1. INTRODUCTION

The injection of liquid fuel and water directly into the combustion chamber of internal combustion (IC) engines or technical combustors is a promising way to improve the combustion behavior and efficiency, and to reduce the combustion and exhaust gas temperatures and pollutant emissions [1,2]. So far, the fuel and water are often injected separately while the injection of a mixture using the same injection system would be favorable. As fuel and water do not form a stable mixture, its preparation and transport to the nozzle are challenging [3]. Generally, the injection system should be suitable for the injection of mixtures at high engine loads and fuel-only injections at lower engine loads. Furthermore, it has to be ensured that the water and fuel are still well mixed and homogeneously distributed in the combustion chamber. Under this prerequisite, maximum cooling effects and—with an optimal fuel/air mixture—a safe ignition and combustion can be achieved. Consequently, the control of the fuel/water mixture quality during injection is of high importance for optimization of the combustion concept. Furthermore, a separation of the fluids during injection must be avoided, as it also affects the jet and droplet breakup of both liquids due to their different fluid properties.

Many experimental and numerical studies already exist in the literature regarding the combustion and soot formation of fuel/water emulsions for IC engine and gas turbine applications [4–16]. These studies are mainly focused on diesel or kerosene fuels containing water, but little work has been conducted for gasoline fuels so far. Furthermore, the characterization of emulsions is important for other technical systems in process engineering or mechanical engineering. This includes maritime technology and the crude oil industry, in which the oil-pollution of seawater and (waste) water processing are relevant topics. Unwanted stable emulsions may form due to certain oil components or fine solids depending on the ambient conditions [17]. These processes are not fully understood, and diagnostic tools are necessary to characterize emulsions under different conditions in a technical system or process. For this purpose, optical techniques are utilized, which are mainly based on light extinction or scattering techniques. The change of the optical density with a variation of the process parameters can be detected with these techniques [17]. Near infrared spectroscopy (NIR) is based on absorption and scattering of light in the emulsion, but it does not provide droplet size distributions directly, and a work-intensive calibration is necessary.

Imaging techniques may provide information on droplet geometry (droplet size) and droplet dispersion (droplet concentration), which are important properties of emulsions. Furthermore, droplet interactions such as collision and coalescence can be studied. All these quantities are crucial for a deepened understanding of emulsion formation and stability. Laser-induced fluorescence (LIF) could be a suitable technique for studying such multi-component mixtures with phase separation. For the LIF approach, usually dyes are added to the investigated fluid, which are excited by a laser pulse and emit redshifted light.

LIF has a wide applicability for mixing and injection studies. For example, two-color LIF thermometry allows the determination of the temperature distribution in liquid flows, individual droplets, and sprays [18–23]. The temperature within the vapor phase can be determined with two-color LIF [24,25] or the two-line excitation laser-induced fluorescence technique [26–28]. The LIF technique can also be utilized to determine the fuel/air ratio within automotive sprays [26,29]. A combination of the LIF signal and the Mie scattering signal allows for the determination of droplet size distributions in terms of the Sauter Mean Diameter (SMD) within sprays [30–34].

Few LIF microscopy, experiments exist on very stable water/oil emulsions (20% oil) in which oil was doped with a dye [35]. The scope of this study was to track the movement of dispersed oil droplets in order to test oil-removal mechanisms. Regarding oil-pollution detection in seawater, the fluorescence spectroscopy of oil can be utilized to identify various types of oils and to measure oil concentrations [36]. In food technology, the lipid concentration of a milk emulsion was studied using the fluorescence spectroscopy of a dye dissolvable in lipids [37]. In general, to the best of our knowledge, no imaging studies exist for the characterization of fuel/water emulsions in which fuel is the continuous phase, and there are no mixing studies on fuel/water mixtures under IC engine relevant conditions. Up to now, no investigations exist on the distributions of fuel and water in the mixture before the injection (i.e., in the fuel system) or during the spraying process.

