Mazda’s New Skyactiv-X Engine Gives New Life to Internal Combustion

For the first time, an engine combines the efficiency of diesel with the cleanliness of gasoline


There are lots of reasons why we’re not all driving electric vehicles now. You’ve probably thought of two or three already, but let me add one that I’m sure you haven’t. It’s a big obstacle to EVs, and it’s rarely remarked upon.

It’s the internal combustion engine, which is no sitting duck. It’s a moving target, and a fast-moving one at that.

There’s no better example of this agile, relentless progress than Mazda’s Spark Controlled Compression Ignition (SPCCI) system, which is scheduled to reach the car-buying public in the form of a new combustion engine in late 2019. Mazda borrowed a trick from the diesel engine, which compresses a fuel-air mixture to the point of ignition rather than igniting it with a spark plug, as gasoline engines do. It’s the biggest advance in combustion engines since electronic fuel injection, which started proliferating in the 1970s.

The new engine operates under some conditions with compression ignition, like a diesel engine, and at other times with spark ignition, like a standard gasoline engine. It will sell under the name Skyactiv-X, building on Mazda’s current engine design, known as Skyactiv-G (G is for gasoline). “We’ve dubbed it Skyactiv-X because it is kind of the intersection of gasoline and diesel technologies,” said Mazda power-train engineer Jay Chen, in a press briefing.

Mazda claims that the 2.0-liter four-cylinder Skyactiv-X provides from 10 to 30 percent more torque and from 20 to 30 percent better fuel efficiency than the Skyactiv-G. So, using the 2.0-L Skyactiv-G as the reference, figure on torque somewhere between 224 and 264 newton meters (165 to 195 foot-pounds) for the Skyactiv-X. If you put it in the Mazda3, a compact car, and assume it has only a minimal hybrid-electric design, then its fuel economy should come to between 6.36 and 5.88 liters per 100 kilometers (37 and 40 miles per gallon). Mazda has not yet announced which model will debut Skyactiv-X.

True, an all-electric car posts better numbers. The U.S. Environmental Protection Agency gives the Chevrolet Bolt EV the e-car equivalent of 119 mpg (1.98 L/100 km). On the other hand, the Bolt will go just 383 km (238 miles) on a charge, while the Mazda3, using today’s Skyactiv-G engine, can manage 785 km (488 miles) on a tank of gas.

“The biggest thing I believe Skyactiv-X does is demonstrate that the internal combustion engine is not dead and that EVs are not a shoo-in,” says George Peterson, president of industry consultancy AutoPacific. “There’s a lot of life left in internal combustion power trains until cost and range issues with EVs are solved.”

To understand how SPCCI works, start with the fundamentals of ignition in the three kinds of combustion engine—the diesel engine, the standard gasoline engine, and the immediate forerunner to the SPCCI, called the homogeneous charge compression ignition (HCCI) engine.

In ideal combustion, each hydrocarbon molecule is paired with an oxygen molecule, producing water and carbon dioxide. The molecules are present in the chemically correct ratio that engineers describe as lambda 1. In a lean fuel condition, when there’s more oxygen, lambda is greater than 1. That’s good when the goal is to reduce fuel consumption. And, because such lean combustion mixtures burn cooler than those at lambda 1, they produce less nitrogen oxide pollution.

However, it’s not always easy to get that lean mixture to burn. “The less and less fuel you have in a mixture, the harder and harder it gets to ignite,” Chen explains. “Just like lighting your barbecue without enough lighter fluid.”

The solution, employed in both HCCI and SPCCI engines, is to keep compressing the air-fuel mixture until it is so hot and under so much pressure that it detonates spontaneously. Diesel engines also use such compression ignition, but they first compress pure air into the combustion chamber, then inject the diesel fuel. Only then does the fuel burst into flame.

This sequence is important because the fire starts at the spot where the fuel is injected and spreads to the rest of the combustion chamber. High temperatures in this expanding flame front cause diesel’s characteristic emission of soot particles and nitrogen oxides.

In HCCI combustion, air and fuel mix together in the cylinder during the compression stroke and spread homogeneously throughout the combustion chamber, as they would in a direct-injected gasoline engine. Only after that spreading and mixing are they compressed to the point of autoignition, as in a diesel engine.

So, in a traditional gasoline engine, combustion begins at the spark plug; in a diesel, it begins at the fuel injector; and in an HCCI engine it happens in all parts of the combustion chamber at once. That makes for an intense explosive reaction, one that puts more downward force on the piston during the engine’s power stroke than the other two engine types do. Gasoline and diesel engines both must light the fuel while the piston is still moving upward on the compression stroke, achieving peak cylinder pressure while the piston is close to the top of its stroke.

“That means the piston is still moving up, already building pressure,” says Chen. “The piston has to fight against the current, if you will, of the pressure.”

