2010-01-2123
Numerical analysis of GDI engine cold-start at low ambient temperatures
Simone Malaguti, Stefano Fontanesi, Elena Severi
University of Modena and Reggio Emilia
Copyright © 2010 SAE International
ABSTRACT
The paper investigates the low-temperature cranking operation of a current production automotive Gasoline Direct Injected (GDI) by means of 3D-CFD simulations. Particular care is devoted to the analysis of the hollow cone spray evolution within the combustion chamber and to the formation of fuel film deposits on the combustion chamber walls. Due to the high injected fuel amount and the strongly reduced fuel vaporization, wall wetting is a critical issue and plays a fundamental role on both the early combustion stages and the amount of unburnt hydrocarbons formation. In fact, it is commonly recognized that most of the unburnt hydrocarbon emissions from 4-stroke gasoline engines occur during cold start operations, when fuel film in the cylinder vaporize slowly and fuel can persist until the exhaust stroke.
In view of the non-conventional engine operating conditions (in terms of injected fuel amount, engine speed, ambient and wall temperature and almost null fuel atomization and breakup), an understanding of the many involved phenomena by means of an optically accessible engine would be of crucial importance. Nevertheless, the application of such technique appears to be almost unfeasible even in research laboratories, mainly because of the relevant wall wetting.
CFD analyses prove then to be a very useful tool to gain a full insight of the overall process as well as to correlate fuel deposits to both the combustion chamber design and the injection strategy. In order to better understand where, and how thick, these wall films are formed during the intake and compression, a detailed description of the spray interaction with both the piston wall and the intake valves was performed by the authors in a previous paper [1]. Subsequently, a wide set of injection strategies was simulated in order to better understand the physics of spray/wall interaction and to minimize the formation of deposits in the combustion chamber most critical locations [2].
In order to limit the overall number of modeling uncertainties (spray evolution, droplet-droplet interaction, droplet-wall interaction, liquid-film) the spray model was at first validated against experimental data under low injection pressure, and results from the comparison were reported in [1].
In the present paper, cold start operations at decreasing ambient temperatures are modeled and results are analyzed in terms of both fuel film distribution on the combustion chamber walls and resulting fuel/air mixture distribution within the combustion chamber. The use of CFD simulations prove to be useful to investigate and understand the influence of both combustion chamber design and injection profile on the amount and distribution of fuel deposits, showing a high potential to address future engine optimization.
INTRODUCTION
Gasoline direct injected spark ignition engines are nowadays becoming a standard, in view of the increase in engine control flexibility, fuel conversion efficiency and subsequent reduction of both fuel consumption and engine-out emissions, especially at part load. Due to the need for both lower size/power ratios and higher fuel conversion efficiencies, in fact, the advantages of GDI versus conventional PFI engines are numerous [3, 4].
A great effort is therefore paid by engine manufacturers in order to gain a deep knowledge of the impact of direct injection on both engine design and management. Particularly, research fields of crucial importance are fuel spray morphology [5] and its interaction with the fresh charge and the combustion chamber walls [6], injector location with reference to the spark plug and injection strategies optimizing the combustion system over a wide range of operating condition.
In order to face the many challenges above, on the injection system side, manufacturers developed a wide variety of different technologies, ranging from hollow cone single hole injectors through multi hole Diesel like injectors up to outward pintle injectors [4]. As a general trend, injection pressure is constantly raising, as well as new complex multi-injection strategies are under development in order to properly match the target performance of the combustion system to the newly developed injection systems. Simultaneously, on the engine side, manufacturers are paying a huge research effort to gain a better understanding of potentials, constraints and limitations of these technologies as well as their impact on engine design (port design, piston design), engine management, turbocharging, variable valve lift, etc.
Such a combined effort is needed in order to properly take advantage of the GDI benefits, while limiting the drawbacks of the in-cylinder injection, i.e. higher dependency of the mixture quality on the in-cylinder flow field, as well as higher interaction between the fuel and the combustion chamber walls.