For the spectral separation of fuel and water in mixtures, it is necessary to find dye substances that are soluble in only one of the fluids, emit fluorescence in separate spectral bands, and show no distinct temperature sensitivity over a wide temperature interval, as such a dependence would drastically complicate measurements (since temperature information would additionally be required). Dyes soluble in polar liquids like water and alcohols are usually not soluble in nonpolar liquids like alkenes or gasoline. Furthermore, suitable dyes for ratio-based thermometry techniques like Rhodamine B, Sulforhodamine B, and Fluorescein are not applicable for the measurements due to their high temperature sensitivity. The dye Eosin-Y (disodium;2-(2,4,5,7-tetrabromo-3-oxido-6-oxoxanthen-9-yl)benzoate, ${\rm C}_{20}{\rm H}_{6}{\rm Br}_{4}{\rm Na}_{2}{\rm O}_{5}$) seems suitable for mixture studies, as it is only soluble in polar liquids and shows a low temperature sensitivity at moderate temperatures [30]. This characteristic has been confirmed for the solvent ethanol, but not yet for water.

In the present study, the dye Eosin-Y dissolved in water is used for studying the water dispersion in fuel. The LIF imaging technique is optimized for investigation of water/fuel mixtures as well as their blends with ethanol. For this purpose, first the photo-physical properties of the dye in water are investigated in a pressure cell under various conditions using a spectrometric LIF setup. In a second step, the dye is added to water and an imaging experiment is conducted using a planar LIF setup. Fuel mixed with 5–10 vol. % of water as well as with ethanol addition are investigated. This investigation provides the analysis of the spectroscopic properties of the dye under varied conditions as well as the characterization of the phase distribution and their miscibility depending on the water and ethanol content. This work is thus an essential step in the development of novel spectroscopic techniques for the analysis of water/fuel mixtures in general and for studying such blends for IC engine injection applications in particular.

2. MEASUREMENT PRINCIPLE

Laser-induced fluorescence is a widely used measurement technique for qualitative and quantitative investigations of mixture formation in IC engines. Tracer (or dye) molecules are excited by a laser pulse to a higher electronic energy level and emit fluorescence light when returning to their ground state. The fluorescence signal intensity ${S_{\rm fl}}$ depends on various parameters [28,38,39]:

Gasoline consists of about 200 different components, and its composition varies from country to country and between summer and winter. The aromatic component content in gasoline is up to 35 vol. %, consisting of toluene, benzene, and xylene. To ensure a well-defined mixture composition, the two-component surrogate fuel Toliso, which was already applied in earlier studies for studying spray and mixture formation and sooting combustion, was utilized [40,41]. Toliso consists of 65 vol. % isooctane and 35 vol. % toluene, and has similar thermo-physical properties as commercial multi-component gasoline. Mixtures of the aromatic hydrocarbon toluene and isooctane also have a similar sooting behavior as commercial gasoline [42]. To reduce the ${\rm CO}_{2}$ emissions and to comply with ${\rm CO}_{2}$-emission limits, conventional gasoline is often blended with biofuel components like ethanol. Thus, besides the base fuel Toliso (E0, 100 vol. % Toliso, 0 vol. % ethanol), the ethanol blends E5, E10, E15, and E20 were studied in this work. Ethanol can be used as a solvent for water and fuel. The physical, chemical, and optical properties of the investigated fuels are shown in Table 1.

Table 1. Physical and Chemical Properties of the Investigated Fuels and Water, 0.1 MPa [43–45]

View Table

The fluid’s surface tension is an especially relevant fluid property that determines the phase separation and miscibility of the components. While the surface tensions of the fuels are similar, water shows a much higher value. The viscosity and density determine the mixing and flow behavior of the phases as well, but in general, the values are comparable for all components of the mixtures. In principle, large viscosity of the continuous phase could reduce coalescence effects of the dispersed phase. The refractive index is similar for all fuels and water so that beam steering as well as light scattering and refraction effects inside the emulsion can be reduced in general.

“Eosin Yellow” (Eosin-Y, ${{\rm C}_{20}}{{\rm H}_6}{{\rm Br}_4}{{\rm Na}_2}{{\rm O}_5}$) belongs to the xanthene group and has a melting point of 568–569 K [46]. Eosin Y is a potential tracer for water, and has already been used for investigations of ethanol sprays [47]. The chemical properties of Eosin-Y provide good solubility in water and ethanol, but no solubility in hydrocarbon fuels such as isooctane and commercial gasoline. It was found that Eosin Y is not miscible in E0 (pure fuel), E5 (fuel containing 5% ethanol), E20, etc.; only high ethanol concentrations enable a limited miscibility (e.g., E80, E85, etc.). It can be reasonably assumed that the dye does not change the fluid properties of water and ethanol.

