Skip to content.

Berkeley Engineering

Educating Leaders. Creating Knowledge. Serving Society.

You are here: Home News Center Publications Forefront Archive Forefront Fall 2008 Features The smart little engine that could: sensing a new automotive feature
Document Actions

The smart little engine that could: sensing a new automotive feature

By Erica Klarreich | Photos by Roy Kaltschmidt

At UC Berkeley’s engine shop in Hesse Hall, (from left) Ph.D. student Samveg Saxena and post-doctoral researcher Hunter Mack adjust the fast thermal management system of a four-cylinder Volkswagen diesel they are using to study the HCCI, or homogeneous-charge compression ignition engine, a new type of internal combustion engine.

It’s 2018, and you’re pulling into a gas station to fuel up. Your engine is 25 percent more efficient than today’s gasoline engine and can run on virtually any fuel. That’s a good thing, because this station is offering not only gasoline and diesel but also biodiesel, hydrogen, natural gas, alcohols such as ethanol and a host of alternative fuels that haven’t yet been designed.

Luckily, you don’t have to navigate this bewildering array on your own. Before you’ve even stepped out of the car, your engine’s wireless sensor network has started a “conversation” with the gas pump’s computer. Translated into colloquial human-speak, it might run something like this: “We’re on our way across the country, and the trunk’s loaded with everything but the kitchen sink. I’m running too hot, and I need a fuel that will burn cooler and lower my emissions.” The car’s fuel vapor sensors report to the pump exactly what fuel molecules are present in the tank, and the pump recommends a blend containing a bit more methanol than usual, which will make the engine run as cleanly and efficiently as possible.

This scenario may seem like science fiction, but mechanical engineers are working on making it reality within the next 10 years. They know that one new car joins the world’s roadways every second. Demand for cars is soaring in China, India and other developing nations. Here at home, transportation accounts for 65 percent of U.S. oil consumption and is straining the world’s capacity to generate fuel. As gas prices fluctuate and Americans grow increasingly aware of the environmental and social costs of our heavy fuel use, the imperative to make cars run more efficiently grows more urgent.

Three UC Berkeley mechanical engineering researchers are tackling this challenge, each with a different approach. Robert Dibble is leading the effort to develop an engine that uses a new, highly efficient form of combustion, known as HCCI, which can run on virtually any fuel. Albert Pisano’s team is developing super robust sensors to live inside the engine that would constantly be “thinking”—checking temperature, pressure, emissions and fuel—and adjusting conditions inside each cylinder from moment to moment to maximize efficiency. And Van Carey is conceptualizing the gas station of the future, designed to help drivers choose the optimal fuel blend at that moment in that particular car’s lifetime.

“The objective is for gas stations not just to sell fuel,” Carey says, “but also to help drivers navigate fuel choices and support a more environmentally friendly lifestyle.” By adopting a customer service–oriented approach, the gasoline industry can support sustainable alternatives that are better for the planet and better for your bank account.

Internal combustion, reconsidered
The new engine being researched by Dibble first appeared on the scene in the late 1970s at Japan’s Nippon Clean Engine Company. Shigeru Onishi and his team noticed that, under certain circumstances, a spark-ignition engine would run with surprising efficiency even though the spark plug wasn’t firing. Normally, researchers would classify this as errant behavior that needs to be controlled. But in this case, the engine seemed to run better when it, not the scientist, was in charge. The researchers had discovered a new form of combustion engine based on a simple principle: a mixture of fuel and air, if compressed enough, will spontaneously ignite. This process came to be known as homogeneous-charge—meaning the fuel and air are premixed—compression ignition, or HCCI.

Over the next two decades, researchers realized that HCCI could potentially avoid the main drawbacks of both gasoline and diesel engines. In a gasoline engine, fuel and air are premixed, then ignited by a spark. The proportion of fuel to air is low, making the fuel burn at a relatively low temperature. That quality, together with the homogeneity of the mixture, makes for a clean, low-pollution burn, since it is hot-burning pockets of fuel or air that are responsible for most emissions. The downside is that the compression ratio inside the cylinders must be kept fairly low, since otherwise some of the fuel might auto-ignite, ultimately damaging the engine. This low compression ratio reduces the engine’s efficiency by a good 15 to 20 percent.

