Experimental Study of the Effects of Pre-Chamber Geometry on the Combustion Characteristics of an Ammonia/Air Pre-Mixture Ignited by a Jet Flame

Abstract

In the future, ammonia is expected to become a carbon-free fuel for internal combustion engines. However, the flammability of ammonia is poorer compared to conventional fuels such as gasoline and diesel fuel. Pre-chamber jet ignition may be an effective way to ensure stable ignition and enhance the combustion of ammonia. In this paper, the effects of pre-chamber geometric parameters, including volume and orifice diameter, on the jet ignition and combustion processes were studied using visualization methods, combined with pressure acquisition. The results showed that ignition energy increased and the jet duration was prolonged with the increase in pre-chamber volume, resulting in a higher maximum pressure and pressure rise rate in the main chamber. The jet characteristics of a larger volume pre-chamber exhibited higher stability when the ambient parameters were changed. The smaller volume pre-chamber showed the superiority of a shorter flame propagation distance inside the pre-chamber, which advanced the timing of the jet appearance and shortened the ignition delay when the flammability of the pre-mixture was adequate. The larger pre-chamber orifice diameter caused an earlier jet ignition timing, shorter ignition delay, and higher ignition location. The jet duration for the pre-chamber with a smaller orifice was longer, which was beneficial for increasing the pressure rise rate in the main chamber. Too small a pre-chamber orifice led to ignition failure in the main chamber.

1. Introduction

The internal combustion engine is one of the most widely used power machines. However, the long-term usage of fossil fuels, including gasoline and diesel, has caused energy shortages and environmental problems [1].

CO₂ is one of the elements of engine emissions, and it is well known to be the main component of greenhouse gas. Scholars have been trying to reduce CO₂ emissions by optimizing the combustion processes of internal combustion engines or seeking alternative fuels. Hydrogen (H₂) and ammonia (NH₃) may be the ideal carbon-free fuels because they contain no carbon element.

For hydrogen, although it has many advantages, including a low ignition energy requirement, a wide flammability limit, high flame propagation velocity, and low emissions [2], the explosive property and difficulties in storage and transportation hinder hydrogen from becoming a widely used fuel for internal combustion engines.

Ammonia is one of the most abundant inorganic chemical products in the world, and it has been widely used in agriculture and industry [3]. Like hydrogen, ammonia is a carbon-free fuel. Moreover, ammonia can be easily liquefied at a pressure of about 8 bar at 294 K, which brings great convenience to the storage and transportation of ammonia [4]. Based on the advantages mentioned above, ammonia is expected to be a favorable alternative fuel for internal combustion engines, solving many energy and environmental problems. Compared with conventional fuels, the low flame propagation velocity, high auto-ignition temperature, and narrow range of the combustible equivalence ratio of ammonia are problems that remain to be solved [5,6,7]. The research of Grey et al. indicated that in a compression ignition engine, the compression ratio should be as high as 35:1 to ensure the ignition of direct injection ammonia [8]. Garabedian’s study [9] and Starkman’s research [10] have also proved that single-fuel ammonia is difficult to use in compression ignition engines. Using diesel, kerosene, or DME as a pilot fuel may be an effective way to promote the ignition of ammonia [11,12,13,14]. Ammonia may be more suitable for spark ignition (SI) engines because of its low auto-ignition ability and high octane rating of 130 [15]. Charles Lhuillier et al. studied the engine performance, combustion characteristics, and emissions of NH₃/H₂/air mixed fuel in a spark ignition engine [16,17]. The results showed that H₂ obviously accelerated the early stage combustion of NH₃; adding a 20% volume fraction of H₂ to the fuel can improve the cycle stability and avoid misfire. Christine Mounaïm-Rousselle et al. explored the load limit of a spark ignition engine fueled with pure ammonia, or with only a small amount of hydrogen added [18]. The results showed that the low load operation was difficult for pure ammonia, but increasing the intake pressure or adding 10% hydrogen may be effective ways of resolving this concern. Ryu K et al. studied the combustion and emission characteristics of an SI engine with the direct injection of gaseous ammonia along with the assistance of port injection gasoline [15]. The results indicated that with the enhancement of gasoline, the power performance of the ammonia engine was comparable to that of the gasoline engine. However, the NO_x emission increased and an ammonia slip was detected.

