Analyzing Combustion Efficiency According to Spray Characteristics of Gas-Centered Swirl-Coaxial Injector

Abstract

The momentum flux ratio (MFR) significantly affects the mixing characteristics and combustion efficiency of propellants in rocket engine injectors. The spray characteristics of three gas-centered swirl-coaxial injectors used in a full-scale combustion test were investigated according to the change in the momentum flux ratio. The difference in combustion efficiency was analyzed through the comparison with combustion test results using spray visualization and quantification. The spray cross-sectional shape and droplet distribution were measured using a structured laser illumination planar imaging technique. As the swirl effect was more apparent at a low MFR, the flow rate of the liquid that was sprayed outside was high. The flow rate of the liquid sprayed around the gas injection increased with the MFR. The Sauter mean diameter (SMD) of each injector liquid spray was obtained using the laser shadow imaging method. The SMD decreased as the MFR of all injector types increased, and the injector with a high liquid flow rate and small SMD injected towards the gas center exhibited higher combustion efficiency than the injector with a dominant liquid spray and the large SMD at a large injection angle. The outcomes of the study could help contribute to the increase in the combustion efficiency of the full-scale staged combustion cycle engine combustor.

1. Introduction

The liquid engines used in launch vehicles require less fuel with a higher combustion efficiency and can carry a heavier payload into space. Developing high-efficiency engines is essential at a time when the competition for low-cost launch services is intensifying in countries around the world. Thus, there have been several efforts to improve the engine efficiency in recent years [1,2,3,4,5]. One such approach involves increasing the combustion pressure, and for this purpose, a turbo pump that is capable of generating a larger discharge pressure is required. The gas generator cycle, which is a type of rocket engine that incorporates a turbopump, is used to drive the turbine. It results in an ISP loss of approximately 1–1.5% when the combustion pressure is 70 bar mainly due to the gas that is discharged to the outside [6]. This loss further increases with the increase in combustion pressure. Thus, reusing the propellant that drives the turbine is crucial for improving the overall performance. The staged combustion cycle engine, which was primarily designed to maximize all propellants for combustion, is an example of a system that focuses on reusing the propellant. Since the introduction of the staged combustion cycle engine in the late 1950s, it has been extensively developed and commercialized in major countries such as the United States, China, and India. Systems with various specifications pertaining to different propellants, engine shapes, and thrusts have been developed. Among these, the oxidizer-rich staged combustion cycle engines, such as RD-170, which were developed in the former Soviet Union, incorporated liquid oxygen/kerosene as the propellant [7]. Several versions of the engine were developed, including the two-chamber RD180, the one-chamber RD-191, and the RD-151 used in the first stage of the KSLV-1. Owing to the high engine reliability, the RD-180, along with its series, are still used in various launch vehicles, including the Atlas V. The gas-centered swirl-coaxial (GCSC) injector is used in the main combustor of the oxidizer-rich staged combustion cycle engine. It sprays the high-temperature oxidizer gas that is burned in the pre-burner into the main combustion chamber in order to split the liquid propellent into droplets with a high gas velocity.

The characteristics of the GCSC injector are not widely known. Nonetheless, there has been a recent surge in the research on the GCSC injector with considerable attention given to various types of engine development processes. Strakey et al. studied the relationship between the injector atomization conditions and combustion efficiency based on a comparison of the c* efficiency, in addition to the cold flow and combustion tests for various injector shapes [8]. Lightfoot et al. used approaches such as high-speed imaging and laser measurement to understand the spray characteristics of the GCSC injector, and they studied the length and thickness of the liquid film according to the injector shape and momentum flux ratio ($\mathrm{MFR}={\rho }_g{U}_g^2/{\rho }_l{U}_l^2)$ [9,10,11,12]. A larger MFR yields an enhanced atomization and stability, and it further enables a decreased liquid film length. However, the impact on the injection angle is not significant. Wagner et al. conducted combustion tests for various types of GCSC single injectors while changing the mixture ratio and combustion pressure [13]. According to the experimental results, if there is a swirl in the gas injector, the combustion efficiency increases, but there is the disadvantage of increasing the wall heat load. Yang et al. conducted experiments on the droplet size according to the recess and reported that the droplet size was minimized in a specific recess between the outer mixing and inner mixing [14]. Kulkarni studied the breakup behavior according to changes in the liquid and gas flow rates [15]. Kim et al. conducted visualization experiments with the recess ratio (RR), which is the ratio of the diameter of the gas injector to the depth of the gas injector, as a variable in a high-pressure environment [16]. The experimental results confirmed that the injection angle decreased to a specific value as the MFR increased. Park et al. reported an increase in the MFR and RR in the MFR regions below eight that resulted in a decrease in the film thickness and spray angle [17,18]. Jung et al. conducted a droplet size and injection angle study on GCSC, wherein both the gas and liquid were swirls. They compared and analyzed the difference in spray characteristics between internal and external mixing [19]. Anand et al. found that as the MFR increased under the condition that the MFR was less than or equal to seven, the injection angle and droplet size decreased [20]. Lim et al. compared and analyzed the characteristics of the liquid-centered swirl-coaxial injector and GCSC injector. They reported that when the MFR increased, the injection angle decreased at a low MFR and increased again with the MFR [21].