A diesel engine, in which fuel is sprayed into a high-pressure chamber where it auto-ignites, is much more efficient, due to the higher compression ratio. However, the uneven distribution of fuel and air creates emissions: pockets of hot-burning fuel generate soot, and pockets of hot air create nitrous oxide, one of the chief agents responsible for urban air pollutants.

Since HCCI premixes fuel and air and operates at a high compression ratio, it has the potential to combine the best of both worlds: the low emissions of gasoline engines with the increased performance of diesel. But there’s a catch.

“This engine has a mind of its own,” Dibble says, explaining that there is no event that triggers combustion; it simply happens when the pressure and temperature get high enough. It’s a huge challenge to control these factors precisely enough to make the engine fire at the right moment, when the cylinder’s piston is near top-dead-center, the farthest position away from the crankshaft. For that reason, research on the HCCI engine sputtered along slowly for two decades.



By the mid-1990s, improvements in control technologies and the growing demand for high fuel efficiency made HCCI seem more feasible. Berkeley was one of the first institutions to kick into high gear on this research, particularly in the area of the four-cylinder HCCI engine. Most research teams are using single-cylinder engines, which simplify the fundamental research questions but offer only a partial picture of the challenges commercially viable HCCI cars would face.

Berkeley’s work on the four-cylinder HCCI engine has illuminated some of the ways individual cylinders interfere with each other’s performance, says George Lavoie, a mechanical engineer and visiting scholar at the University of Michigan, Ann Arbor, which, with Berkeley is part of a consortium of four universities studying HCCI. “The Berkeley researchers have shown that there is cross-talk between the exhaust pipes. If you close one, it changes the results of the others,” Lavoie says. “That’s something that will have to be controlled in an HCCI engine.”

In a four-cylinder engine, there are inherent differences in the cylinders; unless carefully controlled, they will ignite at four different times, which isn’t optimal. “Understanding what you can do to line up the ignition events at the same time is a large part of what we’ve contributed,” says Hunter Mack, a postdoctoral researcher working with Dibble. The team has developed sensors to detect exactly when combustion occurs in each cylinder, which, Dibble says, is crucial in attaining the degree of control an HCCI engine requires.

“As the car is going down the road,” Dibble explains, “the pressure could go up in a second, and the engine has to be fast enough to compensate so the firing occurs around top-dead-center.” Computer algorithms will be key to achieving this moment-to-moment control, he adds. “We’re getting increasingly close to tackling the challenges automobile companies will face on the production line.”

The small blue square on the engine plug is a microelectromechanical system (MEMS) made of silicon carbide, designed to regulate engine pressure for more efficient power generation. Albert Pisano’s team is developing sensors like these to control factors such as strain, temperature, oxygen and knock.

HCCI engines could hit the market within 10 years, and they may be as much as 25 percent more efficient than gasoline engines. Already, General Motors (GM) and Mercedes Benz have created demonstration HCCI cars. “It’s definitely part of our strategy for the future to have this technology in our cars,” says Matthias Alt, GM’s HCCI program manager. “It’s a very exciting technology to work on.”

When HCCI cars start being sold, they are likely to cost only nominally more than cars with gasoline engines, Mack says. Hybrid electric cars, by contrast, typically cost thousands more than their conventional counterparts, due to changes required in the structure of the car, like a big electric motor or complicated regeneration schemes for recharging the battery.

GM researchers meet twice a year with the academic consortium, relying on them for a better understanding of HCCI’s fundamental physical processes, says Paul Najt, a group manager at GM’s Powertrain Systems Research Lab.

“We monitor their work and absorb the nuggets that fit in with what we’re trying to do,” Najt says, adding that Dibble always “challenges the community to think outside the box.”

Sensing a new automotive future

Professor Robert Dibble (left) and post-doc Hunter Mack install the pressure transducer in the head of the combustion chamber on a single-cylinder Waukesha CFR engine, an industry prototype for researching internal combustion engines.

The enabling technologies that will make Berkeley’s smart engine and smart fueling station a reality are at earlier stages in their evolution than the HCCI engine. These technologies include digital controls, wireless network and diagnostic analysis systems as well as new designs for sensors, also known as microelectromechanical systems or MEMS.