Pre-chamber jet ignition may be another way to ensure the stable ignition of an ammonia/air mixture and promote its combustion processes. A pre-chamber jet ignition system has the advantages of high ignition energy, widely distributed ignition points, and strong turbulence fluctuation [19,20,21]. Additionally, it can save one fuel supply system compared to dual-fuel pilot ignition.

This ignition method has a wide range of fuel flexibility, and many scholars have studied the combustion characteristics of various fuels (including methane, hydrogen, and methanol) ignited by a pre-chamber jet [22,23,24,25]. Their results showed that pre-chamber jet ignition can improve ignition stability, significantly increase combustion rate, and extend the lean burn limit.

The visualization method can provide a deep diagnosis of the ignition and combustion processes. However, existing visualization studies on ammonia are relatively fundamental, mainly focusing on the laminar flame velocity of ammonia or the blend of ammonia and other fuels [26,27,28,29]. There are few studies regarding the ignition and combustion characteristics of ammonia ignited by a pre-chamber jet flame.

According to some researchers’ previous studies [30,31,32], the geometric parameters of the pre-chamber have vital effects on the jet ignition and subsequent combustion in the main chamber of methane/air pre-mixture. In this paper, the effects of the pre-chamber geometric parameters on the jet ignition and combustion characteristics of ammonia/air pre-mixture were studied by shadowgraph, natural luminescence shooting, and pressure acquisition. This work may provide substantial and valuable fundamental data for the development of pre-chamber type ammonia engines.

2. Experimental Setup

2.1. Test Bench

This study was carried out on a constant volume combustion test bench. The experimental setup is shown in Figure 1. The test bench consisted of a constant volume chamber (CVC) equipped with a pre-chamber, synchronous control and data acquisition system, gas supply system, image capture system, and temperature-control system. A Photron SA-Z high-speed gray camera was used for the shadowgraph, and the frame rate was 20,000 fps. A Photron mini UX100 color camera was used for natural luminescence, and the frame rate was 6400 fps. The light path of the shadowgraph is shown in Figure 1. As shown in Figure 2, the rising edge of the TTL signal used to trigger the shooting is synchronized with the charging moment of the ignition system and the start of pressure acquisition.

Figure 3 shows the structure of the CVC. The ammonia/air mixture was heated by an external heating module with six heating bars. The internal temperature could be up to 600 K. The quartz windows provide an optical diameter of 80 mm, and they can bear a burst pressure of 10 MPa. The constant volume chamber was pressurized with high-pressure air, and the equivalence ratio of the ammonia/air mixture was determined by the partial pressure of ammonia. A Kistler 6052C pressure sensor and a K-type thermocouple were installed to monitor and acquire pressure and temperature. The pressure acquisition frequency was 12,500 Hz.

Figure 4 shows the structure of the pre-chamber. The upper part of the pre-chamber was equipped with a spark plug, and the lower part connected the main chamber with a single orifice. In this study, the mixture inside the pre-chamber was not individually enriched; thus, the equivalence ratios of the pre-chamber and main chamber were the same.

2.2. Experimental Parameters

Table 1. Experimental parameters.

Parameters	Value
Volume V/mL	4, 6, 8
Orifice diameter d/mm	1, 2, 3
Initial pressure p0 /MPa	0.8, 1.1
Initial temperature T0/K	500, 550
Equivalence ratio Φ	0.7, 0.9, 1.0
Ignition system charging pulse width /ms	20

shows the experimental parameters. The effects of the geometric parameters of the pre-chamber under different ambient parameters were studied and compared. For different volumes (V), the ratio of length (L) and diameter (D) of the pre-chamber were kept constant.