While in-cylinder flow pattern optimization is usually demanded to the design of the intake ports, both mixture preparation and fuel/wall interaction involve a wider set of factors, among which in-cylinder flow motion, piston design, injector location, injection pressure, timing and strategy, spray characteristics, etc.
Due to the broad range of engine operations, which span from low temperature/cranking conditions at engine start-up to high temperature/high speed full load operations, the optimization task is characterized by an extreme complexity.
Exhaustive experimental understanding of such a complex set of mutually interacting phenomena is far from being feasible from an industrial point of view, and is an open challenge also in research laboratories. Conversely, CFD simulations, in view of the ability to give a fast and complete set of global and local information [7, 8], appear to have a great potential in proficiently supporting the optimization process, where a strong reduction in both development costs and time is expected. Different combustion systems, injectors and injection strategies can be virtually prototyped, for example, in order to optimize fuel/air mixing, as well as to get optimum mixture composition at the spark plug just before ignition. The advantage becomes even more evident if CFD is used to reproduce engine operations where experimental data are not available. A typical example is the very first engine cycles at engine start-up, where only limited information can be provided by the ECU.
Despite the increase of both computational power and CFD software ease of use, nevertheless, both the accuracy and the predictive capability of the numerical forecasts of in-cylinder GDI engine processes are still critical issues, because of the many complex physical phenomena involved [9, 10]. In fact, CFD analyses must accurately reproduce fuel spray evolution, vaporization and mixing with the surrounding air, as well as the interaction between the fuel droplets and the combustion chamber walls.
A CFD investigation on the fuel mixing formation within the combustion chamber of an automotive GDI engine under low-temperature cranking operations, is reported in the present paper. The activity is focused both on the effects of the decreasing ambient temperature and the injection strategies, stressing the attention on the fuel deposits on the combustion chamber walls and mixture formation.
Although many CFD models are available in literature to represent each of the above described processes, none of them is widely recognized as a standard, and a preliminary tuning process is required for most of them. Therefore, experimental evidence is still a mandatory requirement to assess the quality and accuracy of CFD forecasts. A validation of the hollow cone spray generated by the pressure-swirl injector was previously performed and is reported in [1]. The resulting adopted lagrangian model was preliminarily validated against experimental evidence injecting the fuel in a test vessel under quiescent conditions and ambient temperature and pressure. Spray visualization by means of stroboscopic and Schlieren techniques were used to assess the accuracy of the CFD setup. Concerning fuel deposits on the combustion chamber the adopted film model is now validated against experimental data and results are reported in the following sections.
CFD analyses of the low-temperature cranking operation at 273K were presented in previous papers by the authors [1]. Results from the CFD simulations showed that a strong interaction between the injection process and the valve opening and closing events was the primary reason for the formation of fuel deposits at the spark plug, as highlighted in Figure 1 below, where both CFD forecasts and experimental evidence are reported. In order to limit the amount of fuel deposits, different injection strategies were investigated [2]. An optimized injection strategy was then proposed and simulated. The strategy proved to be able to avoid misfire during low-temperature engine start-up operation and led to reasonable fuel vaporization.
Figure 1 CFD vs Experimental: fuel deposits in the spark plug region.
THE engine
Grid Generation
The investigated engine is a four cylinder, four valves per cylinder gasoline direct injected unit. Some key engine features are reported in Table 1 below.
Table 1: engine characteristics
The combustion chamber has a wide spacing architecture with a central mounted spark plug and the injector in-between the intake valves. A wall guided in-piston bowl solution, visible in Figure 2 below, is chosen in order to obtain the desired charge stratification.
Figure 2: engine piston top surface
The injection system is based on a swirl inwardly opened nozzle injector, with a 0.56 mm hole diameter and a 60° nominal cone, while the injection pressure range varies between 0.5 and 15 MPa according to the engine operations.
