European Transportation Biofuel Consumption Up 78% from 2005 to 2006
Electro Energy to Provide Hybrid APUs to Trucking Industry

MIT Researchers Present Latest Results on Dual-Mode SI-HCCI Engine; Variation in Market Fuels Has Little Effect On HCCI Operational Limits

MIT researchers have been developing a spark-ignition (SI) automobile engine that can, under the appropriate driving conditions, move into a spark-free Homogeneous Charge Compression Ignition (HCCI) operating mode that is more fuel-efficient than the SI mode. The mode-switching capability could appear in production models within a few years, improving fuel economy by several miles per gallon in millions of new cars each year.

Members of the MIT team presented the latest results from the project in two papers at the Japan Society of Automotive Engineers (JSAE)/Society of Automotive Engineers (SAE) 2007 International Fuel and Lubricants Meeting this week.

HCCI engines offer the promise of significant fuel consumption benefits (~15 – 20%) relative to spark-ignition (SI) engines; low levels of NOx emissions compared to SI or diesel (compression ignition—CI) engines; and extremely low particulate emissions compared to diesels.

In an HCCI engine, fuel and air are mixed together and injected into the cylinder. The piston compresses the mixture until spontaneous combustion occurs. The engine thus combines fuel-and-air premixing (as in an SI engine) with spontaneous ignition (as in a diesel engine). The result is HCCI’s distinctive feature: combustion occurs simultaneously at many locations throughout the combustion chamber.

With combustion occuring throughout the combustion chamber in HCCI, there is no need for a quickly spreading flame to combust the fuel charge before a new charge enters the chamber. As a result, combustion temperatures can be lower, so emissions of oxides of nitrogen are negligible (although hydrocarbon emissions are higher). The fuel is spread in low concentrations throughout the cylinder, so the soot emissions from fuel-rich regions in diesels are not present.

However, managing combustion in the HCCI process is difficult, especially given different operating conditions and transients. Specifically, precisely controlling combustion phasing is difficult in the HCCI combustion regime due to the lack of an inherent control strategy, such as spark- or injection-timing, used in traditional engines.

According to Professor William H. Green, Jr. of MIT’s Department of Chemical Engineering, ignition timing in an HCCI engine depends on two factors: the temperature of the mixture and the detailed chemistry of the fuel. Both are hard to predict and manage outside of controlled conditions in the laboratory.

The range of conditions suitable for HCCI operation is far smaller than the range for SI mode. Variations in temperature had a noticeable but not overwhelming effect on when the HCCI mode worked. Fuel composition had a greater impact, but it was not as much of a showstopper as the researchers expected.

In work described in one paper presented at the meeting, “Effects of Variations in Market Gasoline Properties on HCCI Load Limits”, the research team, which included Ford Powertrain Research and BP, measured the impact of market-fuel variations on the HCCI operating range in a production Mazda 2.3L, in-line, 4-cylinder, 16-valve engine, modified for single-cylinder operation.

The intake air and exhaust from the firing cylinder were kept separate from the motoring cylinder flow to ensure accurate fuel/air ratio measurements. Other modifications to the engine included: increasing the compression ratio from 9.7 to 11.1; adding continuously variable cam phasing to the exhaust valve train (continuously variable intake cam phasing was a standard feature on the production engine); and using cams with reduced durations and lifts.

To induce HCCI combustion, the MIT engine uses negative valve overlap (NVO)—the exhaust valve closes early during the exhaust stroke to trap hot exhaust gas residuals. The thermal energy of the exhaust gases induces auto-ignition on the subsequent cycle. The use of cam phasing maximizes the HCCI operating range in terms of the high load limit (HLL) and low load limit (LLL) at each engine speed.

When approaching low load limit, the goal was to minimize fuel consumption. The amount of fuel inducted was directly related the amount of residual mass in the cylinder; higher residual fractions resulted in lower fresh charge fractions. To increase the residual fraction, the exhaust cam phasing was advanced (i.e. exhaust valve closing, EVC, occurred earlier during the exhaust stroke).

