This article highlights the new ignition schemes for Otto cycle engines that seem to be bound for extinction. Ever since Nicolaus Otto demonstrated the first working four-stroke engine in 1876, engineers have been struggling to come up with ways to sidestep a fundamental limitation of an otherwise stellar design. The reciprocating engine is capable of generating high pressure with reliable sealing, but the volume swept out by the piston has had to remain fixed. Small engines use less internal reciprocating mass than large ones, so the energy to overcome friction decreases as size drops. Small engines are lighter than big ones, too. By recirculating exhaust gases back into the combustion chamber, however, Mitsubishi uses the exhaust to reduce NOx. Because the air-to-fuel ratio is so high, the exhaust gases, which normally hinder combustion, can be as much as 70 percent of the cylinder volume. At the same time, Mitsubishi uses a lean NOx catalytic converter.


As an automobile mover, the Otto cycle engine seems bound for extinction, its critics say. Meanwhile, backers of alternative systems proclaim fuel cells, diesel engines, and hybrids as the path to clean, efficient mobility.

In its own way, however, the ubiquitous Otto cycle engine thumps along another trail to survival. It follows a course of incremental refinement, begun a century ago. Here is a sampling of what a few manufacturers and researchers have been up to lately with regard to the lowly spark ignition engine.

Engineers at Sweden’s Saab Automobile SA recently unwrapped a concept they call the SVC engine, with claims that it will reduce fuel use by 30 percent compared with a larger, naturally aspirated engine of similar output.

The Saab variable compression engine sprang from seeds germinated back in the early 1980s, according to a Saab spokesman, Kevin Smith. By the end of the decade, a patent application had been submitted, and a prototype was on the way to completion. By the mid-1990s, a second experimental engine was ready for testing.


Ever since Nicolaus Otto demonstrated the first working four-stroke engine in 1876, engineers have been struggling to come up with ways to sidestep a fundamental limitation of an otherwise stellar design. The reciprocating engine is capable of generating high pressure with reliable sealing, but the volume swept out by the piston has had to remain fixed. Thus, carmakers have had to build engines tuned to work most efficiently during periods of high loading.

A car, though, does not spend all its time climbing hills or racing away from a stop. Much of the time the engine is operating below peak demand, and below peak efficiency.

Compression ratio sets the level to which a piston compacts a mix of air and fuel before a spark ignites it. Make the ratio too high and the fuel ignites itself, as in a diesel engine. This leads to the formation of multiple flame fronts in the cylinder, as part of the fuel self-ignites and another part is lit by the spark. The resulting knock can ruin an engine.

According to Smith, a typical car engine runs at a fixed compression ratio, somewhere around 9.3:1. If the ratio could be changed continuously, he said, an engine would run more efficiently at light loads with a reduced incidence of ignition knock at heavy loads. Increasing the compression ratio to as much as 14:1 for light loads is ideal, he said. For heavy loads, dropping the compression ratio as far as 8:1 nearly stops knock altogether.

Saab’s variable compression engine does just that. According to Olle Englund, chief engineer for engines at the company, Saab engineers invented a way to vary cylinder volume without resorting to variable height pistons or eccentric connecting rod bearings. Instead, engineers created the “monohead.” This assembly combines cylinder head and cylinder walls in a single unit.

The monohead pivots on a pin inside the crankcase. On the side opposite the pin, an actuator rotates a cam, driving connecting rods to cant the monohead by as much as 4 degrees, Englund said. Tilting the monohead to any spot along 4 degrees of arc adjusts engine compression to an infinite number of ratios between 8:1 and 14:1.

Showing Off the New Engine

Saab unveiled its new five-cylinder SVC engine at the Geneva Motor Show in February, and displayed it again at the recent New York Auto Show. Smith and Englund, along with Lars Bergsten, who manages Saab’s power train center, were on hand giving the press a preview.

The engine’s most distinguishing feature is the elastomer joint winding entirely around the monohead. The band seals the gap between monohead and crankcase, while flexing as a boot to accommodate the monohead motion. Because the high pressure of combustion is limited to the cylinder, the band merely has to contend with the lower pressure of the crankcase, Smith explained.

Another unusual characteristic is the choice of five cylinders—not six or four. The second prototype engine, built after Saab proved the concept on a 2.0-liter power plant, was a 1.4-liter, in-line Six. The goal of the engine designers was to compete with the output of a naturally aspirated, 3.0-liter V6, Smith said, which the 1.4-liter prototype did. To make room in the crankcase for the SVC’s articulated arms, plus their shaft and its actuator, Saab engineers decided to pull two balancing shafts out of the crankcase.


