UC Berkeley

Homogeneous Charge Compression Ignition (HCCI) engines are one of the most promising engine technologies for the future of energy conversion from clean, efficient combustion. HCCI engines allow high efficiency and lower CO2 emission through the use of high compression ratios and the removal of intake throttle valves (like Diesel), and allow very low levels of urban pollutants like nitric oxide and soot (like Otto). These engines, however, are not without their challenges, such as low power density compared with other engine technologies, and a difficulty in controlling combustion timing.

This dissertation first addresses the power output limits. The particular strategies for enabling high power output investigated in this dissertation focus on avoiding five critical limits that either damage an engine, drastically reduce efficiency, or drastically increase emissions: 1) ringing limits, 2) peak in-cylinder pressure limits, 3) misfire limits, 4) low intake temperature limits, and 5) excessive emissions limits. The research shows that the key factors that enable high power output, sufficient for passenger vehicles, while simultaneously avoiding the five limits defined above are the use of: 1) high intake air pressures allowing improved power output, 2) highly delayed combustion timing to avoid ringing limits, and 3) using the highest possible equivalence ratio before encountering ringing limits. These results are revealed by conducting extensive experiments spanning a wide range of operating conditions on a multi-cylinder HCCI engine.

Second, this dissertation discusses strategies for effectively sensing combustion characteristics on a HCCI engine. For effective feedback control of HCCI combustion timing, a sensor is required to quantify when combustion occurs. Many laboratory engines use in-cylinder pressure sensors but these sensors are currently prohibitively expensive for wide-scale commercialization. Instead, ion sensors made from inexpensive sparkplugs are proposed for sensing combustion timing. Ion sensing, however, is unreliable under certain HCCI conditions. The dissertation presents two strategies for improving the usefulness of ion sensors in HCCI engines: 1) the use of tiny fractions of metal-acetate fuel additives that expand the useful range of ion sensors, and 2) the use of ion sensors for detecting excessive ringing that must be avoided in HCCI engines. These two innovative research efforts make ion sensors viable for sensing combustion characteristics across the full range of HCCI operation, making them effective for use in engine control systems.

In summary, this Ph.D dissertation addresses two important technical challenges facing HCCI engines: power output limits, and difficulty in sensing combustion characteristics for control applications. The strategies proposed in this dissertation research bring HCCI engines closer to widespread commercialization allowing vehicles to operate with significantly higher efficiency and with cleaner emissions.