At the HLL, the goal was to maximize torque output from the engine by increasing the amount of fresh fuel and air inducted during the intake process. To support more fresh charge, the exhaust gas residual fraction must decrease. This was accomplished by retarding the exhaust cam phasing, i.e. EVC timing was pushed later in the cycle. Combustion phasing to control the pressure rise in the cylinder was retarded by adjusting the intake cam to control the amount of fresh charge entering the cylinder.

The researchers tested 12 different fuels blended from commercial refinery streams, which were designed to span the market-typical variation in the select fuel properties: volatility (measured by the Reid vapor pressure: RVP); Research Octane Number (RON); aromatic content; olefin content; and ethanol content.

The researchers concluded that:

  • The effects of market-typical changes in fuel composition on the operational limits were modest; to a first order approximation, all fuels achieved nearly equal operating ranges. This conclusion is only supported for NVO-induced HCCI. “Employing other controlling methods of HCCI combustion may result in larger and more discernible fuel effects on the HLL.

  • The effect of adding ethanol to the test fuels was insignificant at the LLL and HLL.

  • At the high-load limit, some small fuel effects on the operating range were observed; however, the observed trends were not consistent across all the speeds studied.

  • Within experimental measurement error, there was no change in the LLL among the blends of test fuels.

This research was supported by Ford Motor Company and the Ford-MIT Alliance, with additional support from BP.



Rafael Seidl

Nice work. AVL and many others have been researching control strategies for HCCI on both SI and CI base engines for many years now. This includes mode transitions, what's new here is greater robustness wrt variations in fuel properties.

The biggest hurdles to HCCI implementation, especially the type needed for SI engines, remain high cost (fully variable valvetrains, GDI, in-cylinder pressure sensors, precise EGR control) and high noise: simultaneously igniting all of of the fuel throughout the combustion chamber makes for a very loud bang and also very high mechanical stresses.

Split combustion based on an HCCI event, followed by conventional post-injection has been tried for HDV diesel engines. The objective was to extend the portion of the engine map in which at least partial HCCI would be possible. A big issue was the performance / fuel economy penalty of the suboptimal aggregate fuel combustion curve. Another was continued high cold start emissions, because HCCI still requires a warm engine.


Fascinating! Bravo Ford for funding this research.
I hope HCCI engines become a production reality sooner rather than later. It would seem the big 3 stand to gain the most from implmenting this technology.


sounds interesting, but will probably result in higher costs, if you need to build every gasoline-engine as strong as a diesel engine, to cope with the extra pressure of self ignition


This may cause the big three to delay full-hybrid implementation across their product line. But even so, I don't see a down side. Get immediate gains now without complete retooling, and then the split-mode combustion engine could still be used later in a hybrid drive train for even greater efficiency.



What are higher engine costs compared to a hybrid drive train?


What about emissions? My prediction: You'll never see this thing on the road. Not enough bang for the buck. Well, OK, lots of "bang" but not the right kind.


Yesterday I drove both a petrol car and a diesel utility over the same stretch of winding road. The petrol car was more zippy at all RPM and with the diesel you have to allow for the turbo boost 'shoulder'. They both did the job so while the zippy car was more enjoyable to drive I think people can get used to diesel. I'm assuming BTL diesel will be available in years to come but petrol will be scarce.

However it seems to me that an electric drive hybrid can get both zip and mpg so the IC generator motor does not need to be hi tech.


Another interesting ICE exercise for the engineers; however, I agree with Aussie. You don't need an expensive, complicated ICE with complicated, expensive emissions controls to do middle rpm cruse work and act as a genset for hybrids, especially long range serial PHEVs. And, best of all, the ICE is not needed at all for BEVs.

The market place is straining at the bit for one of the large auto companies to introduce a long range, fairly priced PHEV or BEV. The first one that is priced around $20,000, takes the market and sets the standards.

