Small engines, such as those used in ATVs, have high engine out emisisons, and are terribly polluting. This project examines the effectiveness of a new control technology to implement low cost electronic fuel injection on these engines to reduce emissions levels. OBJECTIVES: To date, we have run simple simulations of our technology to identify system accuracy and feasibility. No engine modeling has taken place, and many assumptions have been made. These simulations are representative of only steady-state and piece-wise linear engine operation. We can not obtain dynamic models without more powerful software, engine models, and simulation tools. We wish to model our method of crankshaft position determination using MatLab and Simulink to better understand its dynamic operation. From these models, we should be able to design a more effective control algorithm and determine any system deficiencies before we develop a functional prototype. Long term, we believe that system modeling will reduce our design time, reduce prototype iteration, and increase program efficiencies. APPROACH: To develop models for simulation, we need to instrument and motor an ATV engine on a dynamometer and observe its operating characteristics. These characteristics will allow us to build accurate models for simulation. The characteristics of interest are a model of sub-cyclic acceleration/deceleration of the crankshaft between ISI pulses, engine volumetric efficiency over its operating rpm range, and the angular relationship between the ISI signal and absolute crankshaft position over the rpm range. We plan to test hardware and software implementations of the ISI circuitry to determine the most flexible and cost effective solution. We will alter the hysteresis levels of this circuitry to examine signal integrity of the ISI circuitry and provide a robust signal under all engine operating conditions. In addition to measuring these values, we need to disassemble a complete engine to measure engine component mass, inertia, and certain geometries necessary for development of an accurate engine model in MatLab. With this baseline information, we will develop a more accurate model of this engine and its operating characteristics with MatLab. Once models have been generated, we will modify and improve the algorithm for the determination of crankshaft position. The determination of crankshaft position is crucial for electronic metering of fuel, since the fuel must be injected at the proper time in the engine cycle and the spray pattern cannot overlap multiple engine cycle periods. In the future, we plan to not only control fuel metering for the ATV engine, but also control of spark ignition using our single sensor technology. While fueling has a comparatively large time window for injection timing, ignition timing must be significantly more accurate. In this proposal, we will focus primarily on fuel metering rather than ignition timing