ENGINE OPERATING CONDITIONS - As stated earlier, the paper investigates the cold start operation of the engine, and all the CFD analyses focus on the second engine cycle, i.e. the one conventionally preceding first engine ignition. Simulations are carried out reproducing low- temperature engine startup, in order to mimic a severe test carried out to evaluate engine ignition under extreme environmental conditions.
The main operation characteristics are listed below:
• The engine speed is very low, slowly increasing from zero to minimum engine speed; in the following calculations, temperature-dependent rpms are chosen to mimic the second engine cycle for each tested condition;
• All the engine components, including the combustion chamber walls and the ports, are at ambient temperature; CFD simulations are reported for temperatures ranging from 273 K down to 259.5 K;
• The injection system high-pressure line is deactivated, so that fuel pressure is provided uniquely by the low-pressure pump; a value of 5.5 bar, suggested by the engine manufacturer, is used in the following sections;
• In view of the reduced fuel vaporization, the injected fuel quantity is very high; the injection process is therefore very prolonged; the injection timing is advanced, covering the whole intake stroke.
As a consequence of the above statements, the investigated conditions clearly appear very far from standard engine operations.
Table 2: tested engine conditions
Some key factors for the tested conditions are reported in Table 2: particularly, engine speed and injection duration for all the cylinders are highlighted. Data are derived from confidential ECU datasheets provided by the engine manufacturer: as soon as stable ignition and combustion are achieved, i.e. between the second and the fourth engine cycle, engine speed rapidly increases up to idle conditions. As stated earlier, the paper focuses on the engine cycle preceding first engine ignition, in order to evaluate the spray interaction with the combustion chamber walls and the subsequent fuel deposit formation.
Concerning injection conditions, instantaneous mass flow rate is derived from a combination of the drive wave form of the injector and literature investigations [5, 11], in order to correctly define the spray evolution during the early stage of the injection. An injector opening delay of 0.2 ms, as well as opening and closure ramp durations, is estimated from injector characterization under standard engine operations carried out by the engine manufacturer.
CFD SETUP
Computational domain
All the analyses are carried out by means of the Star-CD code, licensed by CD-Adapco [12, 13]; moving grids covering the full intake and compression processes are based on a combination of mesh distortion, cell-layer activation and deactivation and are generated by means of the es-ice tool, again by CD-Adapco. The overall computational grid is made up of cells of mainly hexahedral shape, whose number ranges from approximately 86.000 at FTDC up to 620.000 at BDC. Intake and exhaust ports are deactivated during the closed valve portion of the cycle to limit the computational cost of the simulations. In order to improve the investigation of the spray impingement and subsequent fuel film formation phenomena, even the smallest geometrical details of both the cylinder head and the piston are taken into account. Figure 3 shows the whole computational domain, and zoomed views of both the injector site and the spark plug electrodes.
Flow model
CFD setup for the in-cylinder flow analysis is the results of the authors' experience: RNG k-ε model is chosen as a closure model [14, 15], heat transfer through the cylinder walls is accounted for by means of setting fixed temperatures on the different wall regions. In order to prevent possible computational instabilities arising at the intake valve opening, all the simulations start 5 CA degrees before IVO. Since the investigated intake stroke falls at the very beginning of the second engine cycle, the whole intake system is considered to be filled with almost still air at ambient conditions, i.e. 1 bar and temperature ranging from 273K to 259.5K. In fact, although the engine is throttled during the cranking operation, pressure losses in the intake system are expected to be negligible in view of the very low engine speed.
Figure 3 Computational domain and detail of the injector location and spark plug location
Spray model
All the analyses are carried out within the framework of the lagrangian approach to the modeling of fuel sprays.
A set of computational parcels is injected, based on a statistical approximation. Thus, the total population is represented by a finite number of computational parcels (samples), each of which represents a cluster of elements having the same properties [16]. Several sub-models describe the spray formation and evolution, as well as its transition to fuel vapor.
A specific subroutine was implemented by the authors in order to introduce the computational parcels in the computational domain, as well as to define a proper set of initial droplets conditions [2, 11].