A common Nd:YAG pulsed laser with a wavelength of 532 nm can be used for excitation. The fluorescence light is emitted between 540 and 680 nm [47,48] (see also below). Eosin-Y shows a high fluorescence quantum yield of $\sim 0.68$ in ethanol at 500 nm excitation [49], which allows for a comparatively low dye concentration of 0.1 vol. % in this work. For IC engine measurements with changing pressure and temperature during injection, the temperature sensitivity of the fluorescence must be low or known for estimation of the local water concentration. Furthermore, effects of laser fluence on the LIF signal must be evaluated as well as concentration dependencies of the fluorescence. These cross-sensitivities are investigated in a pressure chamber, and the results are presented in Section 4.A. Afterwards, different fuel/water mixtures with ethanol addition are investigated; results are shown in Section 4.B.

3. EXPERIMENTAL SETUP

A. Fluorescence Spectroscopy Setup

The liquid solution of distilled water doped with Eosin-Y was studied in a small, optically accessible pressure cell. The corresponding optical setup is presented in Fig. 1. The beam (diameter 4.2 mm) of a pulsed Nd:YAG Laser (Quanta Physics) was divided using a 50/50 beam splitter. One part of the beam is guided to an energy meter for measuring the laser energy of the individual laser shots. The other part of the laser beam (laser pulse energy 3.6 mJ/pulse, laser fluence $26.0\,\, {\rm mJ}/ {\rm cm}^2$) is guided to the cell for excitation of the dye in water. The cell features a compact design to realize homogeneous temperature distributions in the investigated liquid mixtures. Higher liquid pressures up to 20 MPa and higher can be realized although, in this study, only measurements at 0.1 MPa were conducted. The outer dimensions of the cell are ${50}\;{\rm mm}\times {50}\;{\rm mm}\times {40}\;{\rm mm}$, and the inner volume is $6.6 \,\, {\rm cm}^3$. Four windows (sapphire) allow for optical access. An external thermostat (type FP50, Julabo, Germany) heats the walls up to 140°C. A magnetic stirrer (speed was fixed to 1500 rpm, stir bar dimensions ${8}\;{\rm mm}\times {3}\;{\rm mm}$) is installed for homogenization of the temperature inside the cell. The temperature of the liquid is controlled using two thermocouples in the cell (type K, 1.5 mm, tc-direct, Germany, at the top and bottom of the cell).

 

Fig. 1. Optical setup for the investigation of the spectral fluorescence of the tracer Eosin-Y in water (left), and detail of the micro cell (right).

Download Full Size | PDF

The fluorescence spectrum is recorded by using a spectrometer (type USB 4000, Ocean Optics, USA, wavelength range 495.9–831.8 nm, 3648 pixels, slit size 10 µm, integration time 100 ms, 100 subsequent spectra were averaged for each measurement). The broadband LIF emission of the liquid solution occurs between about 510 and 670 nm for all temperatures (Fig. 3) for an excitation at 532 nm. This means that blueshifted (“anti-Stokes” shifted emission) and redshifted (“Stokes”) fluorescence occurs in relation to the excitation wavelength.

B. Imaging Setup for Studying Emulsions

The optical setup for the fluorescence imaging measurements (see Fig. 1) is complemented with a beam expander (expansion factor 2) after the aperture and a cylindrical lens after the beam splitter to generate a light sheet (dimensions ${15}\;{\rm mm}\times {0.5}\;{\rm mm}$, laser pulse energy 3.6 mJ/pulse, laser fluence $48.0\,\, {\rm mJ}/{\rm cm}^2$). The laser sheet is positioned centrally within the measurement volume (i.e., 10 mm distance from the detection window, see Fig. 2). To ensure large optical access of the macro emulsions, a sealed square cuvette (type UQ-751, Portmann Instruments AG, Switzerland, inner dimensions ${20}\;{\rm mm}\times {20}\;{\rm mm}\times {20}\;{\rm mm}$) is used instead of the micro-cell to study the fuel/water emulsions. The cuvette is covered by a fixed plate to avoid mixtures with large amounts of air (in form of additional bubbles in the emulsion). A magnetic stirrer (speed was adjusted to 1250 or 2500 rpm, stir bar dimensions ${12}\;{\rm mm}\times {4.5}\;{\rm mm}$) maintains the emulsion. For fluorescence imaging, a CCD camera (type pco.1600, PCO-AG, Germany) is used in combination with a long-range microscope (type 12X Zoom, Navitar, USA; focus length 260 mm, resolution 59.3 pixel/mm) and a 532 nm long-pass filter (to suppress scattered light).

