Wall Impingement Process of a Multi-Hole GDI Spray: Experimental and Numerical Investigation

Gratis copy for Alessandro Montanaro Copyright 2012 SAE International E-mailing, copying and internet posting are prohibited Downloaded Monday, March 19, 2012 05:04:19 AM Wall Impingement Process of a Multi-Hole GDI Spray: Experimental and Numerical Investigation 2012-01-1266 Published 04/16/2012 Alessandro Montanaro Istituto Motori CNR Simone Malaguti Univ of Modena & Reggio Emilia Salvatore Alfuso Istituto Motori CNR Copyright © 2012 SAE International doi:10.4271/2012-01-1266 ABSTRACT The Direct Injection (DI) of gasoline in Spark Ignition (SI) engines is very attractive for fuel economy and performance improvements in spark ignition engines. Gasoline direct injection (GDI) offers the possibility of multi-mode operation, homogeneous and stratified charge, with benefits respect to conventional SI engines as higher compression ratio, zero pumping losses, control of the ignition process at very lean air-fuel mixture and good cold starting. The impingement of liquid fuel on the combustion chamber wall is generally one of the major drawbacks of GDI engines because its increasing of HC emissions and effects on the combustion process; in the wall guided engines an increasing attention is focusing on the fuel film deposits evolution and their role in the soot formation. Hence, the necessity of a detailed understanding of the spray-wall impingement process and its effects on the fuel distribution. The experimental results provide a fundamental data base for CFD predictions. In this paper investigations have been performed using a 7hole injector, 0.179 mm in hole diameter, spraying in a constant volume vessel with optical accesses. To examine the effects of various factors on development of the spray impinging on the wall, experiments have been conducted at different injection pressures, diverse wall inclination angles and at atmospheric pressure. The acquired images have been processed for extracting the characteristic parameters of the impinging fuel at the different operative conditions. The multi-hole spray has been simulated by Star-CD code taking into account the commercial gasoline properties and the real mass flow rate derived from experimental measurements. In order to correctly reproduce spray impingement and fuel film evolution, a numerical methodology has been defined. Lagrangian sub-models and numerical parameters have been validated against experimental results. INTRODUCTION A great effort is paid by engine manufacturers in order to gain a deep knowledge of the GDI impact on the engine design and operation. It is well recognized the many advantages of the direct injection of gasoline joint with the supercharged technology in terms of fuel efficiency, drivability and lower size/power ratios [1, 2]. The charge cooling through in-cylinder injection and the ability to improve scavenging at full load operations leads to an increase in engine brake specific power. Downsized homogeneous-charged GDI engines with boost lead to relevant fuel savings compared to current production PFI gasoline engine. Stratified-charge lean GDI combustion provides even further fuel consumption reductions. In terms of CO2 tailpipe emissions, boosted downsized stratified GDI engines approaches turbo-diesel one for given displacement. Gratis copy for Alessandro Montanaro Copyright 2012 SAE International E-mailing, copying and internet posting are prohibited Downloaded Monday, March 19, 2012 05:04:19 AM Since the European Commission proposal aims to reduce average fleet CO2 emissions down to 95 g/km by 2020, which would imply a halving of fuel consumption with reference to 1990 and the recently increased limitations to both NOx and Soot emissions, the engine researcher are pushed to rediscover gasoline engines. As a consequence, massive investments for the development of new technologies and strategies have been made, mainly targeted at increasing fuel conversion efficiency together with reducing engine-out emissions. Particular care is devoted to the study of new injection and combustion systems: the injection pressure is constantly raising as well as new complex control strategies are under development. As far as the fuel spray morphology and its interaction with the combustion system has to deeply investigated, the fuel mixture formation within the combustion chamber is continuously growing in importance. Among the many complex phenomena occurring during the fuel mixture preparation such as in-cylinder flow motion, piston design, injector location, injection pressure, timing and strategy and spray characteristics, the fuel/wall interaction could play an important role. When the engine runs at very low engine speeds and low engine loads, as the idle condition, the reduced fuel injection pressure and the need of particular injection timing could lead to fuel deposits on the combustion chamber [3]. For stratified cold operation, the spray impingement on piston may yield to film formation and large HC and soot exhaust emissions [4]. Furthermore wall guided engine architecture is concerned the fuel mixture formation (e.g. stratified charge) is mainly influenced by the spray/wall interaction. It is well known that an impinging spray has a fine atomization and good spatial distribution of fuel droplets. However, both the spatial distribution of fuel and air entrainment into the impingement spray have a great influence on combustion characteristics. Literature reports the impingement angle [5] and the position of the impinging spray [6] are important factors for mixture formation near the combustion chamber wall. Investigations by means of experimental and numerical comparison have been performed by many researchers. A GDI spray-wall interaction inside a heated pressurized chamber, using various visualization techniques, results the effects of the on the upward spray vortex inside the spray and on the horizontal spread after impingement [7]. The predictive capability of the CFD impinging models have been investigated by Walkins and Park [8] and by Bai and Gosman [9], founding that models based on single-droplet experiments are insufficient to calculate the spray-wall impingement process. Finally, literature are deeply studying the impinging fuel spray for GDI application such as spray generated by hollow cone swirl injector [10,11,12]. In this study the fuel and wall interaction is investigated for a spray generated by a second generation multi-hole injector of a current production GDI engine adopting a wall guided architecture. Injection pressures, spacing from idle engine operating condition strategy to full load operating condition, are investigated by means of optical apparatus. Impinging spray is then reproduced using commercial code Star-CD, focusing on the capability of the CFD models. This paper resumes a preliminary activity focused on the characterization on the multi-hole GDI injectors, in terms of wall impact and aims to analyze the effect of the injection pressure and the inclination angle of the impinging wall. The aim of this work is to increase the understanding of the complex phenomena such as the spray and wall interaction in view of subsequent fuel mixing analysis within the combustion chamber of a wall-guided GDI engine. EXPERIMENTAL APPARATUS The GDI injector considered in the present work is the minisac seven-hole Bosch HDEV 5.1 with solenoid actuation. This injector is characterized by a hole diameter of 0.179 mm and with static flow rates of 13.7 g/s at the injection pressure of 10 MPa. On a plane perpendicular to the injector axis placed at a distance of 30 mm from the holes themselves, the seven jet directions give a spray footprint characterized by a hollow-ellipsoid shape (Figure 1). Figure 1. Holes distribution and spray footprint on a plane placed at 30 mm from the injector tip. Commercial gasoline has been used (ρ=740 kg/m3), within all the experimental campaign, as delivered by a hydropneumatic injection system without rotating organs. A reservoir tank of 1.0 dm3, located before the electro-injector, assures the absorption of the pressure oscillations produced from the needle opening and the gas recharge. Further details of the apparatus are reported in ref. [13]. The injection system and synchronized acquisition set up have been managed by a programmable electronic control unit (PECU). This is an open system able to reproduce the injector energizing currents for the desired strategy in terms of number of injection pulses, durations, rise and dwell times. The PECU reproduces in exit a TTL signal, relate to the injection event, for synchronizing the consecutive images management and Gratis copy for Alessandro Montanaro Copyright 2012 SAE International E-mailing, copying and internet posting are prohibited Downloaded Monday, March 19, 2012 05:04:19 AM acquisitions. The injector exhibits a great flexibility to range from low to high load engine conditions [14]. The gasoline mass flow rate has been measured by means of an AVL Fuel Injection Gauge Rate System working on the Bosch tube principle [15, 16]. The fuel is sent into a small chamber kept at the constant pressure of 0.8 MPa and connected to a pipeline of constant internal section, 12 m long. An AVL GM12D pressure transducer collects the pressure variation produced by the fuel delivered from the nozzle. The pressure variation, Δp, is proportional to the injected quantitiy, q, through the following relationship: impingement. The plate has been located at 20 mm from the nozzle tip, thanks to a X-Y-Φ micrometric apparatus, in two different mode: parallel and inclined of 30° respect the total spray axis, fully framed by the CCD camera. The average roughness of the impact-wall is 1.077 µm, measured by the Stylus Profilometer, Model Surtronic 3 by Rank Taylor Hobson. The plate was at room temperature. The experiments on spray development and atomization characteristics of the impinging sprays have been performed in terms of injection pressure (Pinj) and wall inclination angle (α) at the impinging distance (dwall) of 20 mm. All the spray impacts have