As far as the secondary breakup is concerned, Reitz model is used, where a tuning of some model constants was previously carried out in order to correctly account for the effects of Bag and Strip breakup [11, 17, 18]. Turbulent phenomena due to the droplet flow interaction are taken into account by the stochastic approach described in [19]; in order to correctly define the interaction between the droplets into the toroidal zone of the spray, a collision model [20] is adopted together with a condensation model.
Liquid film model
In order to minimize fuel deposits, it is important to find correlations between film distribution, injection timing, combustion chamber design and engine operating conditions, aiming at finding a trade-off between the need for very rich mixtures and reduced wall film deposits at engine start up.
In order to properly model the fuel film formation, two key factors must be accurately analyzed and taken into account: the former is the correct estimation of the impinging droplet regime transition criteria, the latter is the accurate representation of the post-impingement characteristics: the velocity magnitude and direction of the rebound, the fraction of the mass deposit on the wall, the size and velocity distributions of the secondary droplets.
Droplet/wall interaction is modeled by means of the Bai-Gosman model described in [21, 22], which considers four impingements regimes: stick, rebound, spread and splash. The existence of these regimes, as well as the transition between two different regimes, depends on the properties of the impinging droplets and target surface, i.e. dry or wetted.
Concerning post-impingement, two different liquid-film models are currently available in the Star-CD software, a so-called stationary model and a so-called dynamic one. The former neglects both convection within the film and momentum exchange with the continuous phase. However, mass and heat transfer to the surrounding continuum are allowed. The latter accounts for all transfer processes with the continuous phase. In view of the higher accuracy, the Star-CD built-in dynamic model is used in the present paper.
CFD Results
LIQUID FILM MODEL VALIDATION
Of the many methods which have been used to measure wall film thickness, laser induced fluorescence (LIF) has been one of the most successful in i.c. engines. The film is illuminated with a laser and a fluorescent compound in the fuel emits red-shifted light with an intensity which depends on the film depth. LIF offers a robust signal in the dirty optical environment inside a piston engine. Several different optical configurations have been used for the analysis of fuel films formed in intake manifolds and cylinders of spark ignition (SI) engines: using fiber optics to carry the signal and/or exciting light [23-25], to measure a fuel film on a transparent surface by total internal reflection [26-29] or without total internal reflection [30-32], by illumination and detection through a window to measure the fuel film on the metal splitter separating the two branches in a siamesed port [33], and to visualize fuel on the inner metal surfaces of the cylinder, though without quantitative measurement of the thickness [34, 35]. A sketch of the experimental facility is shown in Figure 4 below.
In the present paper, experiments carried out by Le Coz and co-workers [30], also referred in [36, 37], based on the use of a pulsed laser, diffuse illumination and a camera, are used for the preliminary tuning of both the spray impingement and liquid film models. The experiments were carried out in a transparent pipe with constant air flow rate and pulsed injections of iso-octane with a supply pressure of 3 bar, i.e. very close to the startup injection conditions for the considered GDI engine.
Figure 4 sketch of the fuel film experimental facility
The tuning process focuses both on the post-impingement stripping model, which is used to correctly capture the shear effect of the air flow on the fuel film and on the liquid film sliding model by Foucart [13]. According to this model, the fuel adhering on a wall is treated as an hemispherical fuel droplet with radius R, whose value is evaluated at each time step according to the fuel deposit volume. The radius R is then compared to a critical value Rc, which governs fuel film transition behavior as follows:
if R < cRc the film adheres to the wall;
otherwise, the film continues to move;
where c is a constant governing the transition between "stationary" and "moving" film regimes, usually ranging from 0.0 to 1.0.
The effects of both grid spacing and near-wall grid size were preliminarily investigated, showing only limited effects on the obtained results. Several CFD calculations were performed in order to properly match the measurements in terms of both global film distribution and point-wise film thickness. Figure 5 shows some results from the CFD calibration activity. The figure reports the fuel film thickness at different radial distances from the spray impact region after quasi-steady film behavior is achieved. As visible from the picture, a good accuracy in the film thickness spatial distribution is achieved for case "45-mesh2-Setup3", where the constant c is set to 0.9.
