Effect of temperature on atomization in gas centered coaxial injection systems

Transcription

1 Effect of temperature on atomization in gas centered coaxial injection systems V. Sasi Prabhakaran, D. Sivakumar, C. Oommen, and T. J. Tharakan Abstract This paper summarizes the current status of research work carried out at IISc on the characterization of liquid sprays developing from gas centered coaxial injection systems. The coaxial system injects an annular swirling liquid sheet and a central gaseous jet. The interaction process between the swirling liquid sheet and the non-swirling gaseous jet results in a fine spray. The present study attempts to understand the effect of temperature on the atomization process of outer liquid sheet by conducting experiments with custom made gas centered swirl coaxial atomizers. Limited results show that the liquid temperature significantly influences the spray characteristics. The study also highlights the effect of orifice recess on the breakup behavior of liquid sheets. T I. INTRODUCTION he present project study is a part of ongoing research work at IISc on gas centered swirl coaxial injection systems with water and air as simulants. Such injection systems are employed in semi-cryogenic engines operating with liquid oxygen (LOX) and kerosene (RP1) as fuel. A gas centered swirl coaxial atomizer discharges a central gaseous jet surrounded by an annular swirling liquid sheet. The entrainment developed by the gaseous jet bends the liquid sheet inwards and thereby establishes the mixing between the two jets. Very few works have been reported in the open literature on the gas centered injection systems. Muss et al. [1] and Cohn et al. [2] reported systematic experimental studies on the sprays developing from premixing type gas centered injectors where the liquid fuel This work was supported in part by the IISc-ISRO Space Technology Cell under the project ISTC/MAE/SK/21. The authors gratefully acknowledge Prof. B. N. Raghunandan, Department of Aerospace Engineering, Indian Institute of Science for providing his laboratory facilities for this project study. V. Sasi Prabhakaran is with the Department of Aerospace Engineering, Indian Institute of Science, Bangalore 6, INDIA.(phone: ; sasiprabhakaran@gmail.com) D. Sivakumar is with the Department of Aerospace Engineering, Indian Institute of Science, Bangalore 6, INDIA (phone: ; fax: ; dskumar@aero.iisc.ernet.in). C. Oommen is with the Department of Aerospace Engineering, Indian Institute of Science, Bangalore 6, INDIA ( coommen@aero.iisc.ernet.in). T. J. Tharakan is with the Liquid Propulsion Systems Centre, Valiamala, Thiruvananthapuram 697, INDIA ( tharakan@yahoo.com). is injected into the gaseous port tangentially. The spray qualities, analyzed via both the hot and cold tests reveal that the gaseous port geometry plays a significant role in determining the atomization quality. It was evident through hot tests that the premixing type gas centered injection systems are very sensitive to the thermal stresses developed in the combustion chamber. A comparative study between the premixing type and coaxial type gas centered injection systems was conducted by Soller et al. [3]. The authors found that though the quality of atomization does not show a significant variation between the two types, the coaxial type injection system exhibit an improved performance at high combustion chamber pressures. It is evident from these studies that the geometrical parameters such as the recess length (the distance between the exit planes of the annular and inner orifices of the injector) and orifice geometry play a strong role in governing the quality of atomization. Recently a preliminary investigation [] was carried out to understand the atomization process of liquid sheets by the central gaseous jet discharging from the gas centered swirl coaxial atomizers. It was observed that the presence of central air jet in the core of the swirling liquid sheet significantly changes the breakup behavior of the liquid sheet. The breakup region of liquid sheet is marked by features such as ejection of liquid ligaments from the sheet surface, localized droplet clusters, increased surface corrugations on the liquid sheet and cellular structures (thin membrane of liquid sheet surrounded by thick rim) on the liquid sheet. Quantitative measurements of liquid sheet breakup length suggest that the breakup length of low inertia liquid sheets decreases with increasing air jet Reynolds number and the decrease in sheet breakup length is less severe for high inertia liquid sheets. The current research work on gas centered swirl coaxial injection systems attempts to accomplish the following tasks by employing custom fabricated gas centered swirl coaxial atomizers: (a) Effect of liquid temperature on the atomization process of liquid sheets discharging from the outer orifice of gas-centered swirl coaxial atomizers, (b) an elaborate experimental study on the influence of atomizer geometrical parameters on the characteristics of sprays

