Pre-chamber turbulent jet ignition of methane/air mixtures with multiple orifices in a large bore constant volume chamber: effect of air-fuel equivalence ratio and pre-mixed pressure

Abstract

Liquefied natural gas (LNG), mainly composed of methane, is in progress to substitute diesel fuel in heavy-duty marine engine for practical, economic, and environmental considerations. However, natural gas is relatively difficult to be ignited in a large bore combustion chamber. A combustion enhancement technique called pre-chamber turbulent jet ignition (TJI) can permit combustion and flame propagation in a large-bore volume. To investigate the effect of air-fuel equivalence ratio and pre-mixed pressure on pre-chamber TJI of methane/air mixtures with multiple orifices in a large bore volume, experimental tests and computational simulations were implemented to study the discharge of hot turbulent jets from six orifices of the pre-chamber. Different initial pressures and air-fuel equivalence ratios were considered to analyze the characteristics of TJI. The asymmetry of the turbulent jet actuated from six different orifices were found due to the asymmetric orientation of the spark plug, resulting in the inhomogeneous distribution of combustion in the constant volume chamber, which should be considered seriously in the marine engine design. Besides, as the premixed pressure increases, it has more effect on the flame propagation and plays a more important role, as it further increases.

Keywords

marine engine / natural gas / methane / turbulent jet ignition (TJI) / pre-chamber

Introduction

Internal combustion (IC) engines are still widely used in various applications and will remain widespread for the foreseeable future. Heavy-duty marine propulsion and energy generation is an important application of IC engines, and around 90% of world trade is carried by sea using thousands of merchant ships (International Chamber of Shipping, 2013) [¹]. Research is in progress to substitute the commonly used diesel fuel with alternative fuels for practical, economic and environmental reasons [²]. Liquefied natural gas (LNG), mainly composed of methane, is an attractive option as its combustion inherently produces less CO₂ emissions and particulate matter. A natural gas engine inevitably suffers from low reactivity that requires to pre-heat the intake air to a high temperature to avoid long ignition delay which leads to misfire and combustion instability. However, high NO_x emissions and excessive exhaust gas temperature are generated when operated at rich or stoichiometric mixture of fuel and air [³]. The lean burn strategy, in which fuel combusts in excess air, yields lower combustion temperatures, and results in a higher thermal efficiency due to lower heat losses [⁴]. Fewer oxides of nitrogen (NO_x) emissions are formed when combustion occurs at a lower temperature. However, the lean limit is restricted to the capability of the ignition system to reliably ignite the fuel-lean mixture. To overcome this limitation, pre-chamber type ignition devices [⁴] have been proposed and evaluated to a reasonable extent but only a few have been successfully commercialized.

A combustion enhancement technique named pre-chamber turbulent jet ignition (TJI) can permit combustion and flame propagation in a large-bore volume [⁵]. The main characteristic of the TJI procedure is, by increasing the mixing rate and turbulent intensity, the transient reacting turbulent jet from the pre-chamber plays as a distributed energy source to ignite the homogeneous air/fuel mixture in the main chamber [⁶]. Ricardo first proposed a pre-chamber engine and described it in his paper in 1922 [⁷]. Since this inception, various pre-chamber combustion systems have been developed, primarily to increase the dilution tolerance of engines. Gholamisheeri et al. [⁶] analytically and computationally studied the transient jet actuated from a single orifice to ignite combustible mixtures in the main chamber with TJI. The results of the simulation were compared with high speed chemiluminescence images of TJI in a rapid compression machine. The influences of mixture stoichiometry and orifice sizes on jet penetration rates and combustion characteristics were investigated. Gentz et al. [⁸] analyzed the effect of orifice number and sizes by direct visualization for the premixed propane/air mixture combustion initiated by a TJI system in a rapid compression machine as well. They found that for the condition of near stoichiometric air to fuel ratio, a pre-chamber could produce more spatially distributed jets, which could lead to a faster combustion progression. However, under lean conditions, the orifice with a smaller diameter able to generate a more vigorous and faster jet was needed to initiate the combustion. Mohammad et al. [⁹] numerically modeled the flow field and combustion in a pre-chamber equipped single cylinder engine operated in the reactivity controlled compression ignition (RCCI) mode. Their results showed that the implementation of a pre-chamber could extend the operating ranges of the RCCI engine at lower intake temperatures. However, it was also noted that the proposed combustion strategy led to a lower efficiency, with comparison to the conventional combustion mode, caused by the energy loss of the high speed gas mixture passing through the pre-chamber and discharge orifices. Yousefi et al. [¹⁰] analyzed the effect of diesel injection timing and natural gas energy fraction on emissions and combustion performance of a pre-chamber equipped dual-fuel engine at low engine load. Their results showed that the implementation of the pre-chamber could decrease the unburned methane emissions by around 46%, with comparison to the dual-fuel engine without pre-chamber. There are also a few reports focusing on the influence of pre-chamber parameters on the combustion characteristics of large-bore natural gas engines. Li et al. [¹¹] designed combustion systems for a large-bore natural gas engine with a small volume pre-chamber. The results indicated that the flame radial propagation was accelerated as well as the better anti-knocking ability and emissions when the jet cone angle was larger. Bai et al. [¹²] conducted an experimental study of a single cylinder engine in order to get an improved understanding of the combustion process and provide valuable data to allow validation of the computational models. The results demonstrated that the small volume pre-chamber achieves a higher combustion rate. The discharge nozzle number of the injector should match the air/fuel flow field in order to promote the combustion. Comprehensive reviews of pre-chamber ignition concepts and ignition enhancement techniques have been done by Toulson et al. [⁴] and Alvarez et al. [¹³].