Figure 5: results from the liquid film validation process
ENGINE STARTUP SIMULATIONS
The CFD setup for case "45-mesh2-Setup3" is then adopted for the subsequent engine simulations, focusing on both fuel deposits formation and subsequent fuel evaporation and mixture preparation for decreasing ambient temperatures.
Figure 6 shows fuel deposits on the combustion chamber walls for the four different ambient conditions. The analysis of the picture clearly highlights the following issues:
decreasing the ambient/wall temperature, fuel deposits increase mainly in view of the increasing injected quantity, while the fuel film mass trends versus crank angle are similar;
fuel deposits for 268K and 263K cases are almost comparable in view of the reduced difference between the two injected quantities;
fuel deposits undergoing evaporation are almost negligible, and most of the film on the combustion chamber walls persists up to the last stages of the compression stroke. The temperature reaches at the end of the compression stroke by the mixture is not sufficient for an effective evaporation of the fuel deposits.
Figure 6: fuel deposits evolution.
At the end of the compression stroke, for temperatures ranging from 273K down to 263K, the mass of fuel film undergoing evaporation reaches a value of approximately 10 mg for all the tested cases, despite the injected fuel quantity relevantly increases while decreasing ambient temperature. Most of the additional injected fuel sticks on the combustion chamber walls contributing to the increase of film mass (+43% moving from 273K to 263K).
A further decrease of the startup engine temperature, i.e. 259.5K case, leads to a 15% reduction of the vaporized fuel, despite the overall injected quantity is almost 100% higher than that for case 273K. In this extremely critical case, fuel deposits increase by almost 80%, in view of both the slower fuel evaporation and the advanced start of injection, which promotes both fuel/intake valve interaction and fuel backflow within the intake ports.
As stated earlier, previous CFD calculations showed that the advanced SOI was the major responsible for the fuel film formation at the spark plug, in view of a relevant impact of the injected fuel against the intake valves and the subsequent spray deviation towards the centre of the engine head.
Figures 7 and 8 report the overall film mass on the engine head and on the spark plug respectively, for all the investigated conditions and focusing at 20 crank angle degrees before firing top dead center (BFTDC). A constantly increasing trend can be observed for both the deposits, partially explaining the increased startup failures due to misfire which were experimentally detected during ad-hoc cold-start tests carried out on production cars equipped with the analyzed engine [1, 2].
A further investigation is carried out in order to analyze the influence of both temperature and injection strategy on the mixture preparation in terms of relative air to fuel ratio: two different values are considered, the former being the average within the whole combustion chamber and the latter being the local average value at the spark plug. Figures 9 and 10 show the global and local values 20 crank angle degrees BFTDC respectively: all the tested conditions show an almost constant value of 2, which is relevantly lean and might be critical in terms of ignition and flame propagation sustainability. Nevertheless, considering that no spark energy release is introduced in the CFD calculations and that there is a relevant fuel fraction which backflows in the intake ports, ignition conditions are expected to be met at the subsequent third engine cycle.
Figure 7:deposits on the engine head. Figure 8: deposits on the spark plug.
Figure 9: global within combustion chamber. Figure 10: spark plug region
In view of the difficulty of meeting a proper level of fuel evaporation and subsequent mixing, physical factors affecting the fuel phase transition are then deeply investigated. Fuel vapor can be originated either by the evaporation of fuel droplets floating within the combustion chamber or by the evaporation of fuel deposits at the combustion chamber walls. CFD simulations show that:
approximately 10 CA degrees after the end of injection (AEOI) fuel deposits reach a constant value, while liquid fuel keeps floating within the combustion chamber;
fuel vapor reaches a maximum value around 150 CA degrees BFTDC, then slightly decreases until IVC;
as visible from Figure 12 below, where results are reported for the 273K case, the relevant backflow towards the intake ports, which depletes the combustion chamber, causes the simultaneous decrease of both air and fuel vapor;
fuel vapor backflow is much lower than air backflow, thus explaining the continuous decrease of relative air to fuel ratio visible in Figure 11 below; a possible explanation is that the fuel vapor mostly originates from the fuel film at the walls, and it is characterized by very reduced momentum, which prevents the vapor to escape towards the intake ports;
temperature increase during the compression stroke plays no effect on fuel evaporation, since fuel vapor reaches a maximum well before IVC and remains almost constant up to FTDC.