2 discharging from such atomizers, and (c) an understanding on the temporal characteristics of liquid sheet atomization process. along the spray axis at which the liquid sheet breaks up into droplets and the spray cone angle were measured from the II. EXPERIMENTAL DETAILS Experiments were carried out with tap water and air as working fluids. A schematic diagram of the injector assembly is given in Figure 1. It consists of an inner orifice and an outer orifice to discharge air and water respectively. Water inlet Swirl chamber Outer orifice plate Air inlet D s Di D o Figure 2. Test facility to conduct experiments with hot water photographs. Swirler Air injector Inner orifice lip III. RESULTS AND DISCUSSION Figure 1. Schematic diagram of the gas centered swirl coaxial atomizer. Following geometrical parameters are varied: (i) orifice recess,δ (ii) the flow passage dimensions of swirler, and (iii) outer orifice diameter, D o. The swirling motion was imparted to the liquid by passing the experimental liquid through a helical swirler placed upstream of the converging section of the outer orifice. The helical swirler comprises of six rectangular flow passages on the outer surface. The geometrical variation resulted in different gas centered swirl coaxial atomizer configurations (see Table 1). The experiments were conducted with both cold and hot water. Previously arranged spray test facility was used to carry out Table 1. Details of recessed injector configurations Inj. code Di (mm) Do (mm) Ds (mm) w (mm) h (mm) # of starts (mm) OA() OA(+1) OA(-1) OA (-2) experiments with cold water (ambient temperature condition). New test facility was constructed to carryout experiments with hot water as in the Figure 2. No detailed discussion on the test facility is given here for the sake of brevity. A thermocouple was installed just before the water inlet of the atomizer assembly to measure the temperature of water. Photographic techniques were employed for the visualization of spray discharged from the coaxial injector. A Nikon D3 digital SLR camera was used to take the photographs of the spray with good image resolution. The breakup length of the annular liquid sheet, estimated as the distance between the orifice exit and the axial location A. Effect of orifice recess Three different recess configurations of gas centered coaxial swirl atomizers were examined. Figure 3 shows the schematic sketches of orifice exit of the atomizer assembly δ (mm) Figure 3. Orifice exit under recessed configurations. under three recessed configurations. Systematic experiments were carried out with and without the presence of central air jet. Figure shows the breakup length of liquid sheet versus Weber number ( ) plot for the atomizers with recessed configurations without the presence of central gaseous jet. The measurements were Cold water (ambient temperature) No central air jet Orifice recess, δ (mm) Figure. versus for atomizers with recess

3 compared with those obtained without recess. It is very evident from the figure that the orifice recess drastically influences the liquid sheet breakup process. Examination of liquid sheet images reveals that the positive recess reduces the breakup length significantly with lot of perturbations on the liquid sheet surface. The influence of central air jet on the breakup process of outer liquid sheet with recess is shown in Fig.. For each atomizer configuration, the influence of central air jet is B. Effect of liquid temperature (without central air jet) Experiments were carried out with heated water. Note that the increase in temperature is expected to decrease the surface tension of liquid and such an effect can alter the flow behavior of outer liquid sheets. The effect of liquid temperature on the flow behavior of liquid sheets discharging from the gas centered coaxial swirl atomizer with orifice recess -2 mm is illustrated in Fig. 6. Significant influence of liquid temperature is seen from the images 2 1 Cold water(ambient temperature) With central air jet Orifice recess,δ(mm)= Cold water(ambient temperature) With central air jet Orifice recess,δ(mm)= Cold water(ambient temperature) With central air jet Orifice recess,δ(mm)= Reg Figure. versus for atomizers with recess: Influence of central air jet. examined for different values of outer liquid sheet. The influence of central air jet is more pronounced for the atomizer configuration with negative recess. Interestingly, the influence of air jet on the sheet breakup is very negligible for the atomizer configuration with positive recess. Note that the liquid sheets discharging from this atomizer configuration exhibit larger spray cone angle along with shorter breakup lengths. Owing to this behavior, the liquid sheets do not recognize the presence of central air jet. given in Fig. 6. The higher temperature develops perforations on the surface of liquid sheet and thereby makes the liquid sheet to breakup at shorter distances from the orifice exit. Also the hot water liquid sheets show a wider cone angle compared to cold water liquid sheets. The increase in liquid temperature drastically reduces the breakup length of outer liquid sheet. Quantified variation of breakup length of liquid sheet with for the atomizers discharging heated water is shown in Fig. 7. A comparison of measurements between Fig. and Fig. 7 clearly reveals Liquid sheets with cold water (ambient temperature) Liquid sheets with hot water (2 deg C) Figure 6. Effect of liquid temperature on the flow behavior of liquid sheets discharging from the gas centered swirl coaxial atomizer configuration with orifice recess -2 mm. Note that the flow conditions are comparable between the two cases. Liquid flow rate is increasing from left to right.