3. Results and Discussion

3.1. Jet Ignition Processes

Figure 5 shows the three stages of the jet ignition process of the ammonia/air mixture. The main feature of the first stage was that the unburnt gas at the bottom of the pre-chamber was squeezed out via the orifice. The shadow images showed that the timingswhen the gas emerged were about 20.60 ms after the start of shooting in all experiments. This phenomenon might be related to spark plug discharge activities. Differing from high reactivity fuels, such as methane or hydrogen, high ignition energy is required for ammonia. According to the authors’ previous studies, on the same test bench, for methane, a charging time of 5 ms was sufficient to ensure stable ignition in the pre-chamber [25], and 6 ms was enough for methanol [32]. But for ammonia, the charging period should be as long as 20 ms to ensure stable ignition in the pre-chamber, and a spark plug with four electrodes should be used, rather than the single-electrode spark plug used in previous studies. Therefore, when the spark plug discharged, gas around the electrodes expanded rapidly due to the high ignition energy, and a portion of the unburnt gas was squeezed out. Then, the mixture in the pre-chamber was ignited, and the unburnt gas continued to be pushed out because of the combustion inside the pre-chamber. When the flame passed through the orifice and formed a jet flame, the second stage began. As the jet flame continued to develop, the mixture in the main chamber was ignited and burned, and the third stage began.

3.2. Effects of Pre-Chamber Volume

Figure 6 shows the shadow images of jet ignition processes with different pre-chamber volumes, and other experimental parameters were T₀ = 500 K, p₀ = 1.1 MPa, and d = 2 mm. The time shown in Figure 6 starts from the jet appearance, and the subsequent time is marked after the jet appeared (AJA).

As shown in Figure 6, the flame propagation in the main chamber with the 8 mL pre-chamber was the fastest at all equivalence ratios. Under the condition of Φ = 0.7, the ignition in the main chamber with the 8 mL pre-chamber was much earlier than that using the 4 mL and 6 mL pre-chambers. With an increase in the equivalence ratio, the pre-mixture became more combustible, the ignition advanced, the flame propagation was accelerated, and the combustion became more intense. When the flame nearly filled the window, a bright light was emitted. At the conditions of Φ = 0.9 and Φ = 1.0, the ignition in the main chamber with the 4 mL pre-chamber was still the latest, but the ignition timings using the 6 mL and 8 mL pre-chambers were close to each other when Φ = 1.0.

Figure 7 shows the accurate jet appearance timings, the main chamber ignition timings after spark ignition (ASI), and the ignition delays in the main chamber; these data were obtained from the shadow images. The ignition delay is defined as the interval between the jet appearance and ignition in the main chamber. The jet appearance timings were determined by the time needed for flame propagation inside the pre-chamber, and the time was affected by both the flame propagation distance and the flame velocity. The ignition timings and the ignition delays were determined by the ignition energy carried by the jet flame.

Figure 8 shows the schematic diagram of flame generation and propagation inside the pre-chamber. After the spark plug was discharged, an initial flame kernel was formed near the electrodes, and the flame propagated downwards. When the flame touched the wall of the pre-chamber, heat would transfer rapidly to the pre-chamber. The flame propagation during the period before the flame touched the wall can be regarded as an adiabatic process. For a larger volume pre-chamber, its diameter (D) was larger, so the period of the adiabatic flame propagation process was longer and the heat loss was smaller. Moreover, the pre-chamber can be regarded as a relatively enclosed space, and the heat dissipation rate per unit volume q_L can be described by Equation (3). The convective heat transfer coefficient h is a constant value. F is the internal surface area of the pre-chamber, V is the pre-chamber volume, T_b is the temperature of the burnt gas, and T_w is the temperature of the pre-chamber wall. The interior of the pre-chamber was an approximate cylinder, so Equation (1) can be transformed to Equation (2). The heat release rate by unit volume of the pre-mixture can be described by Equation (3). Subtract Equation (3) from Equation (2) to get the net heat release rate per unit volume of the pre-mixture (Equation (4)). H is the reaction heat per unit volume of the pre-mixture, which is a constant value, and k₀ is the frequency factor, which is related to the number of molecular collisions. [A] is the concentration of the pre-mixture, and the n is the total reaction order. E is the activation energy, and R_u is the molar gas constant. L and D of the pre-chamber were proportionally changed when the pre-chamber volume changed; thus, a larger volume pre-chamber exhibited a smaller q_L. Before the reaction started, the q_G of each pre-chamber was the same because of the same initial pre-mixture parameters.