 

Fig. 2. Optical setup for the investigation of the fuel/water emulsions (left), and detail of cuvette assembly (right).

Download Full Size | PDF

4. RESULTS

First, the results of the spectroscopic fluorescence investigations are presented. In this case, only the dye Eosin-Y dissolved in water is characterized. Afterwards, first imaging studies are conducted for fuel/water mixtures testing the capability of the planar LIF technique to characterize the phases and emulsion formation in admixture with the biofuel ethanol. In these studies, different ethanol and water concentrations are tested.

A. Spectral Fluorescence Investigations

In this section, the effects of the dye concentration, photo-dissociation, and temperature on the fluorescence signal of Eosin-Y in water are presented. In Fig. 3, the influence of the dye concentration on the fluorescence properties is shown. The overall signal intensity rises with larger dye concentration without showing significant changes in the shape of the spectrum. The spectra are shifted slightly towards larger wavelengths with increasing dye concentration (5 nm for about 8-fold increase in dye concentration). This is explained by stronger reabsorption of fluorescence between 500 and 550 nm with increased dye concentration due to the overlap of the absorption and emission spectra [50]. The signal increases linearly between 0.016 and 0.125 vol. %. This means that no significant saturation effects of fluorescence and absorption along the beam path are observed. The inserted linearity diagram shows a slight deviation at 0.125 vol. %, which is caused by extinction effects due to the long laser path length within the measurement volume, leading to an overall signal reduction.

This linear behavior is important for application in IC engine measurements in order to quantify local water concentrations. In the subsequent spectral measurements, a relatively low dye concentration of 0.031% was chosen.

 

Fig. 3. (Left) Emission spectra of Eosin-Y in water normalized to maximum intensity at 0.125 vol%, and (right) with all spectra normalized to their respective maximum values; the inserted diagram shows the linearity ($R^2 = 0.969$) of the integral LIF signal for various dye concentrations in water, 293 K.

Download Full Size | PDF

 

Fig. 4. Left: temperature effect on spectral fluorescence of the dye Eosin-Y in water (0.1 MPa, normalized to 293 K) with inserted diagram showing the linearity ($R^2 = 0.998$) of the integral LIF signal. Right: normalized fluorescence signal drop due to photo-dissociation (293 K, normalized to 0 min) with inserted diagram showing the temporal decrease of the integral LIF signal.

Download Full Size | PDF

In Fig. 4 (left), the effect of temperature on the spectral emission is presented in a range of 293–333 K. With a temperature increase of 40 K, the spectrum is slightly redshifted and broadened, while the total intensity rises by less than 14%. For comparison, in previous measurements, the tracer Eosin-Y in ethanol showed a similar emission behavior and an increase of 7% in total intensity for a temperature variation of 28 K [30]. However, the fluorescence of the dye, in particular the position of the fluorescence peak, is sensitive to the polarity of the solvent. The peak position is around 540 nm for water as solvent and around 550 nm for ethanol [30]. This “blueshift” of the fluorescence in water was reported by Fleming et al. [50] to be a result of increasing ISC rate (intersystem crossing), also leading to reduced fluorescence lifetime and fluorescence quantum yield (FQY, which is 0.69 in ethanol and 0.20 in water).

In Fig. 4 (right), the effect of photo-dissociation on the LIF signal is presented in terms of total intensity. Photo-dissociation (or “photolysis”) is mainly relevant during fluorescence calibration measurements. In this case, the dye dissociates after a certain number of laser illuminations. For a constant number of dye molecules in a static calibration cell, photolysis can introduce errors in the data interpretation. Thus, the number of laser excitations at a given laser fluence is limited for measurements in calibration cells without significant dye circulation. The investigated solution was not circulated (no stirrer was applied) for investigation of the general impact of the photo-dissociation on the resulting LIF emission. The total signal is distinctly reduced already after a short measurement time. However, as mentioned previously, during the dye characterization only few seconds (10 s, at a laser repetition rate of 10 Hz) measurement time were chosen in order to reduce the effect of photo-dissociation on the calibration data.