been performed at ambient pressure and temperature. The detailed experimental conditions are listed in Table 1. Table 1. Experimental conditions for impinging spray where ρ is the fuel density, Atube the inner section of the pipe and a the speed of the sound in the fuel. A time resolution of 10 µs is adopted, suited with the 130 KHz natural frequency cut of the transducer (7.6 µs). The measured q value gives the injected quantity in the time interval 10 µs, while the total injected quantity per shot is evaluated as the integral over the entire pulse duration. The measurements are average over 100 shots and the total amount is compared with the one collected at the discharge pipe, as weigthed through a high precision balance and the discrepancies were always lower than 2%. The spray visualization is made by injecting the fuel in a high pressure quiescent vessel, optically accessible, containing nitrogen at the pressure of 0.1 MPa and the temperature of 300 K. Images of the spray, enlightened by powerful flashes, have been collected at different instants from the Start of Injection (SOI) by means of a synchronized CCD camera, 1376×1040 pixels, 12 bit resolution, 0.5 µs shutter time. An overview of the experimental scheme for the image acquisition is reported in ref. [13]. The images have been acquired using a 50 mm focal lens reaching a calibration factor of 10 pixel/mm. The captured images have been processed off-line by means of a proper software able to extract the parameters characterizing the spray dynamics. A set of 5 images has been collected for each injection condition for a statistical analysis of the cycle-to-cycle dispersion. The images processing analysis has been carried out in different steps: image acquisition and background subtraction, filtering, fuel spray edges determination, tip penetration and cone angle measurements. Background subtraction and median filter procedures have been adopted during the image acquisition to remove impulse noise and stray light so to maintain sharp the spray edge. This is determined selecting an intensity threshold level for separating the fuel region from the background ambient gas. More details about image post-processing procedure can be found in [17]. Finally, a stainless steel flat plate has been introduced into the vessel to reproduce the spray-wall The growth of the impinging gasoline spray, which is a function of the radial penetration (L) and spray height (H), has been analyzed in order to estimate the development of the spray on the wall. The radial penetration has been defined by the distance between the spray center and the end of spray behavior while the spray height has been defined by the height between the plate bottom and the upper spray behavior [18]. Figure 1 shows the definitions of the spray height (H), radial spray penetration (Lsx and Ldx), inclination angle of the wall (α) and impinging distance (dwall) both for the vertically (left side) and the inclined impinging spray (right side). For the lateral spray penetration (L) the average between the corresponding values on the left (Lsx) and on the right (Ldx) respect the injector axis has been considered. While for the measures of the spray height only the value on the left side after the impact has been considered and only for the condition α=0°. CFD SETUP The analyses have been carried out by means of the Star-CD code, licensed by CD-Adapco [19, 20]; a computational grid reproducing the experimental quiescent chamber has been used for the validation of the CFD models; the cylinder of 40 mm of radius and 90 mm of height has been made up of 150000 hexahedral cells, with a average cell size equal to 1.5 mm [21]. As the CFD simulations are compared in terms of spray impingement on the wall, the computational grid is a cylinder of 50 mm of radius and 22 mm of height. It is made by 70000 hexahedral cells while the mesh has been refined near the impinging region, resulting cells of 1 mm of dimension. A wall boundary condition has been used to simulate the impinging region. Gratis copy for Alessandro Montanaro Copyright 2012 SAE International E-mailing, copying and internet posting are prohibited Downloaded Monday, March 19, 2012 05:04:19 AM Figure 1. Definition of the impinging spray characteristics. The RNG k-ε model has been chosen as a closure model [22, 23] for the simulation of the turbulence. All the analyses have been carried out within the framework of the lagrangian approach [24, 25] to model the fuel spray and a set equal to 170000 computational parcels has been injected. As far as the primary breakup has been considered, the initial droplets have been introduced adopting the modified linear analysis of Kelvin-Helmholtz instability (KH model) [26, 27, 28]: in order to take into account the effect of the cavitation the resulting distribution of the atomized droplets has been modified. As far as the secondary breakup is concerned a tuning of the Rayleigh-Taylor model is carried out [29, 27]. Turbulent phenomena due to the droplet flow interaction have been taken into account