In addition to the aforementioned studies, there has been significant research on the combustion characteristics and efficiency of GCSC injectors. Schlieben et al. analyzed the dynamic characterization of the GCSC single injector through a combustion test [22]. According to Xuan et al., for zero recess length, the combustion efficiency decreased as the momentum ratio increases [23]. Roa et al. analyzed the flame shape through flame visualization in the reacting flow and reported the generation of soot, which is a combustion product, around the outer periphery of the swirling liquid sheet [24]. Balance et al. analyzed the change in the flame position and the angle of combustion shape based on a single injector combustion test according to the combustion pressure and momentum ratio; the location of soot was also identified in the IR image [25].

Previous studies primarily focused on the spray characteristics of a single GCSC injector. In addition, combustion tests were conducted using a single injector, and the flame visualization and efficiency analysis were performed accordingly. In a full-scale combustor, numerous injectors are arranged in the combustor head. The spray characteristics according to the difference in the flow rate per injector and the injector shape also have a considerable impact on the results of the full-scale combustion test; they directly affect the combustion performance. Therefore, when developing a liquid rocket engine, it is important to evaluate the combustion characteristics and performance of a full-scale combustor through combustion tests. The RD-170 engine performed more than 300 combustion tests for the combustor development and over 900 combustion tests for 200 engines [6,26]. Woo et al. conducted combustion tests on a full-scale staged combustion cycle engine by applying GCSC injectors, and they investigated the characteristic combustion velocity and efficiency according to the MFR. They further stated that the higher the MFR, the higher the combustion efficiency [27]. For the full-scale test of a staged combustion cycle engine with GCSC injectors, an expensive and large experimental device or a high-performance turbo pump that is capable of pressurizing large amounts of propellants at high pressure is required. In addition, due to the nature of the oxidizer-rich gas, it may potentially explode during tests. Thus, studies on the relationship between the results of the single injector and those of the full-scale combustion tests are not well known. In this study, we used GCSC injectors that were primarily designed for the full-scale combustion test of a staged combustion cycle engine, and we compared and analyzed the difference in the spray characteristics according to the injector shape under the staged combustion cycle engine operating conditions. We applied two spray visualization methods in an atmospheric pressure environment. Furthermore, based on the experimental results, the causes and impacts of the differences in spray characteristics according to the shape and operating conditions of the GCSC injectors on the combustion efficiency were analyzed and compared with the full-scale combustion test results. This study aims to reduce the development cost and time duration in the engine development stage by deriving the design requirements for injectors with a high combustion efficiency.

2. Experimental Methods

Three main types of injectors were used for the full-scale combustor and tested for visualization and combustion. The purpose of the difference in the GCSC-A and GCSC-B injector shapes was to study the effect of the MFR through the difference in the gas velocity at the same flow rate. The difference in the shape between GCSC-A and GCSC-C is for analyzing the effect of the MFR through the difference in flow rate per injector at the same gas velocity.

Table 1. Injector configuration.