$$q_L=\frac{hF}{V}\left(T_b-T_w\right)$$

(1)

$$q_L=2h(\frac{1}{L}+\frac{2}{D})\left(T_b-T_w\right)$$

(2)

$$q_G=H{k}_0{\left[A\right]}^n{e}^{\frac{-E}{R_u{T}_b}}$$

(3)

$$q=H{k}_0{\left[A\right]}^n{e}^{\frac{-E}{R_u{T}_b}}-2h(\frac{1}{L}+\frac{2}{D})\left(T_b-T_w\right)$$

(4)

At Φ = 0.7, it can be seen from Figure 7 that the jet appearance timing, ignition timing, and ignition delay decreased with the increase in pre-chamber volume. Under this condition, the flame propagation velocity of ammonia was low due to the low equivalence ratio. Thus, the time needed for flame propagation was extended. In other words, the heat exchange time was long, and thus, the heat loss would play an important role. For a larger volume pre-chamber, due to it having a lower q_L, the heat loss was less. Additionally, as mentioned in Section 3.1, when the spark plug discharged, some of the pre-mixed gas emerged from the pre-chamber through the orifice before being ignited, causing the concentration of the pre-mixture ([A]) in the pre-chamber to be reduced. As for a larger volume pre-chamber, the ratio of the amount of emerged gas to the original pre-mixture amount in the pre-chamber would be less. Therefore, when the ignition started in the pre-chamber, the concentration ([A]) of the ammonia/air pre-mixture of the larger volume pre-chamber would be higher, which means that the heat release rate (q_G) of the larger volume pre-chamber would also be higher. Thus, the burnt area temperature (T_b) and the flame velocity of the larger volume pre-chamber would be higher, which made the jet appearance timing for the 8 mL pre-chamber the earliest, followed by the 6 mL and then the 4 mL pre-chambers. The turbulence kinetic energy and temperature of the jet flame from a larger volume pre-chamber would also be larger, so it could ignite the pre-mixture in the main chamber earlier.

Under the condition of Φ = 0.9, the jet appearance of the 6 mL pre-chamber was the earliest. When the flammability of the pre-mixture increased with the increase in the equivalence ratio, ignition inside the pre-chamber advanced, and the flame propagation inside the pre-chamber was accelerated, causing a reduction in the heat exchange time and the amount of emerged gas. The negative effects caused by heat loss and emerged gas were diminished to some extent, and the advantages of shorter flame propagation distance caused the jet appearance timing of the 6 mL pre-chamber to occur earlier than for the 8 mL pre-chamber. As for the 4 mL pre-chamber, even though it had the shortest flame propagation distance, under this condition, the heat loss and emerged gas still had a great influence on the flame propagation velocity inside the pre-chamber and the ignition energy of the jet flame. Thus, its jet appearance timing was still the latest among the three pre-chambers. The ignition delay of the 8 mL pre-chamber was still the shortest, followed by that of the 6 mL and then the 4 mL pre-chambers.

When the equivalence ratio rose to 1.0, the flammability was further enhanced. Therefore, the heat loss and the amount of emerged gas were reduced to small values, exerting only a tiny influence on the flame velocity inside the pre-chambers. The flame propagation distance inside the pre-chamber became the most important factor affecting the appearance of the jet, so the jet appearance timing retarded with the increase in pre-chamber volume. The orders of ignition timings and ignition delays from small to large changed to the 6 mL pre-chamber, the 8 mL pre-chamber, and the 4 mL pre-chamber, respectively. Compared to the 8 mL pre-chamber, the flame propagation of the 6 mL pre-chamber had a shorter vertical length, so that when the flame propagated from the top to the orifice, the propagation time required for the 6 mL pre-chamber was shorter than for the 8 mL pre-chamber. The cumulative heat loss in the 6 mL pre-chamber would be less, and the temperature and kinetic energy of the jet flame from the 6 mL pre-chamber would be larger than those for the 8 mL pre-chamber, so the ignition delay for the 6 mL pre-chamber was shorter than that for the 8 mL pre-chamber. For the 4 mL pre-chamber, at these ambient parameters, the cumulative heat loss of the 4 mL pre-chamber would still be the largest among the three pre-chambers, due to its having the largest q_L, and the emerged gas ratio of the 4 mL pre-chamber would still be the largest. The advantages of having the shortest flame propagation distance could not counteract the negative effects of it having the largest heat loss and emerged gas ratio. Even though its jet appearance timing was the earliest, its ignition delay was still the longest.