A fundamental understanding of the TJI mixing process in the cylinder is the key for large-bore engine design. The ignition phenomena of the methane/air mixture by the turbulent jets discharged from the pre-chamber must be visualized. Most optical TJI experiments were conducted on rapid compression machines where real engine conditions can be mimicked, especially for vehicle engines with relatively small bores (d<100 mm). However, it is difficult to find experiments conducted on a large bore (d: ~300 mm) rapid compression machine with optical windows, since it is not feasible to construct a rapid compression machine with such a large bore volume. Therefore, a constant volume chamber is probably the only choice to study the TJI in a large bore environment with optical measurement. Biswas et al. [¹⁴] investigated the combustion regimes of a pre-mixed methane/air TJI in a rectangular constant volume chamber (inner dimensions: 304.8 mm × 152.4 mm × 152.4 mm). The effect of different initial pressures (0.1–0.5 MPa) and equivalence ratios (0.45–1) on the ignition performance was studied. They plotted all possible ignition results on the diagram of turbulent combustion regime summarized by Peters [¹⁵]. Almost all the no-ignition cases fell within the broken reaction zone, and most of flame and jet ignition cases were in the thin reaction zones. In addition, with increasing pressure, the flame ignition became more prevalent; the ignition probability will be reduced when decreasing the main chamber equivalence ratio.

The present study focuses on the characteristics of TJI from the pre-chamber with relatively large orifices to the main chamber of a large bore volume. More specifically, the effect of air/fuel mixture quality on the hot turbulent jet propagation characteristics from the pre-chamber is investigated. Computational fluid dynamics (CFD) simulation of the methane fueled TJI system is only implemented to generate more information to facilitate the understanding of the flow field, especially the temperature fields in both the main chamber and the pre-chamber which are not easily measurable by experiment, although the mechanism is probably only valid at low pressures. The study presented in this paper is a part of an ongoing project which aims to deepen the understanding of the characteristics of ignition resulted from pre-chamber devices in practical large-bore marine engines.

Experimental methods

The TJI system studied in this paper has six small orifices which connect the pre-chamber and the main combustion chamber with a relatively small pre-chamber volume. The small orifice makes the burning mixture to be quickly actuated out from the pre-chamber, which seeds the main chamber with reacting active radical species [⁴]. Then the hot products discharged from the pre-chamber ignite the main chamber fuel/air mixture through the chemical, thermal, and turbulence effects away from the pre-chamber, thus producing a distributed ignition system. Turbulent jets which penetrate deeper into the main chamber can also be produced by even smaller orifices. The pre-chamber volume is suggested to be kept relatively small to avoid the impingement on the combustion chamber wall by previous researchers. Attard et al. [⁵] stated that the “benefits of small pre-chamber volumes relative to their larger counterparts include negligible power loss and fewer HC emissions due to the reduced crevice volume and combustion surface area.”

