Purdue University Graduate School
PhD Dissertation - Dong Eun Lee - vF - 072323.pdf (11.3 MB)


Download (11.3 MB)
posted on 2023-08-08, 17:30 authored by Dong Eun LeeDong Eun Lee

In Chapter 2, an experiment has been developed to investigate the passive pre-chamber jet ignition process in gasoline engine configurations and low-load operating conditions. The apparatus adopted a modified 4-cylinder 2.0L gasoline engine to enable single-cylinder operation. To reduce the complexity, the piston position was fixed at a predefined position relative to the top dead center (TDC) to simulate thermodynamic conditions at ignition and injection timings. High-speed Infrared (IR) imaging was applied to visualize the jet penetration and ignition process inside the main cylinder and to investigate the cyclic spatial variability. Two passive pre-chambers with different total nozzle areas and numbers of nozzles were used. In addition, the pre-chamber volume and pressure at ignition timing were varied to examine their effect on jet ignition performance. Misfire behavior was observed in the main chamber of all test cases, and the results suggested that the main cause is a high Residual Mass Fraction (RMF) in the pre-chamber affecting the subsequent cycle. A larger total nozzle area, smaller volume, higher pressure, and fuel-lean operation tended to mitigate the misfire behavior. For a test case with a spark pressure of 6 bar, a reduced cyclic variability in terms of coefficient of variation peak cylinder pressure (COVPmax) from 10.03% to 7.38% and combustion phasing variation from 81 crank angle degree (CAD) to 12 CAD were observed with increasing pre-chamber volume-to-area (V/A) ratio from 59.37 m to 103.11 m, but slightly higher misfire frequency was observed, from 46.67% to 50.00%, suggesting an accurate combination of pre-chamber design parameters is needed to improve overall performance at low-load operation.

In Chapter 3, it examines the influence of passive pre-chamber nozzle diameter and dilution level on jet formation and engine performance. Utilizing a modified constant-volume gasoline direct injection engine with an optically-accessible piston, we tested three passive pre-chambers with nozzle diameters of 1.2, 1.4, and 1.6 mm, while nitrogen dilution varied from 0 to 20%. With the help of high-speed imaging, we captured pre-chamber jet formations and subsequent flame propagation within the main chamber. Our novel findings reveal that asymmetric temporal and spatial jet formation patterns arising from pre-chambers significantly impact engine performance. The larger nozzle diameter pre-chambers exhibited the least variation in jet formation due to their improved scavenging and main mixture filling processes, but had the slowest jet velocity and lowest jet penetration depth. At no dilution condition, the 1.2 mm-PC demonstrated superior performance attributed to higher pressure build-up in the pre-chamber, resulting in accelerated jet velocity and increased jet penetration depth. However, at high dilution condition, the 1.6 mm-PC performed better, highlighting the importance of scavenging and symmetry jet formation. This study emphasizes the importance of carefully selecting the pre-chamber nozzle diameter, based on the engine's operating conditions, to achieve an optimal and balanced configuration that can improve both jet formation and jet characteristics, as well as scavenging.

In Chapter 4, it investigates the influence of passive pre-chamber nozzle diameter on jet ignition and subsequent main chamber combustion under varying load conditions and dilution levels using a constant-volume optical gasoline direct injection engine. The results reveal that as the load decreases, both fuel availability and flow conditions deteriorate, leading to delayed and inferior jet characteristics that affect main chamber ignition and combustion processes. In high and medium load conditions without dilution, the smallest nozzle diameter pre-chamber (1.2mm-PC) shows improved jet ignition and main combustion due to earlier jet ejection, enhanced penetration, and intensified jet. This is facilitated by the smaller nozzle diameter enabling faster and higher pre-chamber pressurization. Conversely, under low load conditions, the largest nozzle diameter pre-chamber (1.6mm-PC) performs better, likely due to improved scavenging and reduced residual levels, resulting in less compromised pre-chamber combustion and subsequent jet characteristics. The nozzle diameter also has a significant impact on cycle-to-cycle variations, with smaller diameters enhancing jet ignition performance but increasing variability. The effect of external residuals (dilution) on jet ignition performance varies depending on the nozzle diameter, with the 1.6mm-PC exhibiting less degradation and demonstrating earlier jet ejection and CA50 timing compared to smaller nozzle diameter pre-chambers at higher dilution conditions. The improved scavenging and relatively lower residual levels in the larger nozzle diameter pre-chamber contribute to its increased resistance to dilution and potential extension of dilution tolerance.