Injector Type	Total Injector Quantity	Dg [mm]	Dl [mm]	L [mm]	Ug [m/s]	Recess Ratio
GCSC-A	60	7.4	10.4	10.4	100	1.0
GCSC-B	60	8.4	11.4	11.4	75	1.0
GCSC-C	36	9.6	13.5	13.5	100	1.0

lists the total number of injectors used in the combustor, diameter of the gas and liquid injectors, gas injector depth, gas velocity, and recess ratio. The combustor to which GCSC-A and GCSC-B are applied consisted of 60 injectors. The diameter of the GCSC-B was 1.135 times that of GCSC-A, resulting in a difference in the gas flow velocity. The combustor with GCSC-C was composed of 36 injectors, which differed in terms of the flow rate from the injectors of GCSC-A and GCSC-B. However, the gas injector flow velocity was designed such that it was comparable to that of GCSC-A. The recess ratio of all injectors was set at one. Kerosene was injected tangentially from the side of the annular injector outside the gas injector. The experimental methods for this study could be classified into a spray visualization test for a single injector and combustion test for a multi-element injector. The fluids used in the spray visualization were air and water, and oxidizer gas and kerosene, which were simulated, respectively (Figure 1). The flow rates of the liquid and gas were controlled by adjusting the supply pressure to each injector, and the gas flow rate was changed such that MFR was the same for each injector, as shown in

Table 2. Experimental conditions.

		Spray Visualization Test		Combustion Test
		Water	Air	Kerosene	Gas
Mass-flow rate (kg/s)	GCSC-A	0.0297	0.0047–0.0125	0.089	0.286
	GCSC-B	0.0297	0.0054–0.0142	0.089	0.286
	GCSC-C	0.0487	0.0081–0.0215	0.146	0.484
Momentum flux ratio		15, 25, 35 (2p-SLIPI) 5, 15, 25, 35 (Particle master)		9–23

. For the GCSC-C, fewer injectors were applied to the combustion test compared to GCSC-A and GCSC-B. Further, the injection flow rate per injection was 1.6 times that of GCSC-A and GCSC-B. Two-phase structured laser illumination planar imaging (2p-SLIPI) measured the overall spray shape, and the experiments were performed with an MFR in the range 15–35, which represented an area wherein spray changes by gas were clearly shown. The SMD measurement experiments in narrow areas were conducted in the MFR of 5–35 to analyze the change in the droplet sizes with respect to flow rate.

The SLIPI measurement method was employed to measure the spray cross-sectional shape (Figure 2). In this method, the spray cross-sectional shape was measured by irradiating two modulated phases in a sheet form for a time of 10 µs and removing multiple scattering errors from the two measured values. An ND:YLF laser (Photonics Industries International Inc) of wavelength of 527 nm was used as a light source for the cross-section measurement, and a 1024 × 1024-pixel CMOS high-speed camera (FASTCAM SA5 Type 775K-M2, Photron) was used for photographing. The optical arm was connected to the ND:YLF laser and SLIPI Optic module (Lavision) for plane light irradiation. By synchronizing the computer and laser controller through a programmable timing unit, the spray measurement was performed at 3500 Hz in the 100 × 100 mm² area.

The droplet size in the liquid rocket is an important characteristic in spray field analysis because it determines the fuel vaporization rate. To measure the droplet size according to the injector operating conditions, a particle master (Lavision), which converts a beam laser to volumetric through a diffuser and acquires a shadow image, was used (Figure 3). The light source and controller were the same as that of SLIPI. A laser diffuser device was connected to the rear end of the optical arm instead of an optic module to spread the beam. A high-magnification lens (Lavision), which measures the size of sprayed small droplets in front of the high-speed camera, was also attached. Once the particle master shadow software calibrated the size of the droplet to be measured through the depth of the field calibration plate beforehand, it performed functions such as the droplet size distribution and calculation of the droplet velocity from the shadow image that was captured. In this experiment, the SMD, $D_{32}=\left(\sum n_i{d}_i^3)/(\sum n_i{d}_i^2\right)$), which is primarily used for combustion environment research, was used for data analysis for the droplet size comparison [28].