Figure 11: in-cylinder relative air/fuel ratio
Figure 12: in-cylinder fuel vapor and trapped air
In order to confirm the above statements, and mainly statements number 4 and 5, Figures 13 and 14 show the separate contributions to the fuel evaporation from film deposits and dispersed fuel droplets respectively for three different crank angles and for all the tested conditions. Fuel vapor mass from the deposits is computed as the difference between the instantaneous film mass on the whole combustion chamber and the peak value (approximately 80 CA degrees AEOI). Fuel vapor from the fuel droplets is derived from a mass balance considering the dispersed phase which is trapped within the combustion chamber and the contribution of the droplets to the formation of fuel deposits.
The following issues emerge:
when fuel vapor reaches its maximum, i.e. approximately at 570 CA degrees, fuel vapor from the deposits has already reached a value of approximately 4 mg, except for the lowest temperature;
before IVC, an additional contribution is given by further evaporation of the fuel film for all the tested conditions;
the above process persists up to the last stage of the compression stroke for all the cases;
concerning fuel vapor originating from the floating droplets, a relevant initial contribution can be observed in Figure 13, where the injected fuel droplets are still characterized by high momentum and droplet crash against the combustion chamber walls enhances droplet breakup;
temperature increase during the compression stroke is not sufficient to promote droplet evaporation; on the contrary, the low temperature levels cause a non negligible part of the injected fuel to undergo condensation, thus explaining the negative contributions to fuel vapor formation visible in Figure 13.
Figure 13: evaporated fuel from deposits. Figure 14: evaporated fuel from fuel droplets
Conclusion
CFD simulations aiming at deeply investigating fuel film formation and mixture preparation for cold-start cranking operations were performed on a GDI engine of current production. The particular engine operations are characterized by very reduced engine speeds, massive injections of fuel covering a relevant portion of the intake stroke, and reduced injection pressure and subsequent fuel atomization and breakup.
In view of the peculiar engine operations, limited data are available from the ECU, and experimental investigations are extremely difficult even in research laboratories, mainly because of the relevant injected fuel quantities. CFD proves therefore to be a very powerful tool to gain a better understanding of the overall process and to investigate the mutual interaction between injection, combustion chamber design, fuel deposit formation and subsequent mixture preparation.
Four different cranking conditions were analyzed, reproducing the second engine cycle, conventionally referred to as the one preceding the first engine ignition. Data from the engine manufacturer were applied concerning fuel injection timing, fuel supply pressure and engine revving speed. The CFD analyses focused both on fuel film distribution through the combustion chamber walls and on fuel vaporization and subsequent mixture preparation just before first engine ignition.
As expected, injected fuel increase while decreasing ambient temperature, in order to reach almost the same value at the end of the compression stroke for all the tested conditions. In view of the increasing injected fuel, deposits on the chamber walls increase. Fuel film wetting the combustion chamber walls proves to be the major responsible for fuel evaporation, while contribution from the floating fuel droplets is relevant and positive only at the initial stages of the compression, where fuel impingement enhances droplet breakup and subsequent phase change. In fact, due to the combination of high fuel quantities and very low in-cylinder temperatures, droplet condensation appears to overcome droplet evaporation, thus reducing the overall fuel vapor formation.
Results from the CFD analyses proved to be fundamental to gain a deep understanding of the origin of spark-plug fuel deposits and engine misfires, as well to suggest some relevant improvements in the injection management at engine startup, among which the activation of the injection system high-pressure line during cold-start cranking operations and the correct injection phasing with reference to the intake valve timing.