Losing the luxury of active balancing, Saab engineers instead turned to an inherently balanced engine design. Thus, they elected to use a 1.6-liter, in-line Five, which better addressed the needs of Saab’s product line anyhow

Bergsten said that the true benefit of the SVC engine comes when it operates with supercharging. Using variable compression, a naturally aspirated engine could only surpass the efficiency of a conventional engine by 4 or 5 percent, he said. By boosting the pressure of intake air, however, the cylinder can burn more fuel during each ignition and so produce more power in every stroke.

The SVC engine accepts up to twice the boost pressure that Saab cars equipped with high-output turbochargers can handle today. Smith said boost pressure for the SVC engine was so high that the company could not find a turbocharger that met demand which did not suffer from turbo lag. Instead, Saab engineers chose a supercharger to boost ambient air by as much as 40 psi.

That kind of boost generates significant heat. But the monohead design eliminates having to bolt through the cylinder head and cylinder to attach the assembly to the crankcase. Losing the head bolts in the cylinder block opens the entire surface of the cylinder walls to water jacketing, Smith said. As a result, the SVC tolerates the higher heat of the augmented supercharged combustion.

Smith said that eliminating the through-bolts also closes up the space between the cylinder valves. Designers of the engine could thus move the valves closer to the engine centerline, increasing efficiency in the train.

On the subject of valves, Bergsten said that an early design constraint in developing the SVC engine was to retain as many conventional engine components as possible. With the most significant design changes occurring below the combustion zone, Saab engineers could continue to use the design of the four-valve combustion chamber, with its many fuel-burning efficiencies that have evolved over the years.


Bergsten pointed out another novelty in the SVC design: no head gasket. Because the cylinder head and cylinder walls are one piece, the same design that eliminates through-bolts also makes one fewer joint to seal. And with cylinder pressures in variable compression engines capable of reaching new highs, eliminating the potential for seal failure goes a long way toward furthering engine reliability.

As for the intake system, Saab uses supercharger intercooling to extract heat from the boost air. With a dual throttle arrangement, the engine, when only lightly loaded, bypasses the supercharger.

Despite these radical changes to the engine’s physique, however, the idea of variable compression that could adjust to an ever-changing demand would have been impossible without sophisticated controls. Saab engineers adapted a modified version of its Trionic system, which has been used on the company’s turbocharged cars since 1991, Smith said.

Advantages of Variable Compression

Being able to run an engine under higher loads more of the time results in better efficiency. For one thing, an engine under high load is operating wide open, unthrottled. There is no restriction on the passage of air through the intake. Letting up on the gas throttles the intake, pinching down the area through which air enters the engine. The piston thus sucks harder against the constricted passageway, leading to so-called pumping losses. Running the engine wide open more of the time means less energy is spent working to draw air in.

Small engines use less internal reciprocating mass than large ones, so the energy to overcome friction decreases as size drops. Small engines are lighter than big ones, too.

That means a variable compression engine can operate under light loads at a higher compression ratio than a conventional engine and get an attending jump in efficiency. It can accommodate greater boost pressure. A smaller engine can sit in for one nearly twice its size, operate for more of the time under wider throttle openings, and relinquish less energy to friction and weight.

There’s still another advantage to variable compression. It widens an engine’s capacity to operate on a range of fuels.

The Saab people said that the SVC engine is a prototype. More work remains before it bolts onto engine mounts in any of the company’s production vehicles.

“The SVC represents a decisive step in the long-term development work aimed at combining the benefits of the Otto engine and the diesel engine,” Smith said.

HCCI Combustion

Meanwhile, engine researchers are experimenting with another internal combustion cycle. Still in its infancy, homogeneous charge compression ignition, or HCCI, may one day combine diesel efficiency with low NOx and particulate emissions.

Salvador Aceves, an engineer at Lawrence Livermore National Laboratory in Livermore, Calif., has been investigating HCCI for the last five years. The clean-burn promise of HCCI combustion stems from a homogeneous fuel and air mixture that limits NOx and soot formation, Aceves said. “Everything is perfectly mixed,” he said. “There are no droplets. There is no stratification at all.”

The HCCI cycle makes use of the very same phenomenon that is so detrimental to spark-ignited engines: autoignition. One of the reasons spark-ignited engines suffer efficiency-wise is that the compression ratio is limited by knock, a phenomenon associated with autoignition.