3. Results and Discussion

3.1. Spray Distribution According to MFR

For all three injectors, the spray cross-sectional structures were studied through the SLIPI measurement. The flow visualization results measured through the SLIPI were in the MFR range 15–35. These values were used for the comparative analysis of the liquid spray flow characteristics that are shown in Figure 4. The measurement results of the cross-sectional structure indicate that as the MFR of all injectors increased, the shape of the liquid injected into the center of the injector became more apparent. Further, the liquid flow rate at the center increased, and the flow rate of the liquid injected outward with a large spray angle decreased. When the MFR was low, the shape of the center resembled a hollow cone with a small injection angle, and as the MFR increased, the shape resembled a hollow cylinder due to the rapid gas flow velocity. The larger the MFR was, the higher the flow rate was of the liquid drawn into the injection center from the sprayed liquid film with a high injection angle. Unlike previous studies that indicated there was an increment in the amount of liquid injected from the liquid injector with a decrease in the injection angle of the liquid injected from the liquid injector, a part of the liquid film that was injected to the outside was torn and drawn in the direction of gas injection [16,18,29]. In addition, the liquid drawn into the interior was characterized by a periodic oblique cross-section toward the center, and this shape became more apparent as the MFR increased. When the MFR was low in the cross-sectional shape of the outer injection of the injector, the sheet was thinner. However, as the MFR increased, the dispersed angle and the thickness of the sheet increased. The higher the gas flow rate, the greater the energy provided to the liquid film injected by the liquid injector in the radial direction through increased gas momentum. This further resulted in an increase in the amplitude of the liquid film splitting into droplets. The difference in the flow distribution between GCSC-A and GCSC-B in the same MFR was not significant. However, in addition to a farther distribution, the GCSC-C with the highest per unit flow rate among the three injectors exhibited a greater flow rate of droplets that were injected in the radial direction compared to other injectors mainly due to the swirl effect.

The intensity distribution was obtained using the Davis 8.4 software (LAVISION), wherein 500 photographs of each spray were acquired according to the MFR at each injector. Figure 5 shows the radial spray intensity distribution at a distance of 30 mm from the injector outlet in the case of when the area was mainly composed of the liquid that was drawn by the momentum of the gas, also referred to as the center of injection. As a result, the flow rate toward the center was clearly distinguished from that sprayed at a large angle by the swirl. The strength of the liquid flowing into the center of the gas was at a maximum at a distance of 8 mm when the MFR was 15, and the spray of all injectors moved toward the center of the injection with a maximum strength of 6 mm for GCSC-A, 7 mm for GCSC-B, and 6.5 mm for GCSC-C when the MFR was 35. When the MFR was 15, the GCSC-A exhibited a maximum strength at 31 mm, GCSC-B at 39 mm, and GCSC-C at 38 mm, and the maximum value of the GCSC-C injector with a high flow per injector was approximately twice the values of the GCSC-A and GCSC-B injectors.

In combustors, injectors are often manufactured separately and are placed on the manifold of the combustor head, ensuring a certain distance between the injectors. The propellant that is sprayed at the center of the injector is predominantly affected by the mixing characteristics of the injector itself. However, the liquid sprayed at a high angle interacts with the one sprayed from the surrounding injector [30]. Figure 6 shows the results of the comparison of the volume of liquid propellant injected toward the center of injector that had less interaction with the surrounding injector, owing to high gas momentum. For different injectors, the sum of the intensity of the liquid flow distributed from the center of the injector in the radial direction, and the value that was twice the diameter of the liquid injector (2D_l) were compared. When the MFR was 15, 16% of GCSC-A, 15% of GCSC-B, and 18% of GCSC-C exhibited a flow rate distribution less than or equal to 20%, and most liquids were sprayed at a high angle owing to the impact of the liquid injector swirl. However, when the MFR increased to 25, approximately half of the sprayed flow rate was dragged into the center of the injector due to the momentum of the gas, with GCSC-A being 48%, GCSC-B 49%, and GCSC-C 45%. When the MFR was 35, GCSC-A increased to 55%, GCSC-B to 58%, and GCSC-C to 49%. The increase in the MFR resulted in a steady rise in the total flow rate of the liquid sprayed at the center of the gas. Nonetheless, the degree of increase decreased. The intensity analysis showed that in an injection environment similar to that with an MFR of 15, the spray of liquid was mainly sprayed outside the injector under the effect of the liquid injector swirl, and the effect of the gas sprayed from the center was relatively small. When the MFR was equal to or greater than 25, the flow rate of the liquid drawn into the gas center increased rapidly. When the operating MFR of the injector in the combustor was less than or equal to 15, the liquid spray that was injected at a high angle affected the combustion reaction, and when the MFR was equal to or greater than 25, the characteristics of the droplets mixed with the oxidant injected from the center of the gas directly affected the combustion reaction.