Roger Pham

Oh yeah? I was able to achieve HCCI as far back as the late 70's in my ~9-mpg Oldsmobile 98 with high-compression-420-cid "Rocket" engine, 11:1 compression and 4-barrels carburetor! Some times, while driving at slow speed, I shut off the ignition just for kicks, and the engine kept running. Not knowing much about engines at the time, I was flabbergasted and really amused, but did not make much of it, since the car was a clunker anyway and closed to be tossed to a salvage yard, back in the days of oil embargo with gasoline ~$1.25 USD/gallon.

But seriously, this achievement at Ford is great stuff! 15-20% increase in cruise efficiency brings its thermal efficiency to the level of diesel, without expensive high-pressure diesel injectors, and without expensive diesel exhaust emission control. Port fuel injection would be OK. Adding cam phasing to exhaust valves would not increase cost by that much. The high compression ratio required for HCCI may be lowered during SI mode by delaying the closure of the intake valve, thereby simulating an Atkinson-cycle engine, or by direct water or alcohol injection at high load limit, as the MIT researchers have announced before. To prevent misfire, the spark plug can be programmed to fire a littler later if spontaneous ignition has not occured, until the exhaust valve timing can be tweaked to achieve contionous HCCI. Retarded ignition is still better than a misfire. Knock sensors already onboard can be tweaked to recognize each HCCI event and timing in the cycle in order to precisely monitor and control the engine's spontaneous ignition. So, may be no expensive sensors will be necessary.

By limiting HCCI to low load limit, the engine already designed for much higher output sure can handle the increase in stress without modification and without being too noisy with some noise isolation carpeting or noise-cancellation speakers inside the cabin.
In fact, my Olds 98 engine knocked most of the time I pushed down on the gas, since I was too cheap to use premium fuel as recommended on it, but it ran OK, well, sort of! I didn't know back then that the engine knocking sound was supposed to be bad until later.

Way to go, Ford! Innovation is job number one!


Yeah, cool, but what does "improving fuel economy by several miles per gallon" exactely means?
there are struggles to reduced GHG emission from engines, not to control NOx where several solutions already exist...
This is a bit disapointing to be so vague in such an important factor as fuel economy.

Rafael Seidl

@ FC -

the only reason for using HCCI in an SI engine is improved fuel economy. The article indicates a potential for 15-20% improvement thanks to this technology. The absolute MPG figure depends on the base engine and vehicle.

Mercedes is further along and feels confident it can claim 6L/100km in the NEDC for its DiesOtto concept, which is still in the R&D stage. This is a staggering 40% less than the roughly comparable 3.5L V6 conventional SI engine MB currently offers as the base for its S-Class in Germany. Of course, the DiesOtto also incorporates a bunch of other, synergistic fuel economy features.

Kevin Moore

I wonder how much additional noise you'll get considering it's only designed to go off spark at low to mid-ranges... I assume their "HLL" point was actually a mid-range point... If you trap exhaust gas at WOT then you lose displacement relative to an engine of the same size that scavenges fully and gets as much O2 and fuel in as possible... Also I wonder how much (if any) additional engine reinforcements would be required vs. a diesel, since the 11:1 CR is modest vs. a diesel, and (again) assuming it goes back to spark at WOT.


Have you look at the videos of the “Bourke Engine” prototype and model? This engine meets “HCCI” engine principles in design, operation, and performance. It is most likely the most powerful green engine every designed and built.

Measured engine efficiency 62%, only 2 moving parts

A 2 cylinder 60 to 80 cubic engine, +140 hp, +100 mpg, very low NOx and CO emissions.

It can't get any simpler than the "Bourke"...

Noel Horne

horsemastership ischiococcygeal megalornithidae queasy alcoholysis phytohormone myringitis presocial
Leopardstown Racecourse

Verify your Comment

Previewing your Comment

This is only a preview. Your comment has not yet been posted.

Your comment could not be posted. Error type:
Your comment has been posted. Post another comment

The letters and numbers you entered did not match the image. Please try again.

As a final step before posting your comment, enter the letters and numbers you see in the image below. This prevents automated programs from posting comments.

Having trouble reading this image? View an alternate.


Post a comment

Your Information

(Name is required. Email address will not be displayed with the comment.)