Auto engineers from Stanford, GM and Bosch team up to make leaner, cleaner engines

By Dawn Levy

Prices that siphon your wallet and pollutants that choke your lungs are reasons you'd want a vehicle that can make the most of its gasoline. Enter HCCI, a new engine technology whose acronym stands for "homogeneous charge compression ignition." It may spawn engines that are 20 percent more efficient with near-zero emissions.

This summer Associate Professor J. Christian Gerdes received $2.5 million to work for three years with General Motors, the world's largest automaker, and Bosch, a leading manufacturer of automotive technologies, to speed the development of HCCI technology. The engines they develop could find their way into conventional or hybrid vehicles in less than a decade, says Gerdes.

His collaborators include Associate Professor Christopher Edwards, who studies combustion processes in part with funding from the U.S. Department of Energy and serves as deputy director of the Global Climate and Energy Project, and Associate Professor (Consulting) Ed Carryer, whose research includes electronic instrumentation. Mechanical engineering graduate students Matt Roelle, Nikhil Ravi and Adam Jungkunz work in the lab. Another student, Greg Shaver, graduated this summer and now works at Entelos, where his career has veered from the automotive realm. He now applies control engineering to create dynamic models of human systems for drug design.

"We are control engineerstrying to manipulate inputs to get a desired output is what we do," Gerdes says. "We've tried to take very simple physical models of the chemical kinetics and use those to tell us when we control the various valve timingswhen we open or close different valves, and as we're moving on to doing fuel injection, it will also tell us when and how much fuel to inject. We're trying to form a link between control theory and the science of chemical kineticshow much of the combustion process do we need to understand to control this?"

Collaborating with industry gives Stanford researchers knowledge about the state of the art and allows them to attack real problems.

"We're a fairly small operation, and it doesn't make sense for us to be competing with GM trying to figure out something they already know," Gerdes says. "It works well when we focus our efforts on learning things that they don't know, and the only way to really know what industry does and does not understand is to collaborate directly because they don't tend to publish everything that they know. [The HCCI collaboration] has been a really good example of that. As we've gotten deeper into the project, it turns out that the issues most critical to industry are also academically the most interesting."

No longer 'a laboratory curiosity'

One of Gerdes' collaborators at GM, Paul Najt, group manager of spark-ignition engine systems, pioneered HCCI in the 1980s. "This form of combustion has been well known for a long time," Gerdes says. "It was originally dismissed as a laboratory curiosity because people didn't think that it could be controlled well enough to really work on an engine. This is another example of something enabled by the advances in microprocessors, in sensors, in control systems."

The "homogeneous charge" part of the HCCI acronym refers to the fact that the fuel and air blend to form a uniform mixture. The "compression ignition" part takes into account that temperature rises with pressure. In the HCCI engine, air and fuel mix with some exhaust gas from the previous piston-stroke cycle to increase the temperature. When the piston compresses the air-fuel mixture, raising its temperature further, the mixture auto-ignites.

This process contrasts with spark ignition, in which air and fuel are drawn into the engine, compressed by the piston and hit with a spark to start combustion. "The problem is that you get a flame front going away from the spark, so you get certain hot spots where you can get emissions, and things burn over a long period of time, relatively, in engine terms," Gerdes says.

The HCCI process also contrasts with that of diesel engines, where pistons compress air in the combustion chamber to make it extremely hot. When diesel fuel is injected into the chamber, it auto-ignites. The process can create soot and oxides of nitrogen, which contribute to ozone depletion and smog formation.

But HCCI engines don't have a direct control over when combustion happens like the timing of a spark, which controls spark ignition systems, or the injection of fuel, which controls diesel systems. "We put everything in the cylinder, we close the valves and we compress itand when the conditions are right in terms of the right temperature and for the concentrations of fuel that we have in there, then combustion occurs," Gerdes explains. "The problem with HCCI relative to these other strategies is that it can be very susceptible to any sort of changes in the environment. If the engine wall temperature is slightly different, if the air temperature is slightly different, you can have combustion happening indeed at a very, very different time, and that's horrible for efficiency."

Gerdes and his colleagues have taken a problem that engineers had been viewing as an engine problem and reformulated it as a controls problem. They employ a five-cylinder Volvo engine for their studies but operate only one cylinder to isolate a single combustion event.

"Things are very interconnected on an engine, so if I change one aspect of it, that propagates through in a lot of complex ways," Gerdes says.

His students are well prepared for handling that interconnectedness. "Our students can take classes in various different areas, so all of my students will take classes in combustion and classes in control," Gerdes says. "Typically speaking, people will study one or the other of those things in graduate school. One of the great things about Stanford is that there are collaborations that allow us to do things that are really interdisciplinary. Now, this is all within [mechanical engineering], but it's shocking how different a language control people and combustion people have. Stanford's a great place to bridge that gap."