3.2. Droplet Size

According to the results of a previous study wherein SLIPI was used, the distribution of the liquid spray varies depending on the MFR, and each distribution has a considerable impact on the combustion within the combustor. Thus, the change in droplet size according to the change in MFR was analyzed for different injectors. The measurement experiments were performed in the form of a rectangle of width 7 mm for the parts that increased by 7 mm in the radial direction from the center to a position at a distance of 30 mm from the injector surface. As shown in Figure 7, the droplet size was smaller at a larger distance from the center of the spray. The SMD comparative analysis was conducted on the zones 1–4, and this was suitable for relative comparison when using the particle master. The measurement range of the droplets was 50–1000 μm, and the average was obtained using 500 images.

Figure 8 shows the measurement results of the change in SMD according to MFR. As the MFR increased in all injectors, the SMD decreased, and it tended to converge at specific values in all locations. In Zone 1, wherein two types of fluids are most actively mixed, when the MFR was 5, the SMD of GCSC-A was 165 μm, GCSC-B was 187 μm, and that of GCSC-C was 197 μm. As the MFR increased, the SMD decreased, and all three injectors were distributed in the range 125–140 μm. In Zone 2, wherein the liquid flow rate rapidly decreased, the SMD converged to 190 μm, and the type of injector did not have a significant impact. As the distance from the center of the injector increased, the SMD tended to turn on. In Zone 3, when the MFR was 5, the SMD of GCSC-A was 460 μm, GCSC-B was 420 μm, and that of GCSC-C was 380 μm. When the MFR was 35, the SMD was in the range 240–300 μm. In Zone 4, which is the farthest from the injection center and mainly consists of spraying due to the impact of swirl, when the MFR was 5, the SMD of GCSC-A was 680 μm, GCSC-B was 640 μm, and that of GCSC-C was 560 μm. Furthermore, in this zone, as the MFR increased, the size of the droplet decreased, and the SMD for GCSC-A was 370 μm, GCSC-B was 450 μm, and that for GCSC-C was 420 μm, all of which tended to converge at approximately 400 μm. In all zones, the SMD of GCSC-A, which was slower in gas velocity than in GCSC-B, was mostly large. The GCSC-C was designed such that it possessed the same gas velocity as GCSC-A, and it was measured at a similar level to GCSC-A and SMD. Therefore, the SMD decreased as the gas velocity and gas momentum increased. In Zone 2, it was observed that the GCSC-C possessed a smaller SMD compared to GCSC-A, which had the same design flow rate. This was because the diameter of GCSC-C was 1.3 times the value of GCSC-A. Having said that, because the flow rate was 1.6 times larger, the mixing characteristics in Zone 2, which corresponded to the main interaction area of the gas–liquid, were better.

3.3. Analysis of Spray Visualization and Combustion Test Results

Combustion tests for each injector were performed, as shown in Figure 9, and the SMD measurement results corresponding to the MFR range during the combustion tests are shown in Figure 10. In the case of GCSC-A with the highest characteristic velocity during the full-scale combustion test, the MFR was in the range 20–23. In the case of GCSC-B and GCSC-C with lower characteristic velocities than GCSC-A, the value of the MFR was in the range 9–16 [27]. For the combustor operating conditions of GCSC-A, it was possible to estimate the SMD from Figure 10 with the MFR in the range 15–25. Within this range, the SMD at the center of the injector was within 130–140 μm and 400–450 μm in Zone 4. Furthermore, GCSC-B and GCSC-C exhibited SMD values of approximately 140 μm and 200 μm at the center when the MFR was between 5 and 15 in the graph, respectively. The SMD value increased in the outward direction, and in Zone 4, the value was in the range 430–650 μm.