The effect of photo-dissociation is only important for measurements where the fuel/tracer mixture is illuminated multiple times or is reused several times like in cuvettes or closed chambers [18,51]. Spray measurements in IC engines or constant volume chambers are typically not affected by photo-dissociation because of the high spray velocities and instant exchange of the dye/tracer mixtures (in the range of about 100 m/s for a direct-injection spark-ignition spray).

B. Imaging Study on Fuel/Water Mixtures and Emulsions

In this section, the dispersion of water in fuel is analyzed in an imaging study for varied water content. Furthermore, the effect of ethanol admixture on the emulsification of the fuel/water mixture is discussed. For this purpose, the fuel Toliso was blended with different ethanol percentages (in 5 vol. % increments up to 20 vol. %), and the added water was doped with 0.5 vol% Eosin-Y. Thus, the resulting dye concentration in the fuel/water (f/w)-ethanol mixture is then 0.05–0.1 vol. % and in a similar range as tested in the preliminary measurements shown in Fig. 4.

First, LIF single-shot images of a f/w mixture with 10% water are presented in Fig. 5 for two conditions (top row: low stirrer speed 1250 rpm; bottom row: high stirrer speed, 2500 rpm). Obviously, the water droplets are relatively heterogeneously distributed in the f/w-mixture for a low stirrer speed (top row), and the water droplets are relatively spherical. For increased stirrer speed, the larger of the dispersed droplets show a more distinct deformation because of the higher turbulent motion introduced by the stirrer. Droplet breakup (or droplet stretching, with deformation and elongation of the droplets in at least one direction) and coalescence (in this case, individual droplets appear more spherical than during their stretching, and are merged) processes are visible in the images. In general, the droplets in the middle of the light sheet appear brighter due to the Gaussian beam shape, which is coupled in from the left side.

 

Fig. 5. LIF single-shot images for f/w mixture containing 10% water; top row, low stirrer speed (1250 rpm); bottom row, high stirrer speed (2500 rpm).

Download Full Size | PDF

 

Fig. 6. LIF single-shot images for various f/w mixtures and ethanol contents, 293 K.

Download Full Size | PDF

At low stirrer speed, a large air pocket remains at the top of the cuvette (the cell was not completely filled with the liquid mixture; the air pocket at the top is obvious for quiescent liquids), while more air bubbles are formed at higher stirrer speed, and some stick also to the windows. Some of these air bubbles, as well as the water bubbles, act as a lens so that the shape of water droplets may appear deformed (this effect is visible in all recorded images).

In the next step, the f/w mixtures are studied at varying ethanol contents and two different water contents, as shown in Fig. 6. All subsequent data is recorded at high stirrer speed. The water droplets in the “coarse” emulsion show a spherical appearance for E0 and E5 for both water contents. For E0 and a water content of 5 vol. %, the water droplets are smaller (about 0.5 mm in diameter) in comparison to the mixture with a water content of 10 vol. % (droplets are in this case about 1–2 mm on average).

With increasing ethanol content, the pictures appear more homogeneous. On the one hand, the sizes of the water droplets may decrease with increasing ethanol content, but this, however, cannot be fully quantified within the present study. For a 15% and 20% ethanol addition (E15, E20), a homogeneous “fine” emulsion is generated for a 5% and 10% water content, which is also stable when the stirrer is switched off (at least for some minutes). On the other hand, the LIF signal intensity is also increased with ethanol addition, which is distinct for the variation from E5 to E10. As mentioned previously, the FQY of Eosin-Y dissolved in ethanol is higher by a factor of 3.45 than for the solution in water.

The homogeneity of the f/w mixture with and without ethanol addition can be assessed by calculation of the coefficient of variation (COV) of the 50 consecutively recorded single images. The COV is shown in Fig. 7 in the form of a spatial distribution and in Fig. 8 as a pixel-wise evaluation. The COV was determined within a given ROI of 10 mm diameter (see Fig. 8). The COV displays the spatial and temporal LIF-signal fluctuations (and thus the variation of the water dispersion) within the mixture.

 

Fig. 7. COV calculated out of 50 consecutively taken LIF single-shot images for various ethanol and water contents, 293 K.

Download Full Size | PDF

 

Fig. 8. Summed up COV values in the ROI for different ethanol and water contents, featuring a detailed view of the processed image with ROI within the cuvette, 293 K.