“If we did compression ignition, it happens over such a short period of time, we can actually target the peak of the pressure right after top dead center of the piston,” Chen continues, using the industry term for the point when the volume of the cylinder is at an absolute minimum. That way, “all the energy is released immediately, and bam!—the piston just pushes down with the greatest amount of force. For the same amount of fuel, we can get a much higher pressure out of our combustion process through compression ignition than we can through traditional spark ignition.”

To make it work, HCCI engines need to run at a very high compression ratio, just as diesel engines do. According to Sandia National Laboratories, one of the few outside sources that gives numbers, HCCI engines typically run at compression ratios as high as 14:1. Conventional turbocharged gasoline engines commonly run at around 10:1, while diesels, such as the familiar Cummins 5.9-L turbo diesel installed in Ram pickups, run at 17.2:1.

However, HCCI engines can’t always time that spontaneous explosion so that it happens just after the piston passes top dead center in its stroke and begins moving downward on its power stroke. They simply can’t be designed to exert such precise control, because they’re harnessing highly exothermic chemical reactions that behave chaotically, in a fast-changing environment.

As Chen puts it, “Whenever the air and the fuel inside the cylinder reaches a critical temperature and pressure, it’s just going to go boom.”

Because HCCI combustion is possible under only the right conditions of load and engine speed, HCCI engines need spark plugs to let them run in conventional, spark-ignition mode as well. And here is where the challenges begin. In an HCCI engine, compression ignition is spontaneous, so it is difficult to know exactly when the cylinder’s air and fuel mixture will ignite. If that rapid, forceful combustion that we prize so much during the power stroke occurs too early, while the piston is still rising for the compression stroke, catastrophic engine damage could occur. But variations in engine load, throttle position, and temperature make it difficult to rule out such premature ignition if some combination of those factors suddenly creates a compression ratio high enough for compression ignition.

Mazda finesses the problem by having the engine initially give just a very small squirt of fuel. That trick ensures that the mixture is so lean, regardless of conditions, that it will never preignite. “Then, during the compression stroke, we give a larger injection of fuel, under higher pressures. That atomizes, but it doesn’t have the same amount of time to heat up. In that way, it doesn’t have enough time to reach the autoignition temperature threshold,” explains Chen.

How, then, to get this lean mixture to light at the most opportune moment in the cycle? Mazda’s creative solution to this problem is to build its SPCCI engines with a compression ratio of about 16:1—just below the threshold for compression ignition in this engine.

The earlier, HCCI engines needed a spark plug for conventional operation when the temperature, engine load, throttle position, and rpms were unsuitable for compression ignition. But Mazda’s engineers realized that by manipulating conditions within the compression chamber, they could use that spark plug to ignite a local fire within the chamber. The expanding flame front increases pressure throughout the combustion chamber, effectively raising the compression ratio high enough to trigger ignition in all parts of the chamber at once.

That left the lighter-fluid problem: How do you light that compression-enhancing fireball in a fuel mixture that’s too lean to catch fire? Mazda’s solution is to create a region near the spark plug that’s just a bit too lean to catch fire by compression alone. The spark can then set off a fireball whose expansion will boost pressure throughout the cylinder and cause compression ignition. In other words, the spark doesn’t so much light the fire as help the fire to light itself.

Creating such a local less-lean zone isn’t easy. “We can’t just put fuel in and make it slightly less lean, because it will just mix with [everything else in the chamber],” Chen notes. “In order to cordon off this region of slightly less lean, and very lean outside of that, we introduce cylinder swirl.”

Just as baristas create artistic images in espresso foam, it is possible to induce the air-fuel mixture inside the cylinder to swirl in a very carefully designed pattern. But rather than drawing a whimsical heart shape, Mazda engineers induce the flowing air to swirl like a hurricane, with a placid eye centered on the spark plug.

“We create this swirl inside the cylinder through our port design in the cylinder head and also because we have a lean supercharger that helps deliver a high amount of flow,” Chen says. “The more flow, or the harder it is blowing, the more turbulence and vortex we have.” It is into this walled-off vortex that the Skyactiv-X engine injects a little extra fuel, just enough extra to let the spark plug set off the fireball that triggers the cylinder-wide spontaneous compression ignition at the correct instant.

Other carmakers that have pursued HCCI engines—notably General Motors and Mercedes-Benz—have had some problems in smoothly switching the engine from HCCI mode to conventional spark-ignition mode. Basically, the vehicle would lose some power for about a second as the transition took place. This hiccup was quite noticeable if some combination of driving conditions meant the engine was switching back and forth between modes frequently.

General Motors insists the problems were mere teething pains. “As we showed with the public demonstration of the GM HCCI development vehicles in 2007–2008, drivability and mode transitions are not a major barrier to commercial implementation,” says Paul Najt, GM engine systems group manager. In his view, the main challenge for commercialization is in economically combining HCCI with other technologies, like the selective deactivation of cylinders, to achieve even greater fuel economies.