Until now, no one has built sensors that can survive the rigors of the combustion chamber. Such sensors would have to endure a rapid flash of heat as high as 2400 degrees Fahrenheit, followed by an intake of outside air that could be as cold as Minnesota on a wintry day—and they would have to go through hundreds of these cycles per minute, millions throughout the car’s lifetime.

At present, Dibble is making do with sensors that sit outside the engine. Soon, however, Albert Pisano’s team hopes to offer sensors that can survive inside the combustion chamber itself and measure a host of parameters in the individual cylinders, such as temperature, pressure, oxygen content, vibration and fuel type.

“Our belief is that the engine’s computer should know everything,” Pisano says, adding that today’s engine control systems are surprisingly unsophisticated. They typically know little more than engine speed, how hard the driver is pushing on the gas pedal and how much oxygen is in the tailpipe. The systems’ preprogrammed instructions offer only a limited capacity to handle the wide range of conditions an engine faces in the real world.



Fuel station of the future.
(Click here to enlarge)

That degree of control won’t cut it in the HCCI engine or in any engine trying to maximize the efficiency of its individual cylinders. Pisano’s sensors could allow car manufacturers to turn cylinder-by-cylinder control into a precise science, he says. Engineers could not only correct for differences in the individual cylinders while they’re making the car, but also program the car to keep an eye on the cylinders throughout its lifetime.

“If, 1,000 miles down the road, there’s a speck of dirt and a fuel injector gets partially clogged, the sensors would sort that data out as you’re driving the car,” Pisano says. For example, if the fuel injector in just one cylinder malfunctioned, the sensors would identify which injector was administering the wrong amount of fuel and correct it without disturbing the others. Current technology would apply a correction to all the injectors, thereby upsetting operation of the ones that are functioning properly.

Pisano’s team of graduate students has started creating robust, millimeter-long sensors out of silicon carbide, a hardened ceramic used in the metal-working industry to cut steel. So far, their sensors have endured temperatures of 1000 degrees Fahrenheit and being shot out of a cannon, a shock amounting to 65,000 times their weight. Next, the sensors will have to survive pressure pulses and corrosive exhaust.

“We’ll have to show there’s still enough accuracy in the sensors after we beat them up in this environment,” Pisano says. Early versions of the prototypes are starting to show success, he says, and the team hopes to create chips with multiple sensors connected by high-temperature wire within a year or two.


UC Berkeley mechanical engineers (from left) Robert Dibble, Albert Pisano and Van Carey say that smart technology can help drivers improve their efficiency, economy and environmental footprint.

The sensor-equipped engine creates new opportunities for passenger cars to not only self-regulate but also convey data wirelessly. With today’s gasoline and diesel engines, selecting the right fuel is pretty much a no-brainer. But the HCCI engine can operate on almost any fuel, and with new biofuels now in development, consumers will need help as choices at the pump explode in the next five years.

Van Carey envisions a not-too-distant future when smart gas stations will assume an entirely new role as virtual diagnostic mechanic, suggesting optimal fuel choices and other ways to improve a car’s performance. Carey is investigating energy strategies and devising computerized systems capable of reading data from the car. In cooperation with the engine’s sensors, the gas pump could check the car’s tire pressure, air filter, recent driving history, fuel use and the condition of the belts and catalytic converter, among other systems.

“The station could suggest ways to maintain the car at a high level of performance and efficiency,” Carey says. The technology already exists to check tire pressures and identify the fuel in the car’s tank. The main speed bump, he says, is getting the sensors to recommend the best fuel blend. “We’re just beginning to develop new fuel options, so there’s going to be a learning curve on what these fuels are and how they perform.”

Berkeley’s assembly of experts in combustion engines, sensors and energy infrastructure will undoubtedly play a significant role in helping the auto industry—and motorists—grow greener. If they have their way, a decade down the road, a car’s biggest selling point may prove to be its IQ.

Erica Klarreich is a Berkeley-based freelance science writer and the mathematics correspondent for Science News magazine. Her work has appeared in Nature, New Scientist, American Scientist and The Sciences.