Download Full Size | PDF

For E0 and E5, it is apparent that the fewer but larger water droplet sizes for 10% water addition result in larger corresponding COVs and thus lead to the largest inhomogeneities of the f/w mixtures (see also Fig. 6). The COV images show an increase in homogeneity of the f/w mixture with increasing ethanol content. Some bubbles at higher ethanol concentrations are visible in the COV image, which are caused by air within the cuvette and not caused by the emulsion itself. These bubbles stick to the glass wall for some time, and fluorescent light is scattered and diffracted at their surface so that they are visible in many single images and in the COV image as well. These air bubbles are also visible in the single-shot images for the mixtures E0 and E5, etc. (see Fig. 6).

The accumulated COVs of the investigated fuel/water emulsions are shown in Fig. 8. The COV curves quantify the inhomogeneity depending on the water and ethanol content in the mixture. For all ethanol concentrations, the COV is larger for increased water content. For E0, the COV is more than doubled for a water volume fraction of 10 vol. %, while for E20, the COV is very similar for 5 and 10 vol. % water. It can be concluded that the water droplets disappear (i.e., they become probably smaller and show a reduced number) with increasing ethanol content, as ethanol is a good solvent for water and fuel. Thus, ethanol addition leads to an overall more homogeneous mixture of the fuel–water–dye–ethanol system.

It can be concluded that the technique is capable of studying f/w mixtures and emulsions with ethanol at varied conditions. The mixture quality in terms of homogeneity and resulting droplet structures can be assessed under simplified conditions. In the future, the LIF spectroscopy technique must be optimized to incorporate further cross-sensitivities and to enable microscopic analysis of fine emulsions under pressurized flow conditions. For very dense mixtures in which multiple scattering is severe, a method to compensate these effects must be applied. One way is the application of structured light illumination planar imaging (SLIPI), which is currently developed for spray analysis [30,31,48,51].

5. CONCLUSION AND FUTURE WORK

This work presents a first study for the investigation of the dispersed water phase of fuel/water mixtures relevant for injection and mixture-formation studies in IC engines. The laser-induced fluorescence of the dye Eosin-Y was used to track water in the liquid phase. First, a spectral characterization of the fluorescence of Eosin-Y in water was conducted in a heated pressure cell to identify cross-sensitivities introduced by temperature, tracer concentration, and photo-dissociation. The fluorescence signal showed a linear behavior as a function of the dye concentrations in the studied range. The temperature sensitivity of the integral LIF signal was investigated between 293 and 333 K, resulting in a value of about 0.4%/K. The effect of photo-dissociation of Eosin-Y in water is not negligible, and has to be taken into account when the fuel/dye mixture is illuminated multiple times or when it is reused several times. This observation is not critical for mixing studies in a flow or a spray, where the fuel/dye mixtures are usually just illuminated once.

The presented LIF concept was then used in an imaging study to analyze the water distribution in fuel/water mixtures. Emulsions with 10 different fuel/water/ethanol compositions were studied. The surrogate fuel Toliso was blended with ethanol (in 5 vol. % increments up to 20 vol. %). The LIF signal of the dye in the emulsions can be utilized for evaluation of the structure and interaction of water droplets, and for sizing of the dispersed water droplets. Furthermore, the homogeneity of the mixture was analyzed in terms of the COV. The investigations showed that the diameter of the water droplets decreases at a given water content with increasing ethanol concentration. The water droplets at a water content of 5 vol. % are much smaller than those at a content of 10 vol. %. The mixture homogeneity is improved for 5 vol. % water mixtures, which is visible in the reduced COV. For an ethanol percentage of 20 vol. %, the homogeneity of the mixture of both water contents is maximal and very similar.

For a deepened understanding of the fuel and temperature effects on the fluorescence of Eosin-Y with various solvents, further fundamental characterization of the fuel/tracer mixtures has to be performed. In the future, this technique will be improved to allow for quantitative microscopic analysis of water/fuel emulsions at different flow conditions (e.g., at varied pressure, velocity, turbulence, composition, temperature, added surfactants). For dense mixtures with multiple scattering effects, a method must be applied in order to compensate these LIF signal disturbances. From the LIF signal, further quantities can be derived such as local water volume fractions, and droplet coalescence and breakup rates. Furthermore, the technique can be applied in sprays under IC engine conditions in order to analyze the fuel and water distribution in the cylinder for different injectors, injection parameters, and ambient conditions, and to correlate the data with flame propagation and pollutant emission results.