by the stochastic approach described in [30]; to correctly define the interaction between the droplets into the toroidal zone of the spray, a collision model [31] has been adopted. Figure 2 shows the various impingement regimes of a droplet-wall interaction. The detailed Bai and Gosman model [32] has been used for the calculation of the impinging droplets. In the stick regime, a droplet with low kinetic energy adheres to the wall in nearly spherical form and continues to evaporate. In the case of spread, the droplet impacts with moderate velocity on a dry or wetted wall, spreads out and mixes with the wall film (wetted wall) or forms a wall film (dry wall). If rebound occurs, the droplet bounces off the wall (reflection) and does not break up. This regime is observed in the case of dry and hot walls, where the contact between drop and wall is prevented by a vapor cushion. Rebound also occurs in the case of a wet wall if the impact energy is low and an air film between drop and liquid film minimizes energy loss. In the boiling-induced break-up regime, the droplet disintegrates due to a rapid liquid boiling on a hot wall. The wall temperature must be near the Nakayama temperature TN, at which a droplet reaches its maximum evaporation rate. In the case of break-up, the droplet deforms into a radial film on the hot surface, which breaks up due to thermo-induced instability. The splash regime occurs at very high impact energy. A crown is formed, jets develop on the periphery of the crown, become unstable and disintegrate into many droplets. Figure 2. Illustration of the impacts regimes Figure 3. Impinging regimes and transition conditions Bai and Gosman [32] have developed a more detailed impingement model considering also the splash regime, as depicted in figure 3 where the transition conditions are reported too. In the case of a dry wall, the stick and spread regimes are combined and called the adhesion regime and the stick regime is neglected in the case of a wetted wall because of its typically very low impact energy. The model varies the transition regimes: Gratis copy for Alessandro Montanaro Copyright 2012 SAE International E-mailing, copying and internet posting are prohibited Downloaded Monday, March 19, 2012 05:04:19 AM RESULTS AND DISCUSSION (1) is the Weber number, which takes into account the ratio of the droplet's kinetic energy (vn is the velocity component normal to the surface, ρl the liquid density, d the droplet diameter, σ surface tension); Different injection strategies have been tested changing the injection duration, pressure and the corresponding total injected mass. All the investigated configurations are collected in Table 2. For each configuration, reported in Table 2, the fuel injection rate has been extracted from the experiments as the average of one hundred of shots. Table 2. Injection strategies. (2) is the Laplace number, which evaluates the relative importance of the surface tension and viscous forces (µl is the dynamic viscosity of the liquid). For a dry wall, the transition criterion between adhesion and splash is given as (3) where the coefficient A depends on the surface roughness; in the case of an already wetted wall the transition Weber numbers for rebound and spread are and for spread and splash are (4) Bai has modeled the post-impingement characteristics introducing experimental correlation and random numbers in the definition of the velocity components for the rebounding droplets and for the radius of the secondary droplets (i.e. in case of splash regime). As the droplets adhere on the wall a liquid film formation could occur. A CFD model developed by Bai and Gosman [33], multi-component evaporation and condensation on the film surface is based on the theory given in [34] and [35]. The model accounts for convective transport of conserved quantities within the film and from/to the gas phase. Figure 4 shows the injection profiles obtained for the investigated injection strategies. The area under each profile is the total amount of the injected fuel per stroke. The profiles depend on the injection pressure and duration. Infact higher is the injection pressure and higher is the stationary phase level while greater is injection duration and longer the corrisponding injection profile. The investigaed strategies represent typical injection conditions for the star-up, idle and maximum power engine operating conditions for a turbocharged GDI engine. It is interesting to note as all the profiles are quite similar in the overlapping parts in terms of rise time and slope. In the early stage of the injection the dynamic of the exiting fuel is independent from the injection pressure and the mass flow rate reaches the peak value after 140 µs for 12 MPa and 5.5 MPa of injection pressure while for the 3.0 MPa after 150 µs. The three investigated cases present an incresed mass flow of 83%, 90% and 130% respectevly for 12, 5.5 and 3.0 MPa, comparing to the mass flow rate during the stationar period of the injection. A subsequent pressure drop occuring in the nozzle generates a