4 Heated water (2 deg C) No central air jet Orifice recess, δ (mm) Figure 7. versus for atomizers with recess. Experimental liquid is heated water that the increase in temperature significantly reduces the breakup length of outer liquid sheet. Observations are made from the experiments carried out under elevated water temperature of 2 o C. The above plot between breakup length and Weber number of the corresponding flow condition lucidly states the variation of the breakup length is associated with the negative recess configuration. For zero recess condition, the breakup of liquid sheet remains in a same range for most of the values between higher and lower Weber number values. Positive recess has a pattern of very low breakup even in cold flow condition; with hot liquid, the pattern remains the same for positive recess configuration. The trend followed by the negative recess can be seen from Figure, which shows the same pattern for both -1 and -2 recesses. The value of liquid sheet breakup length decreases gradually with increase in the Weber number value. C. Effect of orifice recess with temperature (liquid with central gaseous jet) I) Higher of.3 and 9.76 The liquid sheet characteristics is not much altered for higher flow rate condition i.e. the breakup length of the liquid sheet is found to be in the same range for = 13 and 13. The general trend in the breakup length shows the gradual decrease in the breakup length of liquid sheet with increase in the value up to 2x1 and its rapid decrease for higher value of. From the plot shown in the Figure 2, δ = + 1 mm[ hot flow, 2 deg C] configuration yields a very low breakup length compared to other configurations with recess mm, -1mm and -2mm. At lower values of the difference in breakup length for δ = + 1 mm is significantly higher and it gradually reduces for higher. This can be attributed to dominant influence of the dynamics of interaction of the air jet and liquid sheet near the injector exit, which even influences the entrainment levels at axial locations further downstream. The liquid sheet behavior at δ = + 1 mm configuration is marked by a reduction in the perforations on the liquid sheet due to shorter break up length owing to larger cone angles observed. This greater value of cone angle may be due to the increased area of contact of the liquid sheet with the air orifice leading to more friction and also the flow induced by the liquid sheet on the adjacent ambient air is enough to push the sheet outwards. But it is inferred that, the breakup length more or less remains the same for all increasing value of ; which reveals a very low air-liquid sheet entrainment phenomenon, this is also due to higher cone angle. With increasing the effect of projection/ recess seems to diminish as all the breakup length try to attain the same value. This can be explained by the fact that, at higher the air jet may directly hit the enveloping liquid sheet drastically leading to the sheet breakup. II) Lower of 61.9 In the case of lower values of Weber number [61.9] of hot flow, the pattern observed in the breakup length show a drastic decrease with negative recess case of -1 and -2 mm configuration. But the pattern in co planar and positive recess [+1 mm] configurations is the same as observed in higher ; especially in the case of co-planar configuration the value of breakup seems to increase when the sheet is converged and decrease when the sheet is diverged. In general, the negative recess gives a reasonable value of breakup length from an atomization point of view, but