Background
Passive pre-chamber ignition systems have been identified throughout the industry as a promising technology for fuel efficiency improvement and emission reduction. Their performance is quite robust in a controlled laboratory environment, but they have been found to be extremely sensitive to residual exhaust concentrations and the ability to control the fresh air to fuel ratio within the chamber. At the two extreme operating points, low load (idle) and full load, there are different challenges towards the pre-chamber spark plug. In the previous IR&D project (03-R8895), an optimized pre-chamber spark plug design showed potential to work at low load, high dilution operating condition and demonstrated up to 20% faster combustion speed at higher loads in the CFD simulation for a 1.2 liter PSA EB2DTS gasoline engine. This project was carried out to demonstrate the scalability of the SwRI developed novel pre-chamber design on a different light duty gasoline engine platform and validate it with engine testing.
Approach
The project work scope included a combination of simulation and experimental testing to explore the viability of a passive pre-chamber ignition system for improved engine efficiency and combustion performance. The primary engine hardware used for this project was a four-cylinder PSA 1.6 liter EP6CDTx engine. Initial CFD simulations were carried out for five cycles to verify whether the designed pre-chamber spark plug showed stable combustion at low load, high dilution operating condition. The pre-chamber spark plugs (PCSP) were then fabricated for all four cylinders and put into the spark plug hole without modifying the cylinder head.
Accomplishments
The PCSP engine showed faster MFB 10-90 durations than the baseline engine at all operating points. At low loads, the MFB 10-90 durations were 8% faster on average while only 3% faster on average at higher loads. However, the PCSP engine showed 1% higher Indicated Specific Fuel Consumption (ISFC) on average compared to the baseline engine at most operating conditions. One exception was at 3000 rpm, 15 bar case where the PCSP showed a 1% reduction in ISFC.
At higher loads, the PCSP demonstrated an extension of the dilution limit when compared to the baseline engine by up to 3.5%. However, the PCSP also caused uncontrolled, runaway pre-ignition events preventing the full benefit of the systems being realized. A pre-chamber design with the thicker pre-chamber wall, while keeping similar volume, may help achieve the full potential of the PCSP and show greater improvement in efficiency. Later on, engine tests were conducted with the low octane E10 gasoline to achieve knock first before pre-ignition at higher loads for the PCSP engine. In general, the knock tolerance was not improved with the current PCSP engine and even when there was improvement in MFB 10-90 duration with better combustion phasing, it still showed poor efficiency compared to the baseline engine. Further pre-chamber design optimization is needed to fully realize the knock benefit of the PCSP engine at higher loads while still maintaining the current performance seen at lower loads.
SwRI had not previously validated the combustion CFD models with the actual passive pre-chamber engine test data for a gasoline engine. The successful validation of the PCSP engine CFD model with the engine test data at lower loads, including the points which showed multiple misfire events, demonstrated the applicability of high-fidelity simulation tools for future passive or active pre-chamber design projects.