TJI system

The TJI system consists of a pre-chamber mounted on a steel adapter which is mounted at the top of the constant volume chamber. A Teflon gasket is used to seal the pre-chamber. A spark plug (Champion FB77WPCC) is installed onto the top of the pre-chamber. A six orifices injector is installed to the bottom of the pre-chamber, where the orifices are sequentially numbered as shown in Fig. 1. The diameter of each orifice is 2.9 mm, and the total area of these six orifices is 39.63 mm². The volume of the pre-chamber is 15 cm³. A smart high voltage ignition coil with the spark plug is controlled by the data acquisition (DAQ) card (NI PCI-6251) to initiate the ignition in the pre-chamber. A detailed sectional view of the TJI system mounted onto the constant volume chamber is shown in Fig. 1. Once the spark plug in the per-chamber is triggered by the DAQ signal, the methane/air mixture is ignited in the pre-chamber, leading to a rapid increase of pressure in a relatively small volume, resulting in the fact that the pressure in the pre-chamber exceeds the pressure in the main-chamber. Then, the ignited hot products of combustion are discharged from the pre-chamber into the main-chamber and initiate the combustion in the main chamber by the TJI process. In the meantime, the pressure in the pre-chamber decreases until an equilibrium pressure is reached, and the pressure in the main-chamber raises due to the combustion. Each test case is operated three times to demonstrate repeatability.

Constant volume chamber

As it is difficult to build a large bore rapid compression machine with optical windows, all experimental testing in this study was operated in a constant volume chamber whose inner diameter is d = 300 mm, which is large enough for the investigation of the in-cylinder combustion of a large bore marine engine. The designed pressure of the chamber is up to 8 MPa, and two quartz glasses with a visible diameter of 140 mm are mounted onto the chamber to provide a clear view of the turbulent jet flame propagation. To monitor the pressure and temperature variation in the constant volume chamber, a piezoresistive pressure transducer (HM80A) is mounted at the side of the main chamber. In addition, a digital pressure gauge with a range of 0–0.4 MPa, an accuracy of 0.2%, and a resolution of 0.0008 MPa is installed in the chamber to obtain the desired methane-air mixture at different air to fuel ratios (l). The air and methane supply and the discharge processes are controlled by three solenoid valves driven by the DAQ card (NI PCI-6251). A schematic view with the main components labeled and the real photo of the constant volume chamber, are demonstrated in Fig. 2.

Optical setup and test cases

Turbulent jet initiated combustion is directly photoed by a Photron FASTCAM-ultima APX high-speed camera coupled with a Nikon 24–85mm AF lens, at a frame rate of 250 frames per second with an image resolution of 1024 pixel × 1024 pixel. To image the initial illumination of the flame development when the ignited turbulent jets travels through the six orifices from the pre-chamber, the camera lens aperture is set to be completely open.

When the air-fuel equivalence ratio in the main chamber of an actual large gas engine is around 1.8, the calculated air-fuel equivalence ratio in the pre-chamber (with an auxiliary pre-chamber CH₄ inlet pipe) is around 1.0 at the top dead end with the piston compression process. Therefore, different initial air-fuel equivalence ratios (l = 0.8, 1.0, 1.2) were designed to analyze the effect of initial pressures (P_initial = 0.6, 1.0, 1.5 MPa) on the characteristics of the TJI with the pre-chamber, and a set of test cases at different initial conditions as listed in Table 1 are studied. It should be noticed that the pre-chamber and the main chamber are directly connected through the six orifices without any valve engaged. It is difficult to pre-mix CH₄ and air for different fuel-air equivalence ratios in the pre-chamber and the main chamber. Therefore, the initial conditions in both the pre-chamber and the main-chamber are the same.

Experimental results and discussion

Combustion visualization of TJI in the constant volume chamber

Figure 3 displays the TJI initiating combustion in the constant volume chamber at different initial pressures and air-fuel equivalence ratios. The brightness and contrast of the color (RGB) images of the combustion are enhanced by an image processing software Image J, available for download from the website of the National Institutes of Health. The images are cropped by circular outlines to display the optical window boundary of the constant volume chamber which was only illuminated by chemiluminesence light and not clearly recognized in all the frames. It is noteworthy that the frame rate of the camera is reduced to 50 fps to increase the exposure duration for the lean combustion (l = 1.2) when the natural luminosity is weak.