this can rigidly stated with drop size measurement. This characteristic of liquid sheet with negative recess is mainly due to the high possibility of air-liquid entrainment; as the air jet is delivered before the liquid sheet feeding out of the injector exit, it can eventually lead to air-liquid pre mixing. Even though the positive recess provides low breakup length, its atomization performance is low when compared with the negative recess configuration. As the cone angle is 1 δ = -2 mm δ = -1 mm δ = mm δ =+1 mm W e =.3 Liquid Temperature = 2 o C 1 δ = -2 mm δ = -1 mm δ = mm δ =+1 mm W e = 9.76 Liquid Temperature = 2 o C 1 δ = -2 mm δ = -1 mm δ = mm δ =+1 mm W e = 61.9 Liquid Temperature = 2 o C R e [gas] R e [gas] Figure. Cumulative data on breakup length is plotted against the corresponding Reynolds number, for different configuration and Weber number,. R e [gas]

5 comparatively very high for positive recess, this configuration lead to a high thermal load on the combustion chamber wall and the orifice head of injector. D. Comparison of liquid sheet characteristic at cold and hot flow conditions. I. Negative and zero injector recess configurations: The comparisons are carried out between the cold and hot liquid sheets under identical flow condition of psi, psi and 2 psi of liquid. The heated liquid (water) at 2 o C has a lower co-efficient of viscosity value than water at ambient temperature. The corresponding value of Weber number will vary with liquid temperature and hence comparisons are made according to the flow conditions. In the case of δ = -2mm, the cold and hot liquid sheets behave a bit in the same fashion. Plot between higher = 13 [hot] and = 19 [ambient condition] in the figure 9, shows that liquid sheet breakup and cone angle remains not much altered with increase in Reynolds number value of central gaseous jet for hot flow condition. But in the case of liquid at ambient temperature, liquid sheet breakup length shows a decrease in the value as increases. This states that central gas jet has an influence on the liquid sheet breakup. But with very high, the liquid sheet break up length may come down considerably. The same trend depicted in the above case is followed for = 9 [hot] and = 11 [ambient condition] for both hot and cold flow conditions. But for low value of, liquid sheet break up length comes down considerably with increase in value in the case of cold flow; heated liquid sheet is not giving as much response to central gaseous jet as liquid sheet at ambient temperature. With δ = -1mm, liquid at ambient temperature breaks up into drop much shorter than heated liquid for = 11, [ambient temperature] and = 13, 62 [hot]. The negative recess influencing the air-liquid pre mixing may lead to this variation in the breakup length in cold flow condition. The trend followed by this configuration is shown in the Figure 11. The sheet is not tending to bend as in cold sheet; this is due to low viscosity and surface tension. The negative recess of -1mm influences the airliquid entrainment process at the near exit of orifice leading to low cone angle (Figure 1). 2 δ= -2 mm Cold [W e = 19] = 13] 2 δ= -2 mm = 9] Cold [W e = 11] 2 δ= -2 mm = 62] Cold [W e = ] 2 δ= -1 mm =.3] Cold [W e = 1.73] 2 δ= -1 mm = 9.76] Cold [W e = 1.91] 2 δ= -1 mm = 61.9] Cold [W e =.9] 2 δ= mm =.3] Cold [W e = 1.73] 2 δ= mm = 9.76] Cold [W e = 1.91] 2 δ= mm = 61.9] Cold [W e =.9] Figure 9. Hot and cold flow comparison for the configurations of -1,-2 and mm recess for higher and lower values of We for same liquid pressure range.