It can also be noticed that when the equivalence ratio increased from 0.7 to 0.9, and then to 1.0, the values of jet appearance timing, ignition timing, and ignition delay of the 8 mL pre-chamber decreased more slightly than those of the 4 mL and the 6 mL pre-chambers. As discussed above, the 8 mL pre-chamber had the smallest emerged gas ratio, q_L, and the longest adiabatic flame propagation process. It can be speculated that the jet characteristics of the 8 mL pre-chamber were inherently little affected by these negative factors. Thus, the change in the equivalence ratio would not have a significant influence on the 8 mL pre-chamber. However, for the 4 mL and the 6 mL pre-chambers, the increase in the flammability of the pre-mixture would significantly reduce the negative impacts of both the heat loss and emerged gas.

Figure 9 shows the main chamber pressure (MCP) traces and pressure rise rates (PRR) in the main chambers with different pre-chamber volumes. At Φ = 0.7, the maximum MCPs of the three pre-chambers were close to each other. However, the maximum PRR of the 8 mL pre-chamber was much higher than those of the 4 mL and the 6 mL pre-chambers. Moreover, the combustion duration (PRR > 0) decreased as the pre-chamber volume increased. The combustion duration of the 4 mL pre-chamber was more than twice that of the 8 mL pre-chamber. At Φ = 0.9 and Φ = 1.0, the maximum MCPs increased as the pre-chamber volume increased. The PRRs of the 8 mL pre-chamber were still the largest, and its combustion durations were the shortest, especially at Φ = 1.0, although the ignitions in the main chamber with the 4 mL and the 6 mL pre-chambers occurred earlier than with the 8 mL pre-chamber; due to it having the shortest combustion duration, the MCP of the 8 mL pre-chamber reached the maximum at the earliest time.

Figure 10 shows the shadow images of the jet ignition processes of the pre-chambers with three volumes under different initial thermodynamic parameters; the equivalence ratio was kept the same as 1.0. Combined with the data in Figure 11, it was found that under the condition of T₀ = 500 K and p₀ = 0.8 MPa, the jet appearance and ignition orders were the 8 mL pre-chamber, the 6 mL pre-chamber, and the 4 mL pre-chamber, respectively. When the initial pressure was increased to 1.1 MPa, the experimental parameters and results were the same as under the condition of Φ = 1.0, as shown in Figure 7c.

Increasing the temperature to 550 K, the flammability of the pre-mixture was further enhanced. It can be noticed in Figure 11c that the jet appearance timing and the ignition timing of the 4 mL pre-chamber were the earliest, and its ignition delay was the shortest. This may be because when the flammability of the ammonia/air pre-mixture was enough to ensure rapid ignition and fast flame propagation in the pre-chamber, the amount of emerged gas would be very tiny and the heat exchange timing would be short. Thus, the flame velocity and flame temperature in the pre-chamber would not be significantly affected by these two factors. Moreover, as the initial temperature was increased to 550 K, the temperature of the pre-chamber wall (T_w) was also increased, to some extent. Therefore, the decrease in (T_b − T_w) caused the heat dissipation rate to decrease, and the differences in q_L among the three pre-chambers were narrowed. The distance required for flame propagation inside the 4 mL pre-chamber was the shortest, enabling the flame to pass through the pre-chamber in the shortest period of time; therefore, the cumulative heat loss of the 4 mL pre-chamber would be the least. Thus, the 4 mL pre-chamber could generate the jet faster than the 6 mL pre-chamber and the 8 mL pre-chamber, igniting the pre-mixture in the main chamber earlier after the ejection of the jet. The differences in the ignition delay among the three pre-chambers were not large.