Distributed ignition can be observed throughout the jet from the six orifices. The initial jet velocity is essential as it determines the ability to entrain the unburned mixture and the amount of turbulence that the jet produces [⁸]. For each case with the same initial pressure, it can be seen that the natural luminosity of the turbulent jet is brighter with the decrease of air-fuel equivalence ratios, which is mainly caused by the easier soot formation and their “bright” combustion phenomena in the rich fuel condition(e.g. l = 0.8). Besides, the flame propagation and discharge velocity are faster at l = 0.8 and l = 1.0 than those for the lean case (l = 1.2), which can be easily observed from the sequential images that illustrate the combustion process (Fig. 3). As the initial pressure increases, the discharge velocity and flame propagation are inhibited. In addition, the asymmetry of the turbulent jets is observed, especially for the rich fuel case of P_initial = 0.6 and l = 0.8, where the violent combustion phenomena is circled by the red dashed line (Fig. 3). A detailed discussion of the asymmetry of the turbulent jets will be made in Section 4 with the help of computational simulation.

For the high pressure case (P_initial = 1.5 MPa), the signal of the chemiluminesence light can barely be detected by the current optical setup. The flame fronts for different air-fuel equivalence ratios at P_initial = 1.5 MPa can be drawn as depicted in Fig. 4. The flame fronts of all three cases start as the “w” shape which is caused by the turbulent jets discharged from the spatially distributed six orifices of the pre-chamber. As the air-fuel equivalence ratio increases, the flame front forms a “V” shape (l = 0.8), a “⌒”shape (l = 1.0), and a “─”shape (l = 1.2) with the flame propagation which indicates that during the early stage of TJI, the flame propagation is particularly caused by the hot turbulent jets from the orifices, and during the later stage, the flame propagation speed is mainly dependent on the premixed air/fuel ratio, and again the rich fuel mixture leads to faster vertical flame penetration at the injector center.

Pressure variations in the main chamber

Figure 5 exhibits the pressure variations in the main chamber after the spark ignition from the pre-chamber, where DP/P represents the pressure change rate in the main chamber. The peak pressures (1.2, 2.0, and 3.0 MPa) are nearly twice of the corresponding initial pressures (0.6, 1.0 and 1.5 MPa). The time required to reach the peak pressure after the spark mainly depends on the premixed air-fuel equivalence ratio. Double peak pressures are observed for the lean case (l = 1.2) or the higher back pressure case (P_initial = 1.5 MPa), mainly because of the lower flame propagation speed. The earlier peak was caused by the ignition of the mixture around the six turbulent jets from the six orifices, and the later peak was caused by the ignition of the mixture from other places especially the vertical region from the injector center as the flame front development, as shown in Fig. 4. This phenomena is not found for the stoichiometric case (l = 1.0) and rich case (l = 0.8) at P = 0.6 MPa and 1.0 MPa, which is due to their faster flame speed than that of the lean case. However, two pressure peaks are also observed at high initial pressure (P_initial = 1.5 MPa) for the rich mixture (l = 0.8). This indicates that not only the air-fuel equivalence ratio, but also the back pressure affects flame propagation. This also illustrates that the turbulent jets travel faster than premixed flame propagation for the lean case (l = 1.2) and high pressure case (P_initial = 1.5 MPa). The DP/P plots also prove that increasing the pressure has an obvious effect on the flame propagation, and the pressure change rates are reduced. With the increase in premixed pressure (~1.5 MPa), the pressure has more effect on the flame propagation than the condition of 0.6 MPa and 1.0 MPa, and plays a more important role with further increase of the pressure. Especially for the rich case, the peak pressure increasing rate falls from 0.008 (at P = 0.6 MPa) to 0.002 (at P = 1.5 MPa).

Numerical analysis of asymmetry of the turbulent jets from the six orifices

To understand the asymmetry phenomena of the turbulent jets from the six orifices as observed previously in Fig. 3, a CFD simulation of the methane fueled TJI system was conducted to generate more information to facilitate the understanding of the flow field, especially the flow in the pre-chamber which was not experimentally measurable, using the commercial computational fluid dynamic software CONVERGE [¹⁶].