6 θ deg hot flow with central air jet 2 o C δ= δ= +1 δ= 1 δ= -2 We =13 We =13 We = 62 flow with central air jet at δ= δ= +1 δ= 1 δ= -2 ambient temperature We =19 We =11 We = Figure 1. Cone angle θ (deg) versus for various recess θ deg clusters. In the case of heated flow, the sheet is not influenced by the air jet. No convergence of the heated liquid sheet is noticed, this may be due to a considerably low viscosity of heated liquid. In the case of co-planar configuration, the trend followed in liquid sheet breakup is same as observed in δ = -1mm configuration, but in higher magnitude. The hot condition remains not affected by the central air jet and the cold sheet breaks at a lower range than hot liquid for lower Weber number. This phenomenon is depicted in the Figure. Figure 11. Comparison between heated and cold liquid flow conditions for δ = -1mm Note that the liquid sheet with = 1.91 takes a converged jet shape in the presence of central air jet of = 63. as seen in the Figure 11. The presence of central air jet with low pulls the liquid sheet towards the spray axis as seen in the second and third images of the bottom row of Figure 11 and results in a shift in the sheet breakup location towards the orifice exit. The central air jet impinges directly on the inner surface of liquid sheet at these flow conditions which may be leading to the generation of chunks of liquid masses immediately after the breakup. The liquid sheet collapses dramatically with further increase in and the breakup of liquid sheet occurs in the near region of the orifice exit at higher in the case of cold flow condition. The axisymmetric conical shape of liquid sheet as observed at lower values of is no longer seen at these higher values of and the sheet breakup occurs in a more complex manner. Discrete liquid sheets bounded by thick rims are dominantly seen at these flow conditions. The sheet coming closer to spray axis is marked by profuse ligament formation. The ligaments disintegrate into drops in a manner similar to that of liquid columns breaking up into drops and the distribution is in The influence of central air jet on the breakup of heated liquid sheets is shown in Fig.. The trends are similar to the behavior observed for the tests conducted with cold water. The decrease in breakup length with increasing gas Reynolds number ( ) is almost insignificant. A possible reason may be the wider spray cone angle and shorter breakup length of outer liquid sheets with =. These experiments clearly points out that an effective interaction between the central air jet and the outer liquid sheet can be realized only if the jets are located very near to each other. Liquid sheet with hot water (2 C) Liquid sheet with cold water (ambient temperature) Figure. Comparison between hot and cold flow condition for co-planar configuration

7 Attention need to be focused on this aspect in future experiments. II. Positive recess configuration δ = +1mm Positive recess configuration allows the liquid sheet to get diverged yielding to its large cone angle for both hot as well as cold conditions of liquid. Due to this divergence, the air-liquid entrainment is very low in the presence of central gaseous jet. The break up length in hot flow is noticed to be lower in magnitude than in cold flow. The liquid sheet characteristics of hot liquid are compared with liquid at ambient temperature as depicted in the Figure 13. Figure 13. Effect of positive recess [+1 mm] showing variations in the liquid sheet characteristics for different conditions of and under hot and cold condition in comparison with co-planar [δ = mm] configuration. As discussed above, δ = +1 mm configuration is seen to have a lower breakup length in hot flow. But in cold flow condition, liquids sheets with lower are influenced more by the central air jet; but at higher, the influence of air jet is gradual. For liquid flow at ambient temperature, the aerodynamic shear developed over the surface of liquid sheet causes waves to grow and these waves separate from the sheet in the form of ligaments which subsequently disintegrate into droplets. This can be seen in Figure 13. This aerodynamic mode of sheet breakup takes place as a consequence of air friction which arises due to the introduction of co-flowing air in the system. Figure 11 indicates the formation of cellular structures (thin liquid membrane bounded by thick rims) in the liquid sheet with increasing value of. The entrainment developed by the gaseous jet bends the liquid sheet inwards and thereby establishes the mixing between the two jets leading to the formation of cellular pattern and ligaments. These thin membranes break up into finer droplets and the rim falls apart as larger drops thus highlighting the importanc e of s preading the liquid to achieve the finest atomization. E. Temporal behavior of the breakup process of outer liquid sheet The previous investigation [] showed that the breakup process of outer liquid sheet by the central air jet results in the generation of thick liquid ligaments. These ligaments are one of the reasons for the localized droplet clusters present in the spray. The breakup process of