The MCP and PRR traces are displayed in Figure 12. It can be seen that under each of the three conditions, the maximum MCP and PRRs of the 8 mL pre-chamber were the largest, and its combustion durations were the shortest. Although under the condition of T₀ = 550 K and p₀ = 1.1 MPa, the 8 mL pre-chamber exhibited disadvantages with regards to the jet appearance and ignition delay, the larger volume caused the chemical energy stored in the 8 mL pre-chamber to be greater than that in the 4 mL pre-chamber and the 6 mL pre-chamber. As a result, the jet from the 8 mL pre-chamber could carry the largest ignition energy. Moreover, because the three pre-chambers have the same orifice area, it took a longer time for the flame in the 8 mL pre-chamber to be completely ejected than for the 4 mL pre-chamber and the 6 mL pre-chamber, which prolonged the duration of the ignition and the turbulence disturbance in the main chamber.

The effects of pre-chamber volume on the process variables, the jet ignition characteristics, and the combustion characteristics are summarized and shown in Figure 13.

3.3. Effects of Pre-Chamber Orifice Diameter

Figure 14 shows the jet appearance timings, ignition timings, and ignition delays for different pre-chamber orifice diameters under different initial pressures and temperatures. The other experimental parameters were V = 6 mL and Φ = 1.0. The pre-chamber of 1 mm orifice diameter could not ignite the pre-mixture in the main chamber under any of these conditions.

It was found that under the condition of T₀ = 500 K & p₀ = 0.8 MPa, the jet appearance timing of the 3 mm pre-chamber was earlier, and at the other two conditions, the jet appearance timings of the other two pre-chambers were close. Under all the three conditions, the values of the ignition timings and ignition delays decreased as the orifice diameter changed from 2 mm to 3 mm. The 3 mm orifice pre-chamber had a larger outlet area than that of the 2 mm orifice, so the throttling effect of the 3 mm orifice pre-chamber was less than that of the 2 mm orifice pre-chamber. When the flame passed through the 3 mm orifice, the heat loss and kinetic energy dissipation would be less. As a result, the jet generated from the 3 mm orifice pre-chamber could cause the earlier ignition of the pre-mixture in the main chamber.

Figure 15 displays the MCP and PRR traces. As can be seen from these graphs, under the three conditions, the MCPs and PRRs of the 3 mm orifice pre-chamber developed first, which corresponds with the data shown in Figure 14. Before the PRRs of the 3 mm orifice pre-chamber reached the maximums, the slopes of the PRR curves for the 3 mm orifice pre-chamber were greater than those of the 2 mm orifice pre-chamber, indicating the earlier combustion in the main chamber with the 3 mm orifice pre-chamber. However, the maximums of both the MCPs and PRRs of the 2 mm orifice pre-chamber were higher than that of the 3 mm orifice pre-chamber. As for the combustion durations, when T = 500 K and p = 0.8 MPa, the combustion durations of the two pre-chamber were close, and the ignition timing of the 3 mm orifice pre-chamber was earlier, so the MCP of the 3 mm orifice pre-chamber reached the maximum earlier too. Under the conditions of 500 K and 1.1 MPa and 550 K and 1.1 MPa, the combustion durations of the 2 mm orifice pre-chamber were shorter, and the MCPs of the 2 mm orifice pre-chamber reached their maximums earlier.