Numerical specifications

The geometry for this simulation is the same as the internal shape of the constant volume chamber, and an STL file of the flow domain was prepared in Pro/ENGINEER. Based on the imported STL file, a triangulated mesh file was produced in CONVERGE, where the base grid size was chosen to be \mathrm{4}\mathrm{m}\mathrm{m}\times \mathrm{4}\mathrm{m}\mathrm{m}\times \mathrm{4}\mathrm{m}\mathrm{m}. The fixed embedding grid refinement was applied in the pre-chamber and the spark plug regions, and the scale levels are 3 and 5, respectively. To enhance the grid resolution in the regions with large temperature and velocity gradients, the adaptive mesh refinement (AMR) tool was implemented. The scale level was set to 5, which could lead to the minimum grid size of 0.125 mm. To discretize the equations, a second order accurate spatial discretization scheme was implemented. A modified Pressure Implicit with Splitting of Operator (PISO) was used to solve the transport equations. The first order fully implicit scheme was set for the temporal discretization. Variable time stepping was implemented and time steps were calculated on account of the maximum allowed Courant-Friedrichs-Lewy (CFL) numbers and combustion time step control.

It is noteworthy that, in order to capture the main vortices of the turbulent jet flow, especially the combustion initiation in the main chamber at early stage, the large eddy simulation (LES) turbulent model was implemented for the early stage of TJI (from the triggering spark to 0.06 s). However, due to the high computational cost of LES, the rest of combustion period other than the early stage was simulated by the RANS (Reynolds Average Navier-Stokes) turbulent model.

The premixed combustion modeling in this study was conducted by the SAGE detailed chemistry solver [⁶]. A multi-zone model [¹⁷] was used to accelerate the chemistry calculations. To track the CH* radical, which can be visualized as the “blue” color to indicate the flame front, the detailed chemical kinetic mechanism of methane developed and validated by the Gas Research Institute (GRI) was used in the combustion simulations, which includes 53 species and 325 reactions. Although the mechanism is probably only valid at low pressures, it is sufficient enough to explain the flow field at this stage.

Flow region recognition and definition

For a clear discussion in Section 4.3, the orifices of the pre-chamber are numbered in sequence. Based on the temperature field of the numerical results, the tip penetration of the hot turbulent jet actuated from each orifice of the pre-chamber is presented in Fig. 6. The threshold for extracting the hot turbulent jet contour was set at above 800 K, and the jet area through each orifice was defined as the sum of the pixels with the temperature field of above 800 K.

Temperature field prediction of turbulent jets under different initial conditions

The cross section view of the temperature fields of turbulent jets actuated from orifices ① and ②under different initial conditions are compared as shown in Fig. 7. From the evolution of temperature fields after ignition, it can easily be seen that the reacting jet propagates faster at lower initial pressures (0.6 MPa or 1.0 MPa) or lower air-fuel equivalence ratios (l = 0.8 or 1.0) during the initial stage of combustion. Furthermore, the numerical temperature flow contours illustrate that the flow development in the pre-chamber is not symmetric as the spark plug is located at the side of the cavity. Flame generation and propagation in the pre-chamber results in a pressure increase and a pressure imbalance between the pre-chamber and the main chamber. The pressure difference forces the unburned gas mixtures in the pre-chamber to enter the main chamber and generate a relatively cold and unburned jet which has a large impact on the turbulent mixing in the main chamber, leading to the asymmetry of the turbulent jets actuated from different orifices, as illustrated for orifices ① and ② in Fig. 7. This is mainly due to the asymmetrical formation of the initial spark kernel in the pre-chamber. It proves the asymmetry phenomena of the turbulent jet observed from the experimental images in Fig. 3. In addition, at a higher air-fuel equivalence ratio (l = 1.2), the maximum temperatures of the hot turbulent jet are lower. For instance, at l = 1.0, the maximum temperature of the jet is higher than 2300 K in the early stage, but at l = 1.2 the maximum temperature is lower than 2000 K in the same period.