strong decrease of the flow rate equal to the −38%, −26% and −35% respect the stationary mass flow for 12, 5.5 and 3.0 MPa. The two highest injection pressure cases begin the stationary period of the injection after 700 µs, while for the 3.0 MPa case it is evident dynamic and unstable effects are dominant for all the injection period. This injection pressure, adopted during the idle condition in current production GDI engine, appears to be border line in terms of fuel delivering and total injected fuel quantity. Gratis copy for Alessandro Montanaro Copyright 2012 SAE International E-mailing, copying and internet posting are prohibited Downloaded Monday, March 19, 2012 05:04:19 AM Figure 4. Fuel injection rate profiles for the different injection conditions. All the numerical simulations adopt the experimental flow rates. In order to validate the CFD models a preliminary analysis has been performed. The spray has been simulated for the three injection pressure in a quiescent chamber at ambient temperature and pressure. The atomization and secondary breakup models have been tuned comparing the calculated spray with the experimental images, as reported in figure 5. The pictures report the fuel spray behavior in the early stages of the injection, the left side images show the experimental results and the right show the calculated results. The numerical and experimental comparison appears to be quite good, the simulations well capture the spray morphology for all of the three injection pressure. As the pressure increases a strong jet to jet interaction is visible in the experimental images and explained in [21]. Due to the pressure drop occurring into the center of the spray, for the case with the two highest injection pressures, the center line jet shows a greater penetration than the neighbors. The CFD images report the computational parcels colored in function of the droplet diameter, since blue is the lowest diameter and red is the highest, it is visible as the injection pressure increases the injected droplets are characterized by lower droplet size. Figure 6 shows the good experimental and numerical comparison, in terms of spray tip penetration. The CFD simulations of the three cases well capture the experimental penetration in the early stages of the injection, while, after 0.5 ms after SOI, the numeric over-predicts the experiment. The gap is more relevant for 5.5 and 12 MPa cases and almost focused on the stationary period of the injection. After 1.2 ms after SOI the 3.0 MPa and 12MPa better match with experimental results. Figure 5. Experimental (left) and numerical comparison (right) of the spray morphology for the three injection pressure. Gratis copy for Alessandro Montanaro Copyright 2012 SAE International E-mailing, copying and internet posting are prohibited Downloaded Monday, March 19, 2012 05:04:19 AM case. From the above statements, CFD simulation predicts secondary droplets with an higher tangential momentum than experiment, leading to an increased L penetration and a shorter H penetration. Figure 6. Experimental and numerical comparison of the tip spray penetration for all the injection pressures. In figure 7 and in figure 8 the influence of the injection pressure on the lateral spray penetration is reported for both inclination angle of the wall (α=0° and α=30°) with the wall distance of 20 mm. For each curve the experimental starting point indicates the impact time of the spray against the wall. The penetration rates just after impingement begin to increase according to increasing the injection pressure. In Figure 8 the effects of the wall inclination angle on the lateral spray penetration can also be seen, it increases with increase of the wall inclination angle (going from α =0° to 30°). This trend indicates that the effect of wall impingement is reduced owing to a reduction in radial droplet momentum as the inclination angle of the wall is increased. A slight overestimation of the radial penetration can be seen in the numerical and experimental comparison, in figure 7 and 8. It is more evident in the simulations of the impinging spray on the flat plate with inclination angle α=0°. In the first part of the penetration rate the CFD simulations, in figure 8, show a good agreement in terms of L penetration, for all of the injection pressure, while after 0.7 ms after SOI, a slight gap with the experimental data appears. The slight over estimation shown in the numerical penetration in figure 6 is a possible explanation. The simulation computes higher normal component velocity of the impinging droplets than experiment, resulting in higher tangential component velocity of the secondary droplets. Furthermore, as the tangential velocity component of the droplets, after the impingement in the splash regime, is function of the tangential velocity component of the impinging droplets and function of a friction coefficient cf [32], a tuning activity has been performed. Literature suggests for water spray a value of 0.7 while in this activity a cf=0.5 has been estimated. Lower values of cf