Final Report: High-Authority Fuel InjectionEPA Contract Number: 68D02056
Title: High-Authority Fuel Injection
Investigators: van Schalkwyk, Mauritz
Small Business: Mide Technology Corporation
EPA Contact: Manager, SBIR Program
Project Period: June 1, 2002 through June 1, 2004
Project Amount: $224,926
RFA: Small Business Innovation Research (SBIR) - Phase II (2002) Recipients Lists
Research Category: Air Quality and Air Toxics , SBIR - Air Pollution , Small Business Innovation Research (SBIR)
Promising developments in diesel injection technology have given rise to increasing acceptance of diesel engines in the passenger vehicle market, especially in Europe. As with heavy-duty vehicles, more stringent emission levels are achieved via control of the diesel combustion process through precise timing and metering of the injected fuel quantities. By controlling injection timing and duration, fuel quantity, and rate shape (flowrate profile as a function of time), it is possible to effectively control engine performance.
For example, European manufacturers such as Siemens and Bosch are competing to bring the benefits of piezoelectric injection technology to market in high-pressure common rail fuel systems. This change increased the accuracy and individual control over fuel delivery as a function of load, speed, and ambient conditions (air and fuel temperatures and pressures). Two-way solenoid and piezoelectric valves, however, typically are limited to digital operation; they are either fully open or fully closed. This characteristic is beneficial for controlling fuel quantity and injection timing, but generally is poor for shaping the flowrate profile and carries limitations on the lower bound of injected fluid packet volumes. Midé Technology Corporation has developed a hydraulic unit injector prototype with a servo-style control valve, through replacement of the traditional solenoids with a piezoelectric actuator and controller. This approach enables proportional control over the injection process, and myriad degrees of freedom to control the injection event.
The goal of this Phase II research project was to experimentally demonstrate the emission-reducing capability of a prototype injector. One performance metric was to generate characteristic injection profiles for low-emission engine operation. These would be measured on an industry standard rate tube test, in which injected fluid flowrate profiles were detected with a pressure sensor near the injector nozzle. The target injection shape was determined through combustion modeling. A second performance metric was to produce precision microinjection volumes with the prototype injector, measured on an EMI2 flowmeter. Targets were 1 mm3 average injected quantity over many events, with 0.05-0.1 mm3 standard deviation (shot-to-shot repeatability). Meeting these two performance metrics demonstrated that the prototype injector had pushed the state-of-the-art in emissions-reducing diesel fuel injection technology.
Technical risk and program cost were greatly reduced by Midé’s working (not contractual) partnership with Sturman Industries. Leveraging their heritage and intellectual property in precision control valves and diesel fuel injectors shortened the development process considerably. Sturman already had designed and built a fast, new G2 hydraulic unit injector for turbo-diesel engines. With their permission, Midé obtained several control valves and injectors for use in the high-authority fuel injection development. Proportional control of their spool valves was a desirable value that Midé added, increasing injection authority and rate-shaping capabilities to enable reduced emissions.
The piezoelectric actuator system and injector integration was the main development focus of Phase II. The driven mass (load) already had been established as the G2 control valve spool, but the desired stroke of 0.44 mm (17 mils) would require a prohibitively long piezoelectric stack actuator. On the other hand, the required force to drive the 4.5 g spool was well below the force capabilities of most commonly manufactured piezoelectric stacks. These mismatched force and displacement requirements steered development toward gained piezoelectric actuators, which would multiply a given piezoelectric stack’s stroke at the expense of driving force.
Several different mechanical gain concepts were investigated, including an X-frame concept, an O-frame design, a taper gain actuator, and a jack actuator. Prior research using the X-frame actuator indicated that its maturity level was higher than the others at the start of the project. Thus, most of the initial valve testing was conducted using the X-frame. The O-frame was a derivative from the X-frame and was optimized to meet the needs of this diesel injector application. The final injector prototype used in most of the rate tube and EMI2 flowmeter testing incorporated the O-frame actuator.