Mazda’s Chen explains that SPCCI doesn’t have any problem switching from HCCI to conventional spark ignition because it doesn’t turn the spark plug on and off. It simply adjusts how its spark is used—to ignite the fuel mixture or to pump up the pressure so that the mixture ignites itself.

“Because we are running the spark plug all the time, in both compression-ignition mode and spark-ignition mode, we can drastically expand the range of compression ignition throughout most engine rpm and engine loads,” Chen says. “Only at very high engine speeds do we switch back to a spark-ignition mode.”

This is where SPCCI differs from HCCI. “In a traditional HCCI engine, every time it switches modes, there’s a momentary pause,” Chen says. “And that pause causes a stumble in drivability. So every time you step on the gas, you might get one stumble, then another stumble as it transitions modes. And you have this drivability problem, which is why a traditional HCCI engine never made it to the market. It was good in the labs, it was good maybe as a concept, but customers wouldn’t accept it.”

For the SPCCI, Mazda managed to overcome these problems by equipping its engines with fast electronic valve-timing actuators. Mazda also adds sensors that directly measure combustion pressures in each cylinder every time it fires. This high-speed monitoring lets the engine-management computer make adjustments on the next-to-fire cylinder stroke to ensure that it is running optimally.

Overcoming the drivability problems of earlier incarnations of HCCI may have been the most crucial accomplishment to making SPCCI feasible for production. But Chen says he is most proud of the fact that Mazda was able to advance the state of the art in combustion technology while relying almost entirely on existing, off-the-shelf parts.

Mazda is on the small end of car companies. Its sales of about 1.56 million cars a year is dwarfed by Toyota’s 10 million. So Mazda may seem an unlikely candidate to advance the state of the art in internal combustion. But the company has a history of doing exactly this sort of thing. In the 1970s it became the first (and still only) manufacturer to put the Wankel rotary engine into mass production. In the 1990s, it developed supercharged Miller cycle engines, which are relevant to the Skyactiv-X because each engine design employs an engine-driven supercharger to pump a high volume of air into the cylinder. Typical performance-oriented supercharged engines, such as the 527-kilowatt (707-horsepower) Dodge Challenger Hellcat Hemi V8, use the compressor to pack air into the cylinders and so to boost power output.

This air-supply scheme employs an intercooler to help cool the intake charge, just as conventional superchargers do. Much of the incoming air is recirculated exhaust gas. Cooling the air raises its density, which puts that much more oxygen in the combustion chamber.

Skyactiv-X also features a hallmark of economy-focused engines, a hybrid-electric assist motor. The engine is incorporated within a “mild” hybrid power train, which means the electric motor can’t propel the car on its own. A belt from the engine drives the car’s rather small alternator, which is smaller than a “strong” hybrid’s alternator but a bit larger than what you’d find in a conventional car with an ordinary 12-volt battery. In the Mazda, the alternator allows the car to recover a bit of energy during deceleration, store it in the battery and later use it to seamlessly stop and restart the gasoline engine. (Cold starts are performed by a regular starter motor.)

The same advances in digital technology that are boosting the fortunes of EVs are also extending the life span of combustion engines. True, the basic moving parts, such as pistons, crankshafts, and valves, remain largely unchanged, but everything else about the process of capturing energy from burning gasoline is in flux.

Computers are providing modeling and analysis that lend insight into combustion that never existed previously. Indeed, MIT’s Green Research Group has developed a combustion model that can run on PCs. The MIT Engine Simulator (MITES) follows 4,000 chemical reactions that can take place in combustion; this analysis enables it to characterize the operating range of HCCI engines. Other engine-development tricks include using engines with clear quartz cylinders fitted with laser sensors that peer into the fiery cauldron. Of course, carmakers like Mazda have enough computing resources to model complex combustion events before building a test engine, but having a modeling tool that runs on a PC can give others the ability to look into this developing area at much lower cost.

“HCCI engines are more sensitive to the details of the combustion chemistry than [spark-ignition] and diesel engines,” note the MITES developers in a paper describing their tool. “Hence, without a solid understanding of the physical and chemical processes taking place in HCCI engines, it is difficult to develop practical, efficient, and robust engines.”

Other than the cylinder-pressure sensors, all of the Skyactiv-X’s components are substantially the same as those in today’s engines “We did it without reinventing the engine, hardware-wise,” Chen says. “Everything in the engine is a component that existed somewhere on the market before Skyactiv-X.”

That continuity with the past explains a bit of the magic that Mazda has invoked. Internal combustion is no desiccated relic of the past but a living, developing technology. As the heir to untold investments and ingenuity, the gasoline power plant continues to fend off challenges from electric propulsion. It will be in a lot of cars for the next generation of motorists. And for the one after that, too. 

This article appears in the August 2018 print issue as “Not So Fast, Electric Car.”