In Chapter 5, it presents an analysis of the effects of pre-chamber nozzle orientation on dilution tolerance in a constant-volume optical engine. Using a combination of experimental and numerical methodologies, we provide novel insights into how variations in nozzle number, orientation, and size influence combustion performance under different dilution conditions. The findings reveal that an increase in the number of nozzles, for a fixed A/V ratio, tends to enhance ignition performance and stability across a range of dilution scenarios, primarily due to an increase in ignition points and a larger ignition surface area. Meanwhile, swirling pre-chambers, despite their potential to boost initial combustion performance at no dilution condition, may limit dilution tolerance due to the complexity of their internal flow dynamics and increased heat loss through nozzle surfaces. Furthermore, pre-chambers combining swirling and straight nozzle orientations fail to synergize the benefits of each type, and instead, exacerbate challenges such as heat loss, flame quenching, and unfavorable flow dynamics. These findings emphasize the complexity and nuanced trade-offs involved in optimizing pre-chamber design for improved dilution tolerance and suggest potential directions for future research in this area.

In Chapter 6, it investigates the behavior of pre-chamber knock in comparison to traditional spark ignition engine knock, using a modified constant-volume gasoline engine with an optically-accessible piston. The aim is to provide a deeper understanding of pre-chamber knock combustion and its potential for mitigating knock. Five passive pre-chambers with different nozzle diameters, volumes, and nozzle numbers were tested, and nitrogen dilution was varied from 0 to 10%. The stochastic nature of knock behavior necessitates the use of statistical methods, leading to the proposal of a high-frequency band-pass filter (37-43 kHz) as an alternative pre-chamber knock metric. Pre-chamber knock combustion was found to exhibit fewer strong knock cycles compared to SI engines, indicating its potential for mitigating knock intensity. High-speed images revealed pre-chamber knock primarily occurs near the liner, where end-gas knock is typically exhibited. The study identified that increasing pre-chamber nozzle diameter resulted in a larger dispersion of knock cycles and more severe knock intensity, likely due to shorter jet penetration depth requiring more time for end-gas consumption. Strategies for mitigating knock in pre-chamber combustion systems include reducing the pre-chamber volume for a fixed A/V ratio and increasing dilution level. The results of this study offer valuable insights for developing effective knock mitigation approaches in pre-chamber combustion systems, contributing to the advancement of more efficient and reliable engines.

In Chapter 7, a numerical investigation of different premixed gaseous injection strategies was performed to understand their impact on the scavenging and mixture formation of an air-fuel premixed pre-chamber with high exhaust gas recirculation (EGR) operations. EGR dilution is effective for reducing coolant heat loss, pumping work at throttled conditions, and mitigates knock at high-load conditions, thus increasing engine efficiency. To further extend the EGR limit of an air-fuel premixed pre-chamber engine, the effects of different injection strategies (including timing, duration, pressure, pre-chamber volume, and hardware) on the EGR level, trap efficiency, and parasitic loss were determined. Regardless of injection duration and upstream pressure, injecting too early not only increased the amount of the injected premixed gas leaking into the main chamber but also was inefficient in reducing the EGR level in the pre-chamber. To reduce the EGR level in the pre-chamber to a level where successful ignition and combustion of the pre-chamber mixture is possible, the injection timing should be delayed to be close to the ignition timing. A premixed air-fuel injection is thus proposed to reduce the time required for air-fuel mixing in the pre-chamber. With a delayed end of injection (EOI), both leakage amount and EGR level were reduced compared to the cases with earlier injection timings. The results show that an injection with 15 bar upstream pressure, 20 CA duration, EOI of −20 CAD aTDC (ignition timing), and with guided injection hardware for the base pre-chamber volume resulted in about 0.17% air compression parasitic loss, over a 94% trap efficiency, at the same time maintaining the mean EGR level in the pre-chamber below 20%, ensuring good pre-chamber combustion. With a 50% increase in pre-chamber volume from the base case, the parasitic loss increased by 65% (from 0.17% to 0.28% loss), indicating a problem with a larger pre-chamber with a separate air valve and injector.


Aramco Americas

Ford Motor Company


Degree Type

  • Doctor of Philosophy


  • Aeronautics and Astronautics

Campus location

  • West Lafayette

Advisor/Supervisor/Committee Chair

Li Qiao

Advisor/Supervisor/Committee co-chair

Gregory M. Shaver

Additional Committee Member 2

Haifeng Wang

Additional Committee Member 3

Terrence R. Meyer

Usage metrics



    Ref. manager