outer liquid sheet by the central air jet was examined by using a high speed digital imaging system. The following qualitative details were extracted by analyzing these image sequences. a) The spray exhibits oscillatory characteristics with low values of. Such oscillatory behavior could lead to dramatic changes in other spray characteristics such as spray cone angle and liquid mass distribution. b) The development of thick rims at the free edge of outer liquid sheet may be the source for the ligament generation in gas centered swirl coaxial atomizers. More details emerg e out after systematic analysis of these image sequences. The breakup phenomenon incurred in hot flow condition is due to the formation of perforation. As seen in Figure 11, the liquid sheet produced under hot flow is not subjected to cell structure and ligament formation as noticed in the cold flow conditions; rather perforations are noticed. The entire sheet is more or less straight and not curved as in cold flow, even though the hot liquid sheet is influenced by the central air jet. Viscosity and surface tension of water at high temperature tend to change accordingly to influence the nature of the liquid sheet formation more in the case of p ressu re swirl atomization. Perforation mechanism: Water at normal temperature has some amount of dissolved gases in it, micro bubbles of air are noticed in liquid sheet even at ambient temperature. But these bubbles are very less in size and do not cause any variation in breakup length and perforation at ambient temperature. Fraser et al [] investigated the role of dissolved air in the formation of perforations. They obtained photographs of a water sheet produced by a fan spray nozzle. The sheet was formed in a vacuum in order to suppress aerodynamically induced waves. Both aerated and deaerated distilled water sheets were examined. The content was reported to have no effect on the formation of holes or on the breakup length of the sheet. Clark and Dombrowski [6] also investigated the role of the release of dissolved air on hole formation. The y s prayed tap water, which was found to be supersaturated

8 high speed imaging reveals that the free edges of liquid sheet develop rims thereby generating thick ligaments in the spray. Figure 1. Mechanism of perforation formation in hot liquid sheet. with air, and distilled water with the air content reduced to 1 p.p.m. At conditions when perforations were formed, no difference in the breakup of the sheets could be detected. In the case of heated liquid, the dissolved air will club together and expand to form air bubbles within the heated liquid sheet. As the liquid sheet thickness (t f ) gradually decreases descending the injector exit, the air bubble will be exposed out of the sheet and this eventually causes air bubble rupture and perforation growth. Perforation forms randomly with some pattern confirming the breakup of the liquid sheet into droplets soon after the perforation. REFERENCES [1] J. A. Muss, C. W. Johnson, R. Cohn, P. A. Strakey, R. W. Bates, and D. G. Talley, Swirl Coaxial Injector Development. Part I, Test and Results, JANNAF Combustion Subcommittee, 22, Florida. [2] R. K. Cohn, P. A. Strakey, R. W. Bates, D. G. Talley, J. A. Muss, and C. W. Johnson, Swirl Coaxial Injector Development, AIAA 23-, 23. [3] S. Soller, R. Wagner, H. -P. Kau, P. Martin, and C. Maeding, Characterization of Main Chamber Injectors for GOX/Kerosene in a Single Element Rocket Combustor, AIAA 2-37, 2. [] D. Sivakumar, C. Oommen, and T.J. Tharakan, Atomization of Co- annular Swirling Liquid Sheet by a Central Gaseous Jet, Technical Reprot ISTC/MAE/SK/179. [] Fraser RP; Dombrowski N; Routley JH (1963) The atomization of a liquid sheet by an impinging air stream. Chem Eng Sci 1: [6] Dombrowski N; Johns WR (1963) The aerodynamic instability and disintegration of viscous liquid sheets. Chem Eng Sci 1: D. Future work Considerable amount of experimental measurements have been obtained on the breakup length and spray cone angle characteristics. The measurement of drop size distribution using Malvern particle sizer was attempted. However, a few modifications in the injection systems are needed to obtain meaningful measurements. Attempts will be made in this direction. Data of spray mass distribution will also be obtained in near future. More experiments with heated liquid will be conducted. IV. SUMMARY An experimental study was carried out to understand the effect of liquid temperature on the flow characteristics of the liquid spray developing from coaxial type gas centered injection systems. The experiments were carried out using locally manufactured injection systems with water and air as working fluids. The increase in temperature of water makes the liquid sheet to breakup at a shorter distance from the orifice exit. In addition, the rise in temperature also develops perforations on the surface of liquid sheet. The current measurements suggest that an effective interaction between the outer liquid sheet and the central air jet can be realized only if the outer liquid sheet and the inner air jet are positioned closer to each other. The experiments using