Engine knock damages pistons because of high heat transfer, Aceves said. During normal combustion, a cooler layer of gas clings to the walls of the cylinder and the surface of the piston. When the fuel self-ignites, it sets up pressure oscillations within the cylinder that sweep aside the protective boundary layer of cold gas. The fuel mixture, which burns at a temperature of 2,000°C or more, is hot enough to scorch the exposed aluminum pistons.

It is the use of a lean fuel mixture in the HCCI engine that prevents the autoignition of the fuel from damaging the engine. Because the fuel-air mixture has a high air content, and because fuel is fully mixed with air before it reaches the combustion chamber, the fuel burns cooler.

Cooler burning not only brings the burn temperature below the threshold where it damages pistons, it reduces NOx production.

Unlike a spark-ignited engine or a diesel engine, an engine that relies on autoignition to burn the fuel mixture has no control over when combustion begins. Yet, the timing of the start of combustion is critical for any reciprocating engine.

A spark-ignited engine, Aceves said, initiates combustion by firing a spark plug. A diesel engine uses no spark plugs. Instead, it begins combustion by injecting fuel into the cylinder at just the right moment in the compression stroke. Both spark-ignited and diesel cycles offer exquisite control of the fuel burn.

By modeling the HCCI cycle, Aceves has been investigating ways to predict the moment of autoignition.

“We decided some time ago that HCCI combustion was limited by chemical kinetics,” he said. “In a spark-ignited engine you have turbulence playing a big role.” After the spark plug ignites the fuel mixture, a flame front propagates along the cylinder. “You have turbulence in the cylinder pushing the flame front and doing all kinds of things to it,” Aceves said. Turbulence plays no similar role in HCCI combustion. It is all chemical kinetics.

Initially, Aceves and fellow researchers Ray Smith, Daniel Flowers, Joel Martinez, and Robert Dribble had the idea of modeling the cylinder with a mesh comprising more than 100,000 cells. But even for the DOE’s fast computers, predicting the combustion turned into a tough exercise. “Each cell would ignite whenever its temperature was high enough,” he said. “By knowing every cell and its temperature, we would know the whole combustion process.” But the models required calculations beyond the powers of the DOE’s fastest computers.

“After deciding we couldn’t do that, we realized we only had to divide the cylinder into zones that had a given temperature,” Aceves said. He said that 10 zones were enough to predict maximum pressure, burn duration, indicated efficiency, and combustion efficiency. Increasing the number of zones to 20 added little to the accuracy of the model.

Together with his mechanical engineering colleagues at the University of California at Berkeley and the Lund Institute of Technology in Sweden, Aceves has verified the predictions experimentally on a single-cylinder engine operating on natural gas. Data from the model and the experiment have correlated quite closely, he said, especially with respect to pressures and the length of burn.

The university’s laboratory has also begun experimenting with a four-cylinder HCCI engine. Predicting the combustion process for this engine turns out to be more complicated than the single-cylinder problem. Heat transfer in a multicylinder engine is harder to stabilize, so finding ignition points is tougher as well.

“In a multicylinder engine, if you have one cylinder that’s, say, 5°C hotter than the other one—which isn’t very much—that may be enough for it to burn way ahead of the other cylinders,” Aceves said. “Or maybe the colder cylinders won’t burn at all.”

Together with ignition timing concerns, Aceves worries that the power output from an HCCI engine would be lower than an equal size diesel, “because the peak pressure limits how much power you can get from a given engine carcass.” Because of the high peak pressure, an HCCI engine would need to reduce its output below a diesel of the same displacement. That would be a bigger consideration for mobile applications than it would for stationary ones, he added.

Like the Saab SVC engine, an HCCI engine could use a multitude of fuels. “HCCI is very much fuel flexible,” Aceves said. The engine could run everything from diesel to natural gas simply by changing the compression ratio. “Diesel is easy to ignite while natural gas is difficult,” he said. “With diesel, you’d go to a compression ratio of 8:1 or so, which is a problem because it reduces efficiency. With natural gas, a compression ratio of 19:1 or 20:1 results in very high efficiency.”

In fact, Aceves and his colleagues considered variable-compression engines as one possible way of controlling initiation of an HCCI cycle. He said the spark-ignited Saab engine was “good as another way of optimizing efficiency in the low-power range, where most driving occurs.” But he suspected the high pressures of the HCCI cycle might overload the articulated arm that raises and lowers Saab’s monohead. “For a spark-ignited cycle, I think it’s a good concept,” he said.

Aceves said he believed the HCCI cycle could eventually meet any emissions restrictions tossed its way.