The combustion efficiency for each injector that was calculated through combustion tests is summarized in

Table 3. Comparison of spray visualization test and combustion test results according to injector types.

Injector Type (MFR Condition)	Spray Visualization Test			Combustion Test
	MFR	Intensity within 2Dl [%]	SMD within 2Dl (Zone 1 & 2)	MFR	Total Injector Number	C* Efficiency [%]
GCSC-A (High MFR)	25	47	139–190	20–23	60	93.2–93.8
GCSC-B (Low MFR)	15	15	164–214	9–16	60	90.4–92.4
GCSC-C (Low MFR)	15	18	143–193	9–16	36	89.4–91.4

along with the spray visualization measurement results. The combustion efficiency represents the ratio of the experimental characteristic velocity to the ideal characteristic velocity as ${\mathsf{\eta }}_{c^{*}}=c_{exp}^{*}/c_{ideal}^{*}$. The combustors with GCSC-A were tested at high MFRs of 20~23, and thus, they were compared to the MFR 25 of the spray visualization test. The combustors with GCSC-B and GCSC-C were tested at MFRs equal to or less than 16, and thus, they were compared to those of the MFR 15 of the spray visualization test. From the comparison, the GCSC-A with the largest MFR had the largest intensity within 2D_l, the SMD was the smallest, and the combustion efficiency was the largest. The GCSC-B and GCSC-C with relatively small MFRs possessed an intensity that was 30% smaller than that of GCSC-A, the SMD was larger than GCSC-A, and the combustion efficiency was low.

In the case where GCSC-A and GCSC-B possessed the same gas design flow velocity of 100 m/s, the SMD was similar in Zones 1 and 2, and in the case of GCSC-B with a low design flow velocity of 75 m/s, the SMD was larger than that of other injectors. With regard to the aforementioned factors, it can be inferred that the SMD was primarily affected by the gas flow velocity. The GCSC-B and GCSC-C possessed similar intensities and SMD sizes, but the combustion efficiency of GCSC-C was relatively low. The total number of injectors in the GCSC-B was 60, and the number of injectors located on the outermost side of the combustion chamber cylinder wall was 24, which corresponded to 40% of the total number of injectors. On the other hand, the total number of injectors in the GCSC-C was 36, and the number of injectors on the wall was 18, which corresponded to 50% of the total injectors. Accordingly, although the MFRs of the two injectors were similar, the fuel flow rate of GCSC-C injected to the wall of the combustion chamber was high and the amount of fuel for complete combustion was low. Thus, the combustion efficiency of the GCSC-C type with a small number of injectors was low.

The cause and effect of the difference in combustion efficiency were inferred by comparing the internal wall conditions of the combustor after combustion tests. A cross-sectional view of a full-scale combustor manufactured for the evaluation of the combustion performance is shown in Figure 11. Figure 12 shows the internal area of the combustor from the perspective of the spray surface of the injector in the nozzle direction, after conducting a combustion test with a combustor installed with GCSC-A, GCSC-B, and GCSC-C. In the flame visualization experiment, soot is mainly generated around the liquid injection, which corresponds to an area with a low mixture ratio [24,25,31]. By identifying the deposition location and deposition concentration of the soot on the wall within the combustion chamber after combustion, the location of the part with a low mixture ratio during combustion can be identified. The soot generated due to the low mixture ratio was only on the wall of GCSC-A close to the spray surface of the injectors, and the bright-colored thermal barrier coating was visible on the remaining portion of the wall. Furthermore, in GCSC-B, a long soot area was formed from the place where the main injectors were located to the part where the secondary film cooling occurred. The baffle injector, which prevents instability in combustion, was designed such that it possesses a higher mixture ratio than the GCSC main injector. Thus, there was a clear color distinction from the dark soot area owing to incomplete combustion of the propellent. Similar to GCSC-B, GCSC-C exhibited a longer soot area; a darker soot was deposited on the wall compared to GCSC-B. The GCSC-C also demonstrated a low MFR operating condition similar to GCSC-B, and thus, a majority of the flow rate was not mixed with the gas at the injector outlet and was sprayed onto the wall. Furthermore, because the flow rate per injector was higher than that of GCSC-B, a relatively large amount of soot was deposited on the combustor wall owing to the low mixture ratio. According to Kirchberger et al., as the flow rate of the propellant used for film cooling increases, the wall temperature and combustion efficiency decrease [32]. In an ideal combustion environment, the sprayed propellant burns with maximum heat according to the design mixture ratio. However, on the wall, the film cooling and unmixed propellants at the injector outlet are combined to form a low mixture ratio, thus resulting in a low combustion temperature and efficiency [33]. The injectors that are operated at a low MFR exhibit a large amount of fuel that is injected toward the inner wall of the combustor, and soot is deposited on the wall owing to incomplete combustion. In addition, the liquid propellant that was not mixed with the gas is mixed with the film cooling propellant that is sprayed to protect the wall inside the combustor, thus reducing the combustion efficiency.