The CFD activity is still going on in order to extend the simulations to the subsequent engine cycle and to include both ignition and combustion in the analyses, in order to assess combustion occurrence and combustion stability.
REFERENCES
Fontanesi, S., Malaguti, S., Cantore, G., Lupi, R., Rosetti, A, "CFD investigation of wall wetting in a GDI engine under low temperature cranking operations", SAE2009 Congress, 2009-01-0704.
Malaguti S., Fontanesi S., " CFD investigation of fuel film formation within a GDI engine under cold start cranking operation ", 2009, Proceedings of the ASME, ICES2009-76055.
Solomon, A. S., Anderson, R. W., Najt, P. M., Zhao, F. 2000. "Direct fuel injection for gasoline engines", SAE.
Zhao, F., Harrington, D. L., Lai, M. 2002. "Automotive Gasoline Direct-Injection Engine", SAE.
Valentino G., Auriemma M., Caputo G., Corcione F.E., "Near Nozzle PIV Measurements and Droplets Size Pattern of a GDI Swirled Spray", COMODIA 2004, The Sixth International Symposium on Diagnostics and Modelling of Combustion in Internal Combustion Engines, August 2-5, 2004, Yokohama, Japan.
M.C. Drake, D.C. Haworth, "Advanced gasoline engine development using optical diagnostic and numerical modelling" Proceeding of the Combustion Institute, 2007.
Edward S. Suh and C. J. Rutland "Numerical Study of Fuel/Air Mixture Preparation in a GDI engine", SAE paper, 1999-01-3657.
L. Fan, R. D. Reitz, "Spray and combustion modeling in gasoline direct-injection engines", 2000, Atomization and Sprays, vol.10, pp. 219-249, 2000.
P.K. Senecal, D.P. Schmidt, I. Nouar, C.J. Rutland, R.D.Reitz, M.L. Corradini, "Modelling High-speed Viscous Liquid Sheet Atomization", International Journal of Multiphase Flow, No.25, pp.1073-1097, 1999.
Jordan A. Snyder, Ronald 0. Grover, Volker Sick, Dennis N. Assanis, 2004, "Transient Spray Cone Angles in Pressure-Swirl Injector Sprays", SAE Paper 042939.
S. Fontanesi, V. Gagliardi, S. Malaguti, G.Valentino, M. Auriemma "Detailed experimental and numerical investigation of the spray structure in a GDI high-pressure swirl injector", ICLASS paper,Kyoto,2006.
Computational Dynamics, "STAR-CD User Guide", 2009, London (UK).
Computational Dynamics, "STAR-CD Methodology", 2010, London (UK).
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.
Cantore, G., Fontanesi, S., Mattarelli, E., and Bianchi, G.M., "A methodology for in-cylinder flow field evaluation in a low stroke-to-bore SI engine", SAE paper 2002-01-1119, 2002
D. P. Schmidt, I. Nouar, P. K. Senecal, C. J. Rutland, J. K. Martin, R. D. Reitz, J. A. Hoffman, "Pressure-Swirl Atomization in the Near Field", SAE Paper,1999-01-0496.
N. Dombrowski, W. R. Johns, "The aerodynamic instability and disintegration of viscous liquid sheets", Chem. Eng. Sci., v. 18, p. 203, 1963
Reitz, R.D., and Diwakar, R. 1986. "Effect of drop breakup on fuel sprays", SAE Technical Paper Series 860469.
Gosman, A.D., and Ioannides, S.I. 1983. "Aspects of computer simulation of liquid-fuelled combustors", AIAA, J. Energy, 7(6), pp. 482–490.
O’Rourke, P.J. "Collective Drop Effects on Vaporising Liquid Sprays". PhD Thesis, University of Princeton, 1981.
Bai, C., and Gosman, A.D. 1995. "Development of methodology for spray impingement simulation", SAE Technical Paper Series 950283.
Bai, C., and Gosman, A.D. 1996. "Mathematical modelling of wall films formed by impinging sprays", SAE Technical Paper Series 960626.