Figure 8 shows temporal variations of hot turbulent jet areas from six orifices at different initial pressures (T_initial = 300 K and l = 1.0), where the jet area (A_jet) is defined as the number of pixels of the temperature region above 800 K, and the image area (A_T) is defined as the total number of pixels of the whole image. Again, due to the spark plug orientation [⁶], the asymmetry of the turbulent jet can be quantitatively reflected from the areas of the flame jets from different orifices. For example, the hot jet area from orifice ③ (A_jet/A_T≈ 0.030) can be more than six times larger than that from orifice ② (A_jet/A_T≈ 0.005) at 0.018 s (P_initial = 0.6 MPa and T_initial = 300 K). The area of the jet from orifice ② is the least, because most of the cold/unburned mixture in the pre-chamber is actuated through this orifice. At P_inj = 0.6 MPa and T_inj = 300 K, the start of the hot turbulent jets discharging for the orifices is around 0.013 s, and at a higher initial pressure (P_inj = 1.0 MPa) the starts of the hot turbulent jets extend to 0.016 s. Obviously, higher pressure (P_inj = 1.0 MPa) prohibits the flow propagation and penetration.

Figure 9 shows temporal variations of hot turbulent jet areas from six orifices at different initial global air-fuel equivalence ratios (T_inj = 300 K and P_initial = 0.6 MPa). At l = 1.0, the start of the hot turbulent jets discharging from the orifices is 0.013 s. However, at a higher air-fuel equivalence ratio (l = 1.2, lean condition), the starts of the hot turbulent jets are postponed to 0.03–0.04 s, which are three times later than the stoichiometric mixture of methane and air.

CH* radical distribution

It is known that the short lived radical species plays an essential role in combustion initiation with TJI [¹⁸]. As it is formed spatially close to the first sharp rise in temperature within the reaction zone, the CH* radical is an indicator of the flame front position [¹⁹]. Figure 10 compares the experimental and numerical results of TJI combustion characteristics in the constant volume chamber at T_inj = 300 K, P_initial = 0.6 MPa, and l = 1.0. The CH* radical can be visualized as the “blue” color to indicate the flame front. The “blue jet” in the middle circled by the red dashed line is caused by the overlapping effect of the inner boundaries of the six turbulent flame jets as shown in the numerical simulation. It is also observed that the relative concentration of CH* decreases with time, which proves that this radical species is short lived [⁸]. Furthermore, the flame propagation from the experimental images is slower than that from the numerical simulation, which is probably due to the heat losses of the flame impinging onto the combustion chamber wall, where the numerical simulation assumes that the wall is adiabatic, and which is probably also caused by the fact that the GRI mechanism is only valid at low pressures.

Conclusions

To investigate the effect of air-fuel equivalence ratio and pre-mixed pressure on pre-chamber TJI of methane/air mixtures with multiple orifices in a large bore constant volume chamber in the early stage, a set of experimental analysis and computational studies were conducted to analyze the discharge of hot turbulent jets from six orifices. The results were compared with different initial conditions (P_inj = 0.6, 1.0, and 1.5 MPa and l = 0.8, 1.0, and 1.2). The following conclusions can be made:

The flame propagation rate slows with the increase in the air-fuel equivalence ratio in the pre-chamber. At a higher air-fuel equivalence ratio (lean condition), the starts of the hot turbulent jets are postponed to 0.03–0.04 s, which are three times later than the stoichiometric mixture of methane and air.

The reacting turbulent jet propagates slower at a higher initial pressure, the ambient pressure plays a more important role with further increase in the pressure. Especially for the rich case, the peak pressure increasing rate falls from 0.008 (at P = 0.6 MPa) to 0.002 (at P = 1.5 MPa).

The CFD simulation illustrates that the spark kernel formation in the pre-chamber is asymmetric as the spark plug is located at the side of the cavity, and results in the asymmetrical distribution of the turbulent jets actuated from six different orifices.

To sum up, the injection strategy is essential to provide a rich fuel/air mixture in the pre-chamber to initiate combustion quickly but a lean mixture in the main chamber to maintain low NO_x emission levels. Therefore, the channel of the auxiliary pre-chamber fuelling is definitely required in the large-bore engine operation. However, it requires space to install the auxiliary pre-chamber fuelling pipeline inside the pre-chamber, which results in the spark plug to be located at the side of the cavity which probably causes the asymmetrical distribution of the turbulent jets observed in this work. Further optimization of the pre-chamber system is still required.

