Previous chapters discussed nationwide mobile communication system technologies and modeled user costs and benefits. This chapter provides a brief overview of mobile communication system economics, the projected U.S. market for nationwide mobile communications, and the potential revenues. The basic cost structure of satellite and meteor-burst systems are also modeled. A combination of all these factors will ultimately govern which systems will be commercially successful. Because of improvements in technology, communication system costs and user terminal costs (A) per unit of capacity tend to decline over time. However, inflation tends to drive the price of an identical product upward over time. GRA N91-12001# Naval Postgraduate School, Monterey, CA. Dept. of Administrative Sciences NATIONWIDE MOBILE COMMUNICATION SYSTEMS. VOLUME 3: APPENDIX B D M.S. Thesis William Joseph Schworer, III Jun. 1990 191 p The following contents include: Communication, Satellite and Navigation Concepts, Mobile Communications Cost/Benefit Spreadsheet, and Operating Statistics and Cost Analysis. GRA N91-12002# Naval Postgraduate School, Monterey, CA. Dept. of Electrical and Computer Engineering. AN ELECTROMAGNETIC RADOME MODEL USING AN Robert Johnston Vince Dec. 1989 175 p The goal of this thesis was to develop and validate a set of microcomputer programs using, in part, a previously written finite-element algorithm to analyze the perturbation of an incident electromagnetic field as it penetrates a missile radome. An interactive program was developed to design the radome structure using aerodynamic shaping functions and provide structure and material files as input to the finite-element code. A second program was developed to use the spherical harmonic expansion coefficients provided by the finite-element code to assemble the electromagnetic fields within the radome core and to display the fields that appear across a planar antenna with three-dimensional graphics for any orientation of the antenna. Algorithms were included which compare the computed field components to the theoretical incident plane wave for each stipulated angles of incidence, in order to determine the perturbation due to the presence of the radome. Validation of the computational method was attempted by analyzing the perturbation indicated for an ideal radome with relative permittivity of unity. The attempted validation showed phase errors in the computed fields which were minimal for axial incidence, but became significant for highly canted incidence. GRA N91-12003# Harry Diamond Labs., Adelphi, MD. THE NORMAL MODE E- AND H-FIELD DISTRIBUTION IN THE CONCENTRIC SPHERICAL CAVITY Final Report, Dec. 1987 Dec. 1988 Louis F. Libelo and Guy E. Pisane Apr. 1990 117 p (AD-A225388; HDL-TR-2168) Avail: NTIS HC/MF A06 CSCL 20/3 The eigenfields of the concentric spherical cavity are investigated in this report. This simple cavity structure serves as a foundation toward understanding a number of important practical problems. The fields are developed mathematically, and extensive computer-generated graphs are presented. A thorough exposition accompanies the graphs explaining how the fields vary with the cavity parameters in a systematic fashion, and how they relate to the corresponding eigenvalue mode charts. N91-12004# GRA Ohio State Univ., Columbus. ElectroScience Lab. TM SCATTERING BY A VARIABLE SHEET IMPEDANCE IN A MULTILAYERED SLAB Matthew E. Peters and Edward H. Newman Mar. 1990 73 p (Contract N00014-89-J-1007) (AD-A225586; TR-721566-1) Avail: NTIS HC/MF A04 CSCL 20/3 The problem considered is two-dimensional transverse magnetic (TM) scattering by a variable sheet impedance in a multilayered slab. An integral equation is formulated for the sheet impedance in the multilayered slab. This integral equation is solved for the surface currents flowing on the sheet impedance using a spectral domain moment method (MM)/Green's function solution. The MM solution is outlined and expressions for the expressions for the matrix elements are obtained. A user oriented computer code was written to implement the solution. Numerical results are obtained and are compared with measured or previously calculated results. GRA ANALYSIS OF IRIS AND VARAN-S IMAGE DATA, AUTOMATED GEOMETRICAL PROCESSING, AND DATA EXTRACTION [ANALYSE VARAN-S EN IRIS RADARBEELDEN GEAUTOMATISEERDE GEOMETRISCHE VERWERKING EN DATA EXTRACTIE] B. M. Braam, H. J. Buiten, D. H. Hoekman, and H. J. C. vanLeeuwen (Wageningen Agricultural Univ., Netherlands) Feb. 1990 134 p In DUTCH (BCRS-89-37; ETN-90-97686) Avail: NTIS HC/MF A07 The data collected during several flights with the VARAN-S and IRIS SAR (Synthetic Aperture Radar) imaging radar systems over the Fleroland test site were analyzed in the framework of the AGRISCATT-87 campaign. The image processing and data extraction were automated because of the large amount of work and routine actions for every single image. A method was developed which makes it possible to process the (multidimensional) radar data. It is possible to join all the images to a geometrical reference so that production of tables, for example field averages, are easy and can be easily adjusted according to the wishes of the researchers. Topographical overlays can be made. The end product is a basic technique delivering tables and geometrically corrected images which are needed for the analysis of the VARAN-S and IRIS data. 33 ELECTRONICS AND ELECTRICAL ENGINEERING ESA International Conference on Large High Voltage Electric Systems) as a U.S. technical expert advisor. Over 200 papers were discussed, contributed from over 45 countries at the conference on all aspects of electric power generation and transmission. Of special interest was a panel session on superconducting technology for electric power systems and the participation on a new task force on the electrical insulation at cryogenic temperatures. Significant insight was gained into the development of superconducting power technologies in Europe and Japan. CIGRE has set up a committee to follow the development in research on the biological effects of electric and magnetic fields. The traveler also visited the Centre for Electric Power Engineering at the University of Strathclyde, Glasgow, Scotland and discussed research on degradation of polymeric cable insulation and gas insulated equipment. DOE N91-12008# Illinois Univ., Urbana-Champaign. Dept. of Electrical Engineering. CURRENT DENSITY CALCULATION USING RECTILINEAR REGION SPLITTING ALGORITHM FOR VERY LARGE SCALE INTEGRATION METAL MIGRATION ANALYSIS M.S. Thesis Hungse Cha Jul. 1990 63 p (Contract F30602-88-D-0028) (AD-A224750; UILU-ENG-90-2223; DAC-22) Avail: NTIS HC/MF A04 CSCL 20/3 In Very Large Scale Integration (VLSI) chips, metal migration (MM) is an important problem from the reliability standpoint. Furthermore, as the feature size is scaled down, MM becomes an even greater problem because of the higher current densities that would exist in the power and ground busses. Because of the complexity of VLSI power busses, there exists a need for a computer-aided design tool to correctly predict the likely failure site(s). A primitive splitting algorithm that calculates current density waveforms efficiently is examined. These waveforms are used to find the Median Time to Failure (MTF), a major parameter of concern in predicting MM. This algorithm was motivated by examining the equipotential plots obtained through finite element method analysis of simple regions. It was successfully implemented and tested, and some examples are described. GRA N91-12009# California Univ., Los Angeles. Dept. of Chemistry. ELECTRICAL RESISTIVITY, MAGNETIC SUSCEPTIBILITY, AND THERMOELECTRIC POWER OF PtGa2 Technical Report No. 1, 1 Oct. 1989 - 31 May 1990 Submitted for Li-Shing Hsu, Lu-Wei Zhou, F. L. A. Machado, W. Gilbert Clark, (Contract N00014-90-J-1178; NSF DMR-84-09390) The electrical resistivity (p), magnetic susceptibility (X) and thermoelectric power (S) of PtGa2 were measured as a function of temperature (T). This compound is metallic at high temperatures, as shown from the room-temperature resistivity value and the linear dependence of the S vs. T curve at temperatures above the