Gasoline Injection

As for gasoline injected engines, Aceves said, “Mitsubishi has been working for some time on a gas engine where the fuel is injected late.” But that results in a stratified charge, he said, where one part of the cylinder is rich and another part is lean.

“When you stratify the charge you have the same problem you have with diesel. You have areas in the cylinder that are going to burn hotter than other areas do, and that makes NOx. If the temperature is higher than 2,000 kelvins, you make NOx. That’s why those engines are not sold in this country; they make NOx,” he said.


Still, Mitsubishi announced the production of its 700,000th gasoline direct injection engine this past February. The company began making the GDI engine in 1996. In addition to providing the power plant in its own cars, Mitsubishi reported that it is supplying the engine technology to European carmakers Volvo and Peugeot SA, as well as to Korean manufacturer Hyundai Motor Co.


In its 1999 environmental report, Mitsubishi described the technology as a direct injection gasoline engine that uses an advanced method of controlling the air-fuel mixture. Unlike diesel fuel, gasoline is fussy about air-fuel ratios, making combustion possible only within a tight range, Mitsubishi said. Soot in a diesel engine is the result of partial combustion of the diesel fuel in a low-air mixture. Soot is less of a problem with gasoline because the mix stands a lower chance of partial combustion.

According to Mitsubishi, earlier approaches to gasoline injection located spark plugs and injectors close together to ensure that the fuel mix ignited. But, liquid fuel would coat the plug, making it smolder. For its gasoline diesel injection engine, Mitsubishi increased the distance between plug and injector, then aimed the spray at a depression in the piston. By using what the company calls swirl injectors, special spherically crowned pistons, and upright intake ports, the GDI engine controls the formation of air-fuel mix in the combustion chamber with high precision.

Thanks to gasoline injection, the engine operates in two combustion modes: stratified charge and homogeneous charge. In the stratified-charge mode, used in steady state driving, the injectors fire late in the compression cycle to place an “optimally stratified air-fuel mixture beneath the spark plug.” This is the mode for lean combustion and efficiency, Mitsubishi said. It is also the mode the engine assumes after it is started, to quickly heat the catalytic converter.

Under acceleration, the GDI engine injects fuel during the intake, producing a homogeneous mixture in the cylinders. The fuel spray vaporizes, cooling the air in the cylinder and contracting it, thereby improving volumetric efficiency. Cooling the intake air prevents knock and increases power, Mitsubishi said.

Mitsubishi has also used gasoline injection technology to develop two-stage mixing as a way of preventing ignition knock. According to the manufacturer, two-stage mixing divides the total cylinder fuel charge into two shots. The first is injected during intake. The second is injected during compression.

The initial shot uses about a quarter of the fuel available for the cycle, producing an ultralean mixture. The second shot creates a rich mixture immediately before ignition. Since neither mixture is the ideal 14.6:1 ratio for gasoline and air combustion, knocking is less likely. Two-stage mixing also prevents soot.

In response to criticism that the GDI engine, while low in C02 emissions, emits more than its share of NOx and HC compared with conventional engines, Mitsubishi agreed that three-way catalysts work well if the engine operates at a conventional air-fuel ratio. Such catalysts are less successful with lean mixes, such as the 40:1 air-fuel ratio of the GDI engine.

Better Breathing

By recirculating exhaust gases back into the combustion chamber, however, Mitsubishi uses the exhaust to reduce NOx. Because the air-to-fuel ratio is so high, the exhaust gases, which normally hinder combustion, can be as much as 70 percent of the cylinder volume. At the same time, Mitsubishi uses a lean NOx catalytic converter.

In its environmental report, Mitsubishi wrote, “Utilizing the superb stability of combustion of the GDI engine and mass exhaust gas recirculation to reduce NOx emissions from the engine and fitting a lean NOx catalyst capable of purifying NOx under lean conditions ... enables the engine to meet Japanese and European exhaust standards. In August 1998, low-emission versions of the Galant were launched on the Japanese market, and their emissions are more than 80 percent below the standards.” Junzo Ishino, vice president of media relations with Mitsubishi Motors America of Southfield, Mich., said that gasoline in the United States has a higher sulfur content than fuels sold in Japan and Europe. Higher-sulfur gasoline can poison a catalyst, he said.

“We believe our GDI engine offers a 20 percent increase in fuel economy and a reduction in C02 that is significant in limiting greenhouse gases,” Ishino said.

The company plans to test the GDI engine in California, Ishino said, where the sulfur content of gasoline is lower than it is in the rest of the country.