of 30 to 200 bar (compared to around 4 bar for PFI and up to 2000 bar for diesel).

Ford evaluated a total of 24 different injector spray patters with CFD modeling, which predicted mixing homogeneity and piston fuel impingement at warmed-up conditions, and air-fuel ratio at spark timing at cold-start conditions.

Dynamometer data at 1500 rpm/5 bar BMEP confirmed that the optimized spray injector pattern allowed 15° earlier injection timing with low soot emissions.

In a PFI engine (left), fuel is injected onto intake port surfaces and closed intake valves, while in a DI engine (right) the fuel is injected directly into the cylinder, making cold-start emissions more challenging. Source: Ford. Click to enlarge.

Piston bowl optimization and cold-start performance improvement. The cold-start period constitutes the majority of engine emissions due to zero or low conversion efficiency of the aftertreatment system. Testing has shown that both UHC and NO_x emissions during the first 50 seconds account for about 80% of the total emissions of the entire FTP75 drive cycle.

Cold starts in a DISI engine are challenging, because during cold-start, the liquid fuel droplets may not fully evaporate in the cylinder due to the colder thermal environment. Some of the droplets may escape into the exhaust system and result in high UHC emissions.

A turbocharged DI engine makes the cold-start emissions even more challenging...twice as much heat flux as in a naturally aspirated DI engine is required to light-off the catalyst due to increased heat loss across the turbocharger. The higher heat flux requirement forces the engine to burn fuel less efficiently inside the cylinder with overly retarded ignition timing. This tends to make the engine combustion very unstable during cold-start operation.

—Yi et al. (2009)

Cold-start is typically considered in two stages: crank and run-up. Ford determined that an optimized intake-compression stroke split injection strategy reduces the UHC by 30% during the cranking period.

The challenge in the second phase is quickly to warm-up the catalyst, commonly achieved by retarding combustion. In a turbocharged DISI engine, the combustion process must be further retarded than in a PFI engine. However, combustion stability deteriorates very quickly as the combustion phasing retards.

Ford identified stratified-charge operation as an enabler to achieve such very retarded combustion timing. Bowl-in-piston geometry and split fuel injection are key to producing the stratified charge. Ford evaluated 15 piston bowl designs using CFD modeling.

The split injection strategy applies the first injection in the early intake stroke and the second injection in the compression stroke. Based on CFD and optical images, Ford engineers further improved the injection strategy by retarding the first fuel injection. The retarded first fuel pulse generates a preliminary fuel-air mixture stratification before the second fuel injection. The second fuel injection reinforces the stratification and forms a much more stable fuel-air mixture around the spark plug.

Testing showed an improvement in combustion stability, and NO_x and UHC emissions below the engineering PZEV targets.

Resources

• Jianwen Yi, Brad VanDerWege, Corey Weaver, Zheng Xu, George Davis, Brett Hinds, Andreas Schamel, Steven Wooldridge, Gary Coulson, James Hilditch, Claudia Iyer, Peter Moilanen, George Papaioannou, David Reiche, Michael Shelby. Development and Optimization of the Ford 3.5L V6 EcoBoost Combustion System (SAE 2009-01-1494)

• Claudia Iyer, Jianwen Yi. 3D CFD Upfront Optimization of the In-Cylinder Flow of the 3.5L V6 EcoBoost Engine (SAE 2009-01-1492)

• Zheng Xu, Jianwen Yi, Steven Wooldridge, David Reiche, Eric Curtis, George Papaioannou. Modeling the Cold Start of the Ford 3.5L V6 EcoBoost Engine (SAE 2009-01-1493)

• David B. Reiche, Steven T. Wooldridge, Peter C. Moilanen, George C. Davis. Experimental Optimization of the Cold Start for the EcoBoost Engine (SAE 2009-01-1491)

• Robert Stein, Christopher House, Thomas Leone. Optimal Use of E85 in a Turbocharged Direct Injection Engine (SAE 2009-01-1490)

• Anand H. Gandhi, Mark Meinhart. Fuel Injector Flow Rate Analysis for the Duratec 35 EcoBoost (SAE 2009-01-1505)