Taking the condition of T₀ = 500 K and p₀ = 1.1 MPa as an example, the natural luminescence images of the process from the time of the appearance the jet flame to the ignition in the main chamber are shown in Figure 16, and the jet penetration lengths are measured. The time when the jet flame obviously propagated laterally was regarded as the time when the main chamber ignition began. The luminous intensity of the jet from the 3 mm orifice pre-chamber was higher, and before ignition in the main chamber with the 3 mm orifice pre-chamber, the jet penetration length of the 3 mm orifice pre-chamber was also longer. We noticed a phenomenon that, for the 3 mm orifice pre-chamber, after about 5 ms AJA, the portion of the jet near the orifice became wide, and the jet flame oscillated. This may be because as the combustion in the pre-chamber progressed, the pressure difference between both sides of the orifice increased, the speed of jet flame was accelerated, and the increased air resistance caused the vertical velocity of the flame edge to decrease. This phenomenon for the 2 mm orifice pre-chamber occurred later, and the oscillation degree of jet flame was relatively small. After about 9 ms AJA, the jet penetration length of the 2 mm orifice pre-chamber grew at a faster rate. The area of the 2 mm orifice was less than half that of the 3 mm orifice; therefore, the resistance of the 2 mm orifice was much larger when the flame passed through it. As a result, the pressure was easier to establish inside the 2 mm orifice pre-chamber after ignition. The ignition delay of the 2 mm orifice pre-chamber was larger; thus, the jet remained affected by the main chamber flame for a longer period time, and the pressure difference between both sides of the 2 mm orifice was further increased, which caused the jet to grow faster in the late stage. Before the ignition in the main chamber using the 2 mm orifice pre-chamber, the jet penetration length of the 2 mm orifice pre-chamber was larger than that of the 3 mm orifice pre-chamber.

The ignition in the main chamber always occurred at the jet flame front, so the ignition position in the main chamber with the 2 mm orifice pre-chamber was located close to the center of the main chamber, while that of the 3 mm orifice pre-chamber was close to the top. The centrally near ignition positions caused the flame propagation distance in the main chamber with the 2 mm orifice pre-chamber to be shorter, shortened the combustion duration, and increased the PRR of the 2 mm orifice pre-chamber. Additionally, the jet duration of the 2 mm orifice pre-chamber was longer, and the jet continued to provide turbulence fluctuation for some time after the pre-mixture in the main chamber was ignited, which also contributed to the higher PRR maximum. Under the conditions of 500 K and 1.1 MPa and 550 K and 1.1 MPa, the PRR curves of the 3 mm orifice pre-chamber showed the same development trend—the PRR first rose rapidly, fell to a plateau, and finally reached zero. It can be speculated that due to the turbulence fluctuation of the jet, the burning rate of the pre-mixture in the main chamber was high at the early stage. When the fluctuation effects dissipated, the PRR decreased, and the flame in the main chamber continued to propagate at a lower rate, which caused the plateau of the PRR curves. For the 2 mm orifice pre-chamber, due to its longer jet durations, the PRRs could continue to rise, and the curves were unimodal.

The effects of pre-chamber orifice diameter on the process variables, the jet ignition characteristics, and the combustion characteristics were summarized and are shown in Figure 17.

4. Conclusions

In this paper, the jet ignition and combustion processes of ammonia/air were studied by visualization methods, pressure data acquisition, and heat release analysis. The conclusions can be drawn as follows.

(1) When the flammability of the pre-mixture was poor, the emerged gas caused by spark plug discharge and the heat exchange between the flame and pre-chamber wall had a vital influence on the jet and ignition characteristics. A larger volume pre-chamber yielded an earlier jet appearance timing and a shorter ignition delay due to its smaller heat dissipation rate and emerged gas ratio.

(2) Compared to the smaller volume pre-chamber, changes in ambient parameters exerted less influence on the jet appearance timing and ignition delay of a larger volume pre-chamber.

(3) With the increase in the pre-mixture’s flammability, the advantage of shorter internal flame propagation distance produced earlier jet appearance timings and shorter ignition delays using the smaller volume pre-chamber.

(4) A larger volume pre-chamber could provide higher ignition energy and longer jet duration, which shortened the combustion duration and increased the pressure rise rates of the main chamber.

(5) The 1 mm orifice pre-chamber failed to ignite the pre-mixture in the main chamber.

(6) Compared to the 2 mm orifice pre-chamber, the 3 mm orifice pre-chamber exhibited earlier ignition timings and shorter ignition delays at various ambient parameters.

(7) The ignition position in the main chamber with the 2 mm orifice pre-chamber was lower than that of the 3 mm orifice pre-chamber because of the larger ignition delay and longer jet penetration length of the 2 mm orifice pre-chamber.

(8) The jet durations of the 2 mm orifice pre-chamber were longer due to its smaller outlet area, which caused the pressure rise rates to increase.