Rosneft and Eni partner to explore fields in the Barents and Black Seas
Alkaline surfactant polymer flooding shown to improve recovery in very mature oilfield by more than 300%

Delphi and partners advancing with high-efficiency, low-emission gasoline compression-ignition combustion

Engine test results with one of the five injectors for 1500rpm-2bar IMEP, 1500 rpm-3bar IMEP, and 1500 rpm-6bar IMEP. Sellnau et al. Click to enlarge.

In 2010, The US Department of Energy (DOE) selected Delphi, along with partners Hyundai America Technical Center, Inc (HATCI); Wisconsin Engine Research Consultants (WERC); and the University of Wisconsin-Madison (UW) for a $7.48-million grant to develop and to demonstrate a new high-efficiency vehicle concept. (Earlier post.) A key strategy for achieving the project goals is the further development of a new low-temperature combustion system: gasoline direct-injection compression-ignition (GDCI).

Mark Sellnau, Engineering Manager of Advanced Powertrain Technology at Delphi Powertrain, reported on the progress with GDCI at both the SAE 2012 High Efficiency IC Engines Symposium and the SAE 2012 World Congress in Detroit this week.

GDCI, a low-temperature combustion (LTC) process for gasoline partially premixed compression ignition (PPCI), has been under consideration and development for about 5 years, Sellnau said, with efforts predating the 2010 DOE funding. GDCI uses a high compression ratio with multiple late injection (MLI)—similar to diesel—along with intake boost and moderate EGR for high efficiency with low NOx and PM over the entire speed-load map.

The relatively long ignition delay and high volatility of pump gasoline combined with an advanced injection system and variable valve actuation provides controlled mixture stratification for low combustion noise.

Among the objectives of the work reported in the paper presented were 1) to determine the best injection strategies for low NOx and PM using low-to- moderate injection pressures; and 2) to evaluate an engine concept for full-time operation over the speed-load map from idle to full load. Use of variable valve lift profiles was instrumental in enabling full-time GDCI operation, the team noted. The team developed and tested five different injectors.

Sellnau reported results derived from testing on a single-cylinder research engine. The cylinder head of the Ricardo Hydra light-duty single-cylinder engine has four-valves with double-overhead camshafts and central injection. The aluminum cylinder head is rated at 200 bar peak cylinder pressure (PCP). For all tests, intake air temperature was 50 C.

  • At a low-load condition of 1500 rpm-2 bar IMEP, Delphi used a secondary-exhaust-valve-lift event to rebreathe hot exhaust gas and promote autoignition. A “BDC” (Bottom Dead Center) intake cam was also used to maximize the effective compression ratio. Even though heat losses increased somewhat due to the rebreathing, they obtained good indicated specific fuel consumption (ISFC) of about 230 g/kWh, stable combustion, and exhaust port temperatures of about 250 C.

  • Sellnau
    Example of a triple-injection process. The piston is seen rising in each frame. Sellnau et al. Click to enlarge.
    At a medium-load condition of 1500 rpm-6 bar IMEP, injector developments combined with a MLI strategy (triple) and low swirl produced the best ISFC and lowest smoke. The most advanced injector design did not require swirl to achieve very low smoke and NOx levels. Measurements of exhaust particulate size distribution showed that very low PM emissions could be obtained with this combustion system.

  • At higher loads, late intake valve closing was used to reduce cylinder pressure and temperature, and increase ignition delay. Delphi obtained a minimum ISFC of 181 g/kWh. Combustion noise, maximum pressure rise rate, and ringing intensity were in acceptable ranges, however, the correlation among these noise parameters was poor.

  • For IMEP from 2 to 18 bar, engine-out NOx and PM emissions were below targets of 0.2 g/kWh and 0.1 FSN, respectively, indicating that aftertreatment for these species may be reduced or eliminated.

  • Measurements of exhaust particulate size distribution indicated very low particle count, especially for a preferred injector with low levels of in- cylinder swirl.

Overall, single-cylinder engine tests of a GDCI combustion system indicate good potential for a high- efficiency, low-emissions powertrain. Additional testing and development on a multi-cylinder engine is needed, including cold-starting and transient operation.

—Sellnau et al.

In his talk at the High Efficiency IC Symposium, Sellnau said that a modeled 1.8L GDCI engine for vehicle simulations showed large regions with fuel consumption of less than 190 g/kWh: loads of 6-20 bar, and speeds of 1800-3500 rpm. Applied in a mid-size passenger car, such an engine could potentially—without optimization or with a start-stop system, although with variation of gear ratios and shift schedules—deliver 60% improvement in city driving fuel economy, and 40% on US06, for a combined fuel economy improvement of about 51%, Sellnau said.


  • Sellnau, M., Sinnamon, J., Hoyer, K. and Husted, H., Full-Time Gasoline Direct-Injection Compression Ignition (GDCI) for High Efficiency and Low NOx and PM. SAE Int. J. Engines 5(2) doi: 10.4271/2012-01-0384.



I always wondered why we don't make diesel engines powered with gasoline, this study shows that it can works even if soot, NOx and volatility of gasoline are a problems in the 1st place

Roger Pham

Technologies for high-quality high-pressure gasoline injector and for variable valve lift and timing and for precision electronic monitoring were not available in the days when the diesel engine was invented, until recently with the motivation for engine efficiency. Necessity is the mother of invention!

This is quite remarkable that the NOx emission is below 0.2gm/kWh WITHOUT post-combustion treatment. For a car at 330 Wh/mi, this equates to Tier 2 Bin 5 or better for NOx. For cars at 250 Wh/mi or better, this can be comparable to the best of current gasoline cars with 3-way catalytic converter. Quite an achievement!

Regarding the issue that you raised with the Tour engine, this engine uses Atkinson cycle at high loads to maximize peak efficiency, while uses less asymmetrical com/exp at part loads to raise compression ratio, thus can eliminate the disadvantage of Atkinson-cycle engine at low part loads as I've discussed in connection with the Tour engine. 230 gm/kWh at 2 bar-BMEP is very very good! Adding a turbo for Miller cycle and one can have it all! High peak efficiency with high part-load efficiency.


Smoke, though... black carbon is the new bogeyman.  If a particulate filter can catch it, this looks a lot better.


Small particles emitted by ICEs (all types) and coal fired power plants etc nay be responsible for children born with major defects such as Autism etc.

One year after the BP Gulf of Mexico disaster, lots of fish and seafood are born without eyes and other major defects.

Is there a similar relationship with air borne particles and our new borne?


It is a low temperature engine with premixing so that is why NOx and soot are low, avoiding too high compression ratio also helps in that regards. last the high volatility of gasoline probably helps avoiding soot formation


What is it that is allowing these new engine types now - is it high speed engine management systems, or sensors, or expensive gasoline or what ?

Roger Pham

High-speed engine management system and sensors are crucial to allow these achievements in compression ignition of homogenous charge in preventing destructive detonation. Variable valve control is another necessity. High pressure and precision GDI injector is another requirement for high efficiency.

But most important of all, expensive gasoline is the mother of all the above, for these above technologies are expensive and if gasoline remains cheap, no one would consider any of this.

Roger Pham

Avoidance of misfire and formation of a homegeneous charge to lower emission is another challenge in HCCI of gasoline that would require sophisticate engine control technologies, beside detonation prevention.


If smoke was the cause of autism, you think we'd have had a lot more of it before the Clean Air Act.

We didn't.

Assortative mating seems a much better explanation.


Micron sized particles in the air from diesel soot definitely plays a role in the occurance of lung cncer.

The comments to this entry are closed.