have not been taken into account. As far as the numerical calculations are concerned, reaching the mass flow rate the steady condition, the 5.5 MPa and 12 MPa injection pressure show an under predicted height penetration H of the spray, as visible in figure 9 for the α=0° Figure 7. Radial spray penetration (L) at various injection pressures for α=0° inclination angle of the wall. Figure 8. Radial spray penetration (L) at various injection pressures for α=30° inclination angle of the wall. Figure 9. Height spray penetration (H) at various injection pressures for α=0° inclination angle of the wall. Figures 10 (3.0 MPa), 11 (5.5 MPa) and 12 (12MPa) report the mean Weber Numbers of the spray averaged on a section, for the two inclination angle: red points regard the impinging droplets while the black points regard the droplets after the impact on the wall. Since the spray is hitting the dry wall at the beginning of injection process, the transition regime is Gratis copy for Alessandro Montanaro Copyright 2012 SAE International E-mailing, copying and internet posting are prohibited Downloaded Monday, March 19, 2012 05:04:19 AM regulated by (3). Due to the high pressure peak, previously described in the figure 4, the impinging occurs in the splash regime. Later in the injection period, as the mass flow rate reaches the steady condition, the Weber Number of the droplet decreases, as visible in the figure 10, 11, 12 and spread regime could occur for some droplets, even if the splash is still the main impinging regime as Wedroplet is greater than (4). Figure 12. Weber Number of the droplets before the impact (red) and after the impact (black) for the 12 MPa case. Figure 10. Weber Number of the droplets before the impact (red) and after the impact (black) for the 3.0 MPa case. Figure 11. Weber Number of the droplets before the impact (red) and after the impact (black) for the 5.5 MPa case. An overall qualitative investigation is reported in the last figures, from 13 and 14. Each figure depicts impinging spray from a front view, comparing the experimental images with the simulations for all of the investigated injection pressures. The figures show the effects of the injection pressure on the macroscopic behavior of the fuel spray impinging on the wall. As shown in the figures, the injection pressure hardly affects the spray development process; at the highest injection pressures the development of spray is very high. The spray tips on the impingement wall for the injection pressure of 12.0 MPa are sharper than 5.5 and 3.0 pressures because of the smaller droplet size produced. Higher is the velocity at the point of impingement, due to the high injection pressure, and higher is the radial velocity near the wall and results in a sharper shape of the spray tip. In figure 13 as the injection pressure has increased, the spray height increases. In the case of inclined wall (figure 14), the downward penetration of the spray is very large. It can indicate that the increased angle of wall inclination reduces the effect of impingement on the macroscopic spray characteristics (the momentum loss during the impingement is less). As the inclined angle of the impingement wall has increased, the spray path behavior is enhanced. Gratis copy for Alessandro Montanaro Copyright 2012 SAE International E-mailing, copying and internet posting are prohibited Downloaded Monday, March 19, 2012 05:04:19 AM Figure 13. Experimental (top) and numerical (bottom) comparison, images of impinging spray for all injection pressures and α=0° The simulated spray has been reproduced by means of computational parcels depicted in constant diameter aiming to better show the after impinged droplets behavior. Focusing on the figure 13 the simulations well describe the wake of secondary droplets and despite at the end of the injection CFD could be improved, globally the details are captured. Similar considerations are evident for the 30° of inclination angle (figure 14): vortexes due to the lost radial momentum of the droplets are described by simulations since the end of the injection where some differences appear. As for the 12 MPa injection pressure case which shows for the simulation a middle wake located between the impinging spray and the edge of the spray is missing. CONCLUSIONS An experimental and numerical study have been performed to investigate the effects of injection pressure and wall inclination angle on the macroscopic behavior of a multi-hole GDI spray impinging on the wall. The experimental characterization has been carried out in terms of fuel injection rate and spatial and temporal evolution of the spray impinging on the wall. The spray visualization has been made by injecting the fuel in a high pressure quiescent vessel, optically accessible, containing nitrogen at the pressure of 0.1 MPa and the temperature of 300 K. A stainless steel flat plate, at the temperature of 300 K and distance of 20 mm from injector nozzle, has been introduced into the vessel to reproduce the spray-wall impingement. The growth of the impinging spray has