Forget everything we've said about knock. Knock is now (or soon to be) our best friend. For years, our narrow-minded perspective on making big, dirty, boosted power has always taught us that knock-ping or auto-ignition-was a fate worse than death. Knock happens when localized hot spots in the cylinder ignite the air/fuel charge prematurely-before the spark plug goes off. This can happen as the piston is still on its up stroke and, in terms of pressure, is essentially equivalent to taking a hammer and bashing the piston repeatedly. Which is why it sounds exactly like a hammer whacking against metal.
We've always fought knock with more fuel, so charge and combustion chamber temps are cooled by the extra fuel, lowering the likelihood of unplanned ignition. Richer mixtures are also less combustible. But more fuel means less lovin' for the planet in terms of consumption and all that stuff coming out of the tailpipe.
But now, knock-or to use a better term: auto-ignition-has become a good thing. While I was still in school, we played around with a concept called Homogeneous Charge Compression Ignition, or HCCI. It was a concept that uses auto-ignition instead of spark ignition to save fuel and also clean the sky, but-on a university research budget-the hardware and super-smart engine management systems to make it happen weren't available.
Essentially, if Karl Otto and Rudolph Diesel had ever got together, HCCI would be their bastard love child . As the name states, it's a homogeneous charge, much like a four-stroke gasoline engine. The air/fuel mixture inside the cylinder is, for all intents and purposes, mixed evenly, at a uniform ratio. When a spark is put to it, the charge will ignite like any normal spark-ignition engine, just as Otto intended.
The compression ignition part comes from the Diesel end of the gene pool. Instead of using a spark plug to light the charge, as long as the mixture is lean enough and the compression ratio high enough, the charge will auto-ignite-or knock itself into combustion. Where it differs from the average diesel engine is that the charge is homogeneous. A typical diesel cycle uses a heterogeneous or stratified charge, where the mixture inside the cylinder is non-uniform-a result of diesel injectors spraying fuel into the combustion chamber during the compression stroke, thereby creating localized rich spots sufficient to initiate auto-ignition.
So you can make your engine knock intentionally-what's the advantage? The advantage of HCCI is that it runs on leaner mixtures while still making power. Since the mechanics of lean auto-ignition mean that combustion no longer has a conventional flame front, combustion temperatures are uniform and cooler. HCCI combustion happens almost instantaneously throughout the entire cylinder, instead of a combustion flame front that spreads out from the spark source. Cooler lean combustion results in cleaner NOx emissions and a boost in fuel efficency by 20 to 30 percent. The excess CO and HC from lean combustion can all be cleaned up by a standard three-way catalyst. HCCI is also better than the average modern passenger-car diesel engine, because burning gasoline doesn't create all the soot and smoke that necessitates particulate filters (which need regular servicing). Also, gasoline auto-ignites at lower pressures and temperatures than diesel, so blocks don't have to be as robust and heavy, thus saving weight and space.
But there are some hurdles for making HCCI an operating reality. Making an engine knock is easy, but controlling when and how it knocks (so it happens only as the piston is ready for its expansion stroke) is hard. This is where things like direct injection, variable valve actuation, charge temperature control and even variable compression come into play. Auto-ignition typically happens as pressures and temperatures exceed the knock threshold of the air/fuel mixture. To make HCCI useful, the mixture not only has to be even (so everything combusts simultaneously), but pressures and temperatures must all be controlled precisely. Even changes in inlet and block temperatures can affect how and when a charge auto-ignites.
To this end, engineers have developed concepts like variable compression, an old idea pioneered by Saab (subsequently dropped by GM, go figure) to help manage auto-ignition. Variable compression and valve timing can help control auto-ignition by regulating cylinder pressures via compression ratio or valve overlap and charge scavenging.
So here's the beautiful future. While several auto manufacturers are taking their own approaches to HCCI, it's something they're all looking into seriously. Mercedes-Benz is championing this concept, using it in a new DiesOtto engine tipped for production. It combines HCCI, direct injection, variable valve control, variable compression and-just to make it unimaginably efficient-a turbo and a regenerative starter for possible hybrid operation.
The engine would operate as a standard turbocharged engine during start-up (no cold start-issues common with diesel engines) and high-load operation. At other times, the engine management changes to HCCI mode, operating on part-load at low and moderate engine speeds, like cruising. HCCI is limited to below 4000rpm because of the time required for the combustion event and the limitations of valve overlap at high engine speeds. Variable compression will be working the entire time to provide higher compression during HCCI and off-boost operation, with lower compression at peak boost. The starter motor/generator can help get the vehicle rolling from standstill.
A four-cylinder, 1.8-liter DiesOtto prototype is already running around the Fatherland and is claimed to be making 238bhp and 295lb-ft of torque. Outside the peak numbers, the motor (mounted in an S-Class-sized chassis) still pulls off fuel economy figures in the 39mpg range. Much of this economy comes from the small displacement and low cylinder count, while efficiency for these high-output numbers comes from proper turbocharging and engine management.
While integrating HCCI into your current daily driver is possible, making it reliable and more efficient is probably out of the question. Even without the issues of variable compression, direct injection and turbocharging, finding a standalone engine management system fast and smart enough would be a big hurdle. Then there's the rocket scientist needed to tune the car. So we'll just have to sit and wait for it.