Cavity ringdown specectroscopy (CRDS) methods have become widely used for trace gas concentration measurements as an enhanced absorption spectroscopy method. We have applied CDRS to a sample cell and simulated particulate soot to develop a CDRS system that will allow rapid measurement of soot absorption and hence density in raw diesel engine exhaust. Results will be shown from several similar systems illustrating the range of improvements available from system optimization. A final system will be shown that appears to be reasonably successful in terms of sensitivity, but is at present not able to provide reliable single-shot measurements. Illustrations of future modifications we plan to develop the single-shot capability will be shown.
Cavity ringdown spectroscopy (CRDS) is one way of measuring small fractional absorption for concentrations down to sub- ppm levels in an optical cavity. The applicability of CRDS for measurement of diesel exhaust is assessed. By use of CRDS, instantaneous exhaust characteristics can be immediately available for engine control. Of specific interest in this set of experiments are measurements of soot or particulate matter. The goal of these measurement would be to estimate the number density of soot particles based on line-of-sight absorption and an assumption regarding the size distribution of the particles. To test the system, the absorption coefficients of Nitrogen, Carbon Dioxide, and 2',7'-Dichlorofluorescein in a test cell at known number densities were measured instead of diesel exhaust. An Nd:YAG-laser provided light at 532 nm for the cavity ringdown measurement. Cavity ringdown waveforms were analyzed with an exponential least squares fit. An equation relating the absorption coefficients to the ringdown time in a test cell was developed including the two parallel windows in the optical cavity. This analysis is based on the assumption that the measured output of the cavity decays exponentially according to a first order expression and the Beer-Lambert law. In the case of nitrogen, the measurement result were in good agreement unit, typically used for diesel exhaust soot measurement.
Evolving ideas of how high pressure atomization occurs, why it occurs, and what parameters might have a significant influence on atomization present some new and interesting challenges for measurement system. Of particular interest for transient fuel sprays is the region near the fuel injector tip itself, where the spray seems to begin the atomization process. This region of a spray is generally optically dense, has a large number of ligaments and non- spherical liquid elements, and probably has a large number of droplets as well. The high number density of liquid elements as well as absorption in the liquid makes optical probing of this region difficult. Reliable data in this region of the spray would be very useful for better understanding of the atomization process in these systems, as well as providing data for model validation for computer models being developed to approximate this behavior. This paper will try to briefly review the most common droplet sizing methods currently used with an emphasis on their performance in dense spray regions.
A two dimensional optical diagnostic technique based on light extinction was improved and demonstrated in an investigation of diesel spray characteristics at high injection pressures. Traditional light extinction methods require the spray image to be perpendicular to the light path. In the improved light extinction scheme, a tilted spray image which has an angle with the light path is still capable of being processed. This technique utilizes high speed photography and digital image analysis to obtain qualitative and quantitative information of the spray characteristics. The injection system used was an electronically controlled common rail unit injector system with injection pressures up to 100 MPa. The nozzle of the injector was a mini-sac type with six holes on the nozzle tip. Two different injection angle nozzles, 125 degree(s) and 140 degree(s), producing an in-plane tilted spray and an out of plane tilted spray were investigated. The experiments were conducted on a constant volume spray chamber with the injector mounted tilted at an angle of 62.5 degree(s)$. Only one spray plume was viewed, and other sprays were free to inject to the chamber. The spray chamber was pressurized with argon and air under room temperature to match the combustion chamber density at the start of the injection. The experimental results show that the difference in the spray tip penetration length, spray angle, and overall average Sauter mean diameter is small between the in- plane tilted spray and the out of plane tilted spray. The results also show that in-plane tilted spray has a slightly larger axial cross- section Sauter mean diameter than the out of plane tilted spray.
Studies of direct injection, compression ignition engines demand a particle sizing technique which can reliably retrieve droplet sizes in transient fuel sprays. Many of the existing sizing techniques cannot analyze the dense, transient fuel sprays or provide information which is awkward to interpret. This work compares the performances of a diffraction based particle sizing technique and a light extinction sizing technique on a series of Diesel sprays. Both techniques were effective in acquiring useful data from the sprays. Under equivalent conditions, the light extinction technique provides somewhat smaller sizes than the diffraction based method.
Current emphasis on improving performance and reducing emissions in direct injection compression injection (Diesel) engines demands detailed understanding of high pressure transient sprays. Combustion studies require quantitative information, specifically local droplet sizes and velocities, throughout the spray field. The turbulent and unpredictable nature of the sprays complicates the analysis. Most tools currently available for spray analysis provide qualitative information over a large field of view or quantitative information at a single point in the spray field. This paper discusses the implementation of a diffraction-based diagnostic capable of determining particle size information over the full field. This diagnostic and particle image velocimetry together can provide a spatial map of local particle sizes and velocities in the spray.
Application of particle image velocimetry (PIV) techniques for measurement of fluid velocities typically requires two steps. The first of these, is the photography step, in which two exposures of a particle field, displaced between the exposures, are taken. The second step of the application is the evaluation of the double exposure particle pattern and production of appropriate particle velocities. Each of these steps involves optimization which is usually specific to the experiment being conducted and there is significant interaction between photographic parameters and evaluation characteristics. This paper will focus on the latter step, that of evaluation of the double exposure photograph.
Among the various evaluation techniques suggested for analysis of PIV images is the evaluation of the scattered interference pattern (Young's fringes) by numerical Fourier transform. An alternative to the numerical calculation of the Fourier transform of the Young's fringes has been suggested, using a modified liquid crystal television as an optical correlator to allow the transform to be performed optically, thus speeding up the interrogation process.
Both transform techniques are affected by the quality of the input function, specifically the Young's fringes. This paper will compare the performance of optical and numerical fourier transform analysis of Young's fringes using speckle images.
The repeatability and an estimate of the accuracy of the particle displacement will be shown for each method.
An optical technique for analyzing particle image velocimetry data has been demonstrated on a variety of particle patterns from actual flows and from similar translate diffuser experiments. The technique uses point-by-point illumination of the particle image and two subsequent 2D FFTs to produce an autocorrelation pattern of the particles from which the particle displacement is determined. The dynamic range in the technique''s configuration is about 30:1, with reasonably good processor speed.