been analyzed in terms of radial penetration (L) and spray height (H). The experimental data together with optical images have been available for the validation by means of the Star-CD code. In order to validate the lagrangian CFD models, preliminary simulations have focused on the spray evolution within a quiescent chamber at ambient temperature and pressure. Since the acceptable experimental and numerical comparisons, both in terms of spray penetration and morphology, the CFD analysis have investigated the spray/ wall interactions. The main results can be summarized as follows: • The fuel injection rate profiles, for the three investigated injection pressures, show a similar behavior in terms of rise time and slope; higher is the injection pressure and higher is the level of the stationary phase. • A complete description of the dynamic behavior of the exiting fuel is described in terms of delay. Gratis copy for Alessandro Montanaro Copyright 2012 SAE International E-mailing, copying and internet posting are prohibited Downloaded Monday, March 19, 2012 05:04:19 AM Figure 14. Experimental (top) and numerical (bottom) comparison, images of impinging spray for all injection pressures at α=30° • The highest injection pressures promote a longer radial penetration and spray height of the impinging spray. • Good agreement between experimental and computed results in terms of spray cone angle, tip penetration and spray morphology for all the three injection pressures before and after the impingement; • Since spray impact is simulated at ambient temperature and at high injection pressure, the splash regime is correctly modeled. A tuning of the CFD sub models has been performed improving the comparison of the radial spray penetration. • Differences are evident in the H penetration of the spray at the end of the injection process leading to the conclusion the CFD simulation slightly over predicts the secondary droplets momentum. This activity aims to analyze the effects of the impinging multi-hole spray in different configurations and conditions and the results reported in the paper are a preliminary step towards the investigation of such a complex set of events. Future works will focus on new impinging configurations (i.e. different injector distances, different wall inclinations) and the effect of the wall temperature will be take into account and investigated. Gratis copy for Alessandro Montanaro Copyright 2012 SAE International E-mailing, copying and internet posting are prohibited Downloaded Monday, March 19, 2012 05:04:19 AM REFERENCES 1. Solomon, A.S., Anderson, R.W., Najt, P.M., Zhao, F. “Direct Fuel Injection for Gasoline Engines”, Society of Automotive Engineers, Inc., Warrendale, PA, 978-0768005363, 1999. 2. Zhao, F., Harrington, D.L., Lai, M-C.D., “Automotive Gasoline Direct-Injection Engines,” Society of Automotive Engineers, Inc., Warrendale, PA, 978-0-7680-0882-1, 2002, doi:10.4271/R-315. 3. Zhao, F., Lai, M., and Harrington, D., “A Review of Mixture Preparation and Combustion Control Strategies for Spark-Ignited Direct-Injection Gasoline Engines,” SAE Technical Paper 970627, 1997, doi: 10.4271/970627. 4. Habchi, C., Foucart, H. and Baritaud, T. “Influence of the Wall Temperature on the Mixture Preparation in DI Gasoline Engines”, Oil & Gas Science and Technology - Rev. IFP, Vol. 54 (1999), No. 2, pp. 211-222. 5. Tsunemoto, H. (1996). “Process of mixture formation of impinging spray on the wall in a hole type nozzle”, Int. J. JSME 27, 2, 36-45. 6. Ishikawa, N. (1998), “Observation of spray impingement behavior in a small DI combustion chamber”, Int. J. JSME 29, 2, 23-28. 7. Park, J., Xie, X., Im, K., Kim, H. et al., “Characteristics of Direct Injection Gasoline Spray Wall Impingement at Elevated Temperature Conditions,” SAE Technical Paper 1999-01-3662, 1999, doi:10.4271/1999-01-3662. 8. Walkins, A.P., Park, K., 1996. “Assessment and application of a new spray wall impaction model” Proc. Inst. Mech. Eng. Conf. Comput. Reciprocating Gas Turbines, 1-10. 9. Bai, C. and Gosman, A., “Development of Methodology for Spray Impingement Simulation,” SAE Technical Paper 950283, 1995, doi: 10.4271/950283. 10. Shim, Y. S., Gyung-Min, C., Kim, D., “Numerical and experimental study on effect of wall geometry on wallimpingement process of hollow-cone fuel spray under various ambient conditions”, International Journal of Multiphase Flow 35, 2009, 885-895. 11. Inamura, T., Tomoda, T. “Characteristics of sprays through a wall impinging injector”, Atomization and sprays, vol. 14, pp. 375-395, 2004. 12. Lee, S. Y., Ryu, S. U. “Recent progress of spray-wall interaction research”, Journal of Mechanical Science and Technology, Vol. 20, N.9, pp. 1101-1117, 2006. 13. Alfuso, S., Allocca, L., Greco, M., Montanaro, A., Valentino, G., “Time- and Space Characterization of Multihole GDI Sprays for IC Engines by Images Processing and PDA Techniques”, Paper ILASS08-071, 2008. 14. Costa, M., Sorge, U., Allocca, L., “Numerical study of the mixture formation process in a four strokes GDI engine for two-wheels applications”, Simulation Modelling Practice and Theory Journal, doi: 10.1016/j.simpat.2010.07.006, 2010. 