References

[1] Tsekenis S A, Wilson D, Lengden M, Hyvönen J, Leinonen J, Shah A, Andersson Ö, McCann H. Towards in-cylinder chemical species tomography on large-bore IC engines with pre-chamber. Flow Measurement and Instrumentation, 2017, 53: 116–125

[2] Shah A, Tunestal P, Johansson B. Effect of relative mixture strength on performance of divided chamber ‘avalanche activated combustion’ ignition technique in a heavy duty natural gas engine. In: SAE 2014 World Congress and Exhibition. Detroit, MI, USA, 2014, 1: 104424

[3] Shah A, Tunestal P, Johansson B. Effect of pre-chamber volume and nozzle diameter on pre-chamber ignition in heavy duty natural gas engines. In: SAE 2015 World Congress and Exhibition, Cobo Center Detroit, USA, 2015: 111931

[4] Toulson E, Schock H J, Attard W P. A Review of pre-chamber initiated jet ignition combustion systems. In: SAE 2010 Powertrains Fuels & Lubricants Meeting, San Diego California, USA, 2010: 2010–01–2263

[5] Attard W P, Fraser N, Parsons P, Toulson E. A turbulent jet ignition pre-chamber combustion system for large fuel economy improvements in a modern vehicle powertrain. SAE International Journal of Engines, 2010, 3(2): 20–37

[6] Gholamisheeri M, Wichman I S, Toulson E. A study of the turbulent jet flow field in a methane fueled turbulent jet ignition (TJI) system. Combustion and Flame, 2017, 183: 194–206

[7] Ricardo H R. Recent research work on the internal-combustion engine. SAE Technical Paper, 1922: 220001

[8] Gentz G, Thelen B, Gholamisheeri M, Litke P, Brown A, Hoke J, Toulson E. A study of the influence of orifice diameter on a turbulent jet ignition system through combustion visualization and performance characterization in a rapid compression machine. Applied Thermal Engineering, 2015, 81: 399–411

[9] Salahi M M, Esfahanian V, Gharehghani A, Mirsalim M. Investigating the reactivity controlled compression ignition (RCCI) combustion strategy in a natural gas/diesel fueled engine with a pre-chamber. Energy Conversion and Management, 2017, 132: 40–53

[10] Yousefi A, Birouk M. Investigation of natural gas energy fraction and injection timing on the performance and emissions of a dual-fuel engine with pre-combustion chamber under low engine load. Applied Energy, 2017, 189: 492–505

[11] Li S S, Bai S Z, Xing X W, Jia Y J, Li G X. Influence of pre-chamber parameters on combustion process in large natural gas engine. Chinese Internal Combustion Engine Engineering, 2012, 33(6): 72–76 (in Chinese)

[12] Bai S, Jia Y, Zhang H, Numerical research on the combustion system in natural gas engine with pre-combustion chamber. In: Annual Academic Meeting of Chinese Society for Internal Combustion Engines, Baoding, China, 2013

[13] Alvarez C E C, Couto G E, Roso V R, Thiriet A B, Valle R M. A review of prechamber ignition systems as lean combustion technology for SI engines. Applied Thermal Engineering, 2018, 128: 107–120

[14] Biswas S, Tanvir S, Wang H, Qiao L. On ignition mechanisms of premixed CH₄/air and H₂/air using a hot turbulent jet generated by pre-chamber combustion. Applied Thermal Engineering, 2016, 106: 925–937

[15] Peters N. Turbulent Combustion. Cambridge: Cambridge University Press, 2000

[16] Jiang W, Zhang Y, Li S, Numerical simulation of oxygen-rich CNG engine based on 1D/3D coupled model. Vehicle Engine, 2015, 6: 39–43

[17] Kodavasal J, Keum S, Babajimopoulos A. An extended multi-zone combustion model for PCI simulation. Combustion Theory and Modelling, 2011, 15(6): 893–910

[18] Murase E, Hanada K. Enhancement of combustion by injection of radicals. In: SAE 2000 World Congress, Detroit, MI, USA, 2000: 2000–01–0194

[19] Karnani S, Dunn-Rankin D. Visualizing CH* chemiluminescence in sooting flames. Combustion and Flame, 2013, 160(10): 2275–2278

RIGHTS & PERMISSIONS

Higher Education Press and Springer-Verlag GmbH Germany, part of Springer Nature