The major drawback of the X- and O-frame designs was that both relied on a scissoring motion for the displacement gain, which increased sliding friction at the spool-valve interface. The slight arcing and foreshortening of the actuator stroke created difficult problems when transferring that motion to the spool in a linear fashion. This prevented a purely axial load on the spool, and the small radial loads resulted in added friction. That friction meant a full closed-loop controller design would be prohibitively expensive to compensate for excessive phase loss in the plant. The final prototype system employed closed-loop control on the DC position only, while open loop control was implemented during the high-speed injection event. This was referred to as the “hybrid” closed-loop control system.
Injector prototype tests were conducted on the rate tube test setup to verify injection profile shaping authority. Results indicated low-emission injection profile shaping capability, comparing favorably to the target shape determined through combustion modeling. Proportional authority over the control valve spool enabled an infinite number of injection profiles to be generated. Timing, rising edge slope, duration, and falling edge slope of the injection trace could be effectively controlled, with polytonic shapes possible and unstable needle “chatter” observed as well.
Subsequently, the microinjection capabilities of the prototype injector (with O-frame actuator) were quantified. Average per injection volume (with distribution) was calculated using calibrated rate tube testing hardware to produce a map of injected volume as a function of spool set point and rail pressure for a given spool trajectory. Microinjection volumes on the order of 0.5 mm3 were calculated through rate tube testing, with standard deviations of approximately 0.2 mm3. Once the level of performance was considered suitable, the hardware was shipped to Sturman Industries for final evaluation on the precision EMI2 flowmeter to actually measure the injection volume map. Precision EMI2 flowmeter test results indicated that the prototype injector was able to meet the average injection volume target of 1 mm3 under certain operating conditions, but shot-to-shot repeatability typically was about 0.3 mm3 (using a signal dither), which was higher than the 0.05-0.1 mm3 standard deviation target.
Although the prototype injector did not meet all of the performance goals, the demonstrated capabilities were very promising for reducing engine emissions through diesel injection control. This experimental success may enable technology commercialization through engine manufacturers as a critical method of satisfying environmental regulatory constraints. Furthermore, the development of breakthrough actuation and control technologies can be used to help realize innovative combustion control concepts, such as the Homogeneous Charge Compression Ignition technology. The resulting high-speed proportional (or servo) control valve also could be further developed into a stand-alone product for many hydraulic applications.
Prior to commercialization of the technology, several technical risk areas must be addressed. First, the piezoelectric actuator (stacks) requires more than 500 W peak power to fully operate the valve at frequencies up to 1 kHz. This is instantaneous power required, and the energy per injection is more manageable. The stacks only require about 130 mJ of electrical energy per cycle for full spool stroke. Still, the power electronics must be able to deliver the peak power to each injector, and 500 W may be a concern. Fortunately, the bulk of injector testing during this research project employed spool profiles that used only a portion of the full spool stroke. If a complete system could be developed that only used a fraction of the full spool travel, peak power could be reduced.
Second, the form factor of the prototype prohibits integration into any existing engine. The actuator, in its current configuration, is too large. On all sides of the injector, valve stems and springs limit the clearance for any actuator in a real engine. Above the injector sits the medium pressure oil rail, and above that is the cam cover, which ensures containment of the vented oil. Most likely, this injector could not be retrofit onto an existing engine, but may be redesigned concurrently with new engine development to enable integration.
Third, there currently is no clear path to driving the overall cost of this injector design that is low enough for market insertion. The most prominent cost additions to the prototype injector, above the standard G2 injector, are two piezoelectric stacks and the spool position sensor. Driving electronics and control algorithms could feasibly be integrated with an engine, similar to all other injector support systems. However, it would be advantageous to convert to a single stack actuator design, as opposed to the dual-stack design currently used. An inexpensive position sensor also would be needed, but the accuracy and resolution of the sensor still must meet the requirements of the injector system.
Ultimately, the prototype injector pushed the state-of-the-art in diesel injection. The G2 injector platform was designed for digital control using solenoid actuation, effectively controlling only timing and duration. Once retrofitted with Midé’s gained piezoelectric actuator, proportional spool valve control dramatically increased the injection authority. Rate shaping capabilities were demonstrated on many levels, including consistent low-emission (clean) injection profiles. Precision microinjection volumes also were demonstrated down to 1 mm3 average volumes with shot-to-shot repeatability of around 0.3 mm3.