15. Bosch, W., “The Fuel Rate Indicator: A New Measuring Instrument For Display of the Characteristics of Individual Injection,” SAE Technical Paper 660749, 1966, doi: 10.4271/660749. 16. Wallace, I. “Injection Rate Gauge: Pass Off Information and User Instructions” - Fuel & Engine Management Systems, Graz - December 2002. 17. Alfuso, S., Allocca, L., Caputo, G., Corcione, F. et al., “Experimental Investigation of a Spray from a Multi-jet Common Rail Injection System for Small Engines,” SAE Technical Paper 2005-24-090, 2005, doi: 10.4271/2005-24-090. 18. Lee, C.H., Lee, K.H. “Experimental Study on Macroscopic Spray Characteristics after Impingement in a Slit-Type GDI Injector”, Internal journal of Automotive Technology, Vol. 9, No. 3, pp. 373-380 (2008), doi: 10.1007/ s12239-008-0045-2. 19. Computational Dynamics, “STAR-CD User Guide”, 2010, London (UK). 20. Computational Dynamics, “STAR-CD Methodology”, 2010, London (UK) 21. Malaguti, S., Fontanesi, S., Cantore, G. “Numerical characterization of a new high-pressure multi-hole GDI injector”, ILASS Europe 2010, Brno, Czech Republic, September 2010. 22. Yakhot, V., Orszag, S.A., Thangam, S., Gatski, T.B., and Speziale, C.G. “Development of turbulence models for shear flows by a double expansion technique”, Phys. Fluids, 1992. 23. Cantore, G., Fontanesi, S., Mattarelli, E., and Bianchi, G., “A Methodology for In-Cylinder Flow Field Evaluation in a Low Stroke-to-Bore SI Engine,” SAE Technical Paper 2002-01-1119, 2002, doi:10.4271/2002-01-1119. 24. Bracco, F., “Modeling of Engine Sprays,” SAE Technical Paper 850394, 1985, doi: 10.4271/850394. 25. Dombrowski, N., Johns, W. R. “The aerodynamic instability and disintegration of viscous liquid sheets”, Chem. Eng. Sci., v. 18, p. 203, 1963. 26. Patterson, M. and Reitz, R., “Modeling the Effects of Fuel Spray Characteristics on Diesel Engine Combustion and Emission,” SAE Technical Paper 980131, 1998, doi: 10.4271/980131. 27. Beal, J. C., Reitz, R. D., “Modeling spary atomizaion with the Kelvin-Helmholtz/Rauleigh-Taylor Hybrid modela”, Atomization and Sprays, vol.9, pp 623-650, 1999 28. Reitz, D. “Modeling Atomization Processes in HighPressure Vaporizing Sprays”, Atomization and Spray Technology 3, 1987, pp309-337. Gratis copy for Alessandro Montanaro Copyright 2012 SAE International E-mailing, copying and internet posting are prohibited Downloaded Monday, March 19, 2012 05:04:19 AM 29. Taylor, G.I., “The instability of liquid surfaces when accelerated in a direction perpendicular to their planes”, In Bachelor, GK, The Scientific Papers of Sir Geoffery Ingram Taylor, 1963, vol. 3, pp 532-536, Cambridge University Press. 30. Gosman, A.D., and Ioannides, S.I. 1983. “Aspects of computer simulation of liquid-fuelled combustors”, AIAA, J. Energy, 7(6), pp. 482-490. 31. O'Rourke, P.J. “Collective Drop Effects on Vaporizing Liquid Sprays”. PhD Thesis, University of Princeton, 1981. 32. Bai, C. and Gosman, A., “Development of Methodology for Spray Impingement Simulation,” SAE Technical Paper 950283, 1995, doi: 10.4271/950283. 33. Bai, C. and Gosman, A., “Mathematical Modelling of Wall Films Formed by Impinging Sprays,” SAE Technical Paper 960626, 1996, doi: 10.4271/960626. 34. Sirignano, W.A. 1999. “Fluid Dynamics and Transport of Droplets and Sprays”, Cambridge University Press, New York. CONTACT INFORMATION Dr. Alessandro Montanaro Istituto Motori CNR Via G. Marconi 8, 80125 Napoli, Italy a.montanaro@im.cnr.it Dr. Simone Malaguti Department of Mechanical and Civil Engineering University of Modena and Reggio Emilia Via Vignolese 905, 41100 Modena, Italy Ph.:+39 059 2056114 simone.malaguti@unimore.it ACKNOWLEDGMENTS The authors wish to acknowledge CD-Adapco Group, for the use of the Star-CD code, granted to the University of Modena and Reggio Emilia. 35. Torres, D.J., O'Rourke, P.J., and Amsden, A.A. 2003. ‘Efficient multicomponent fuel algorithm’, Combust. Theory Modelling, 7, p. 67. 36. Ruge, J.W., and Stüben, K. 1986. ‘Algebraic Multigrid (AMG)’ in “Multigrid Methods” (Ed. McCormick, S.), Frontiers in applied Mathematics, SIAM, 5, Philadelphia. 37. Versteeg, H. K., Malalasekera, W., “An introduction to computational fluid dynamics. The finite volume method”, Longman, 1995. The Engineering Meetings Board has approved this paper for publication. It has successfully completed SAE's peer review process under the supervision of the session organizer. This process requires a minimum of three (3) reviews by industry experts. All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photocopying, recording, or otherwise, without the prior written permission of SAE. ISSN 0148-7191 Positions and opinions advanced in this paper are those of the author(s) and not necessarily those of SAE. The author is solely responsible for the content of the paper. SAE Customer Service: Tel: 877-606-7323 (inside USA and Canada) Tel: 724-776-4970 (outside USA) Fax: 724-776-0790 Email: CustomerService@sae.org SAE Web Address: http://www.sae.org Printed in USA