Ejection Nozzle Research Articles

Carbon Particle Cloud Generation for a Solar Particle Receiver

The development and performance of a full-scale carbon particle cloud generator together with the evaluation of nine commercial carbon blacks is reported. Large variations were found in the dispersability and settling properties of the investigated powders. Scanning electron microscope analysis of cloud samples from different powders showed unequal state of agglomeration and particle size. The particle population distribution of the most suitable powder was determined, showing that the particle cloud consisted of 99.8% spheroid primary particles (25–570 nm dia) and 0.2% large irregularly shaped agglomerates. Although the numerical fraction of the agglomerates was only 0.2%, they contributed 40% to the cloud’s geometrical cross section. Significant variations in the population distribution were found from different batches of the same particle powder. The developed full-scale particle generator was capable of sustained operation, creating a particle cloud with an extinction coefficient exceeding 40m−1 at a nominal flow rate of 25 SLPM. The dispersal efficiency of the system with the optimal ejection nozzle was 25%, compared to less than 1% for free ejection. The particle dispersal rate was 30 g/hr at 25 SLPM at an evacuation efficiency of 87%. Specific extinction cross-sections of 5.8m2/g were measured for particle mass loading higher than 2g/m3.

Measurement of heat loss due to thermal radiation of liquid spray flames using shock tube technique

A single-compression, “tailored” interface shock tube was used to measure thermal energy radiating from diesel combustion. The fuel (light oil) was injected through a throttle nozzle (pintle diameter: 1 mm) into air behind a reflected shock wave, 37, 78, and 120 mg of fuel were injected at pressures of 15 and 20 MPa. Monochromatic emission at wavelengths of 0.6328, 0.9, 1.1, 1.45, 2.5, 3.56, 3.92, and 4.2 μm was followed with IR-detectors at different distances from the injection nozzle. These signals were obtained simultaneously, together with pressure and nozzle lift sensor signals. In another experiment, time histories of the flame radius at diffrent distances from the injection nozzle were determined by observing light signals of near-IR detectors set radially with relation to the shock tube. Heat loss due to thermal radiation was calculated using the time histories of both monochromatic emission and the flame radius. Soot concentration profiles were also obtained using this data. Results were as follows: (1) Heat loss per injection is approximately proportional to the injection pressure. (2) Heat loss is also proportional to the amount of fuel. (3) The ratio of thermal radiation to the lower calorific value of light oil is about 5 to 10%. (4) The volume concentration of soot increases gradually near the ejection nozzle and decays rapidly. At points far from the injection nozzle, soot increases rapidly and decays gradually. The maximum volume concentration of soot is about 10 −4 cm 3 soot/cm 3 gas.

Formation of large metal clusters by surface nucleation

The formation of metal clusters with hundreds of atoms in and near vapor ejection nozzles is analyzed. A previously discussed model, cluster generation by homogeneous nucleation and collisional growth in the vapor phase, is found to be untenable because it cannot account for the large number of collisions required to remove the latent heat of condensation, freed during cluster growth. An alternative mechanism that involves heterogeneous nucleation (i.e., cluster formation on surfaces) is proposed and evaluated. To determine the applicability of this mechanism, the following process steps are considered: formation of an adequate supply of critical embryos on nozzle and adjacent crucible walls, growth of embryos into large clusters, and cluster ejection into the surrounding vapor atmosphere. The conclusion is that, with nonwetting vapor and surface material combinations, substantial quantities of large clusters can be formed and ejected. Specifically, predictions for silver vapor at a few Torr pressure, passing through a graphite nozzle at about 1600 K, are cluster generation rates of 1013–1014 clusters/cm2 s, and cluster sizes up to about 600 atoms. Available experimental observations support these predictions qualitatively.

Acoustic mode determination in solid rocket motor stability analysis

Introduction T HE determination of acoustic reference modes is almost a mandatory step in solid rocket motor stability analysis. An exception to this statement is provided by the recent development of computer codes that solve the full unsteady equations of motions inside the rocket motor, including the choked ejection nozzle. However, these codes presently do not yet represent the state-of-the-art and both the engineer and the research scientist need to rely on more conventional, less expensive methods to study the natural unsteady behavior of solid rocket motors. These conventional analyses — such as the acoustic balance method — call for the determination of reference acoustic modes of the rocket engine combustion chamber. These modes are used primarily to evaluate the stability integrals that define the driving or damping contributions to the chamber stability; consequently, the accuracy of the stability analyses is directly governed by the quality of the acoustic mode determination. The critical steps in evaluating such acoustic modes are 1) the definition of a closed cavity in which to perform the acoustic calculations and 2) the choice of acoustic boundary conditions. The cavity is usually taken to be limited by the head end of the combustion chamber, the burning surface, inert walls in certain cases, and an aft closure plane that formally separates the low-velocity flowfield inside the chamber from the high-velocity flowfield inside the nozzle (Fig. 1). It is of primary importance that the aft closure plane be located within the low Mach number domain because stability analyses assume a quasi-incompressible mean flowfield.' Apart from this restriction, the precise location of this aft closure plane is theoretically free. Obviously, an ideal acoustic mode determination should ensure that the stability integrals are insensitive to the position of the aft closure plane within the incompressible domain. It is the purpose of this work to show that the classical rigid-wall assumption used to evaluate the chamber acoustic modes does not ensure the independence of the stability integrals from the location of the aft closure plane. This point is particularly crucial when the nozzle length is not small compared to the motor length. A simple acoustic admittance boundary condition will be proposed that greatly improves the stability analyses through an improved reference acoustic mode determination.