T. Johnen, and M. Haug. 1995. "Spray Formation Observation and Fuel Film Development Measurements in the Intake of a Spark Ignition Engine", SAE Technical Paper Series 950511.
Chaves, H.; Hentschel, W.; Obermeier, F.; Stasicki, B. 1996. "In Cylinder High Speed and Stroboscopic Video Obervation of Spray Development in a DI Diesel Engine". SAE Technical Paper Series 961206.
P. G. Felton, D. C. Kyritsis, S. K. Fulcher. 1995. "LIF Visualization of Liquid Fuel in the Intake Manifold During Cold Start". SAE Technical Paper Series 952464.
Heavens OS, Gingell D. 1991. "Film thickness measurement by frustrated total reflection fluorescence". Optics and laser technology 23 3.
Evers L. W. and Jackson K. J. 1995. "Liquid film thickness measurements by means of internally reflected light". SAE Technical Paper Series 950002
Coste T. L. and Evers L. W. 1997. "An optical sensor for measuring fuel film dynamics of a port-injected engine", SAE Technical Paper Series 970869.
Kull E., Wiltafsky G., Stolz W., Min K. D. and Holder E. 1997. "Two dimensional visualization of liquid layers on transparent walls", Opt. Lett., 22 9, pages 645-647.
Le Coz, J.F., Catalano, C. and Baritaud, T. 1994. "Application of Laser Induced Fluorescence for Measuring the Thickness of Liquid Films on Transparent Wall", 7th Int. Symp. onApplication of Laser Techniques to Fluid Mechanics, Lisbon, 1994.
Cho H. and Min K. 2003. "Measurement of liquid fuel film distribution on the cylinder liner of a spark ignition engine using the laser-induced fluorescence technique", Meas. Sci. Tech., 14 7 pages 975-982
Senda J., Ohnishi M., Takahashi T., Fujimoto H., Utsunomiya A. and Wakatabe M. 1999. “Measurement and modelling on wall wetted fuel film profile and mixture preparation in intake port of SI engine” SAE Technical Paper Series 1999-01-0798.
Almkvist G., Denbratt I., Josefsson G. and Magnusson I. 1995. "Measurement of fuel film thickness in the inlet port of an SI engine by laser induced fluorescence". SAE Technical Paper Series 952483.
Witze, P. O. and Green R. M. 1996. "LIF visualization of liquid fuel in the cylinder of a spark ignition engine" Proc. 8th Int. Symp on Applications of Laser Techniques to Fluid Mechanics, Lisbon, Portugal.
Witze, P. O. and Green R. M. 1997. "LIF and Flame emission imaging of liquid fuel films and pool fires in an SI engine during a simulated cold start", SAE Technical Paper Series 970866.
G. S. Zhu, R. D. Reitz, "Modelling characteristics of gasoline wall films in the intake port of port fuel injection engines", Int. J Engine research, Vol. 2, 2001.
C. Habchi, H. Foucart and T. Baritaud, "Influence of the Wall Temperature on the Mixture Preparation in DI Gasoline Engines" Oil & Gas Science and Technology – Rev. IFP, Vol. 54, No. 2, pp. 211-222, 1999.
CONTACTS
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
e-mail: simone.malaguti@unimore.it
Dr. Stefano Fontanesi
Department of Mechanical and Civil Engineering
University of Modena and Reggio Emilia
Via Vignolese 905, 41100 Modena, Italy
Ph.: +39 (0)59 2056114
Fax: +39 (0)59 2056126
e-mail: stefano.fontanesi@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, and for the support.
This work was carried out as part of the “Sviluppo di metodologie di base per l'applicazione di tecnologie innovative per riduzione consumi ed aumento prestazioni specifiche per motori ad alte prestazioni” project under the FIRB program ‘Potenziamento e sviluppo dell'industria motoristica incluse le due ruote con motori a basso consumo e a basso impatto ambientale’ by MIUR, code: RBIP06W2MA.
Page 16 of 16