Thus, when the MFR was less than 15, the gas momentum was small, and thus, the liquid film sprayed from the liquid swirl injector could not be divided efficiently. Consequently, the injector simply acted as a liquid swirl injector, and the liquid propellant was sprayed with relatively large droplets at a high angle. However, the injector designed with a high MFR divides a larger amount of propellant into smaller droplets and sprays it into the center of the injector (Figure 13). It is believed that such a difference in spray shape of a single injector affects the distribution of the propellant in the combustor atomization zone, thereby causing a difference in combustion efficiency. With regard to the aforementioned results, quantifying the liquid flow rate and SMD that is sprayed around the gas through the visualization of droplets according to the shape of the injector in the atmosphere is an effective approach to select an injector with a high combustion efficiency at a lower cost, rather than directly performing combustion tests on various types of injectors.

4. Conclusions

Spray visualization tests were conducted to determine the spray characteristics of the GCSC injectors used in the staged combustion cycle engine, and the spray shape and droplet characteristics obtained were analyzed and compared with the results of the full-scale combustion test.

-

Spray cross-sectional shape: As the MFR increased, the liquid spray flow rate increased in the gas injection direction, which was toward the center of the injector. If the MFR was less than or equal to 15, only 20% or less of the spray flow rate was sprayed in the direction of the gas spray. However, when the MFR was 25 or higher, the liquid spray flow rate increased to 45% or higher. In injectors that were operated at a low MFR of approximately 15, most of the liquid was sprayed in a cone shape with a large injection angle. However, as the MFR increased, it was sprayed in a cylinder shape toward the center of the injector.

-

SMD: The lower the MFR, the smaller the SMD. Further, the SMD increased from the center of the spray outward. At the same MFR value, the injectors with higher gas velocities exhibited smaller SMDs. As the MFR increased, the SMD of all injectors decreased.

-

Under the engine operating conditions, nearly half of the GCSC-A was sprayed into the center by an interaction with gas injection, and the droplet size at this instant was in the range 140–150 μm. Furthermore, under operating conditions, 80% of the injected liquids of GCSC-B and GCSC-C comprised large droplets with dimensions equal to or greater than 250 μm and were sprayed to the side with a high angle due to the swirl.

-

When the MFR increased in a full-scale combustor with multiple injectors arranged, the propellant flow rate injected in the form of a cylinder increased in the center of the injector, and the propellant mixing characteristics for the central flow rate played a more important role in the combustion efficiency. An injector operated under conditions with a large MFR exhibited a large ratio of small SMD droplets sprayed toward the center of the injector and the highest combustion efficiency. Furthermore, the combustion efficiency of the injector with a large amount of SMD injected with a high injection angle was relatively low.

The spray shape and droplet distribution were studied through GCSC injectors spray tests according to different shapes and spray conditions, and the combustion efficiency according to the combustion tests results were compared and analyzed. As a result, it was found that there was a difference in combustion efficiency according to the spray characteristics of a single injector. If a spray shape that is suitable for the operating environment of a staged combustion cycle engine is selected through a comparative analysis of the spray characteristics of a single injector in the design stage, the number of expensive full-scale tests can be reduced, thus resulting in a decrease in the development duration and costs. Future work may involve experimental studies on the design of GCSC injectors, which may further increase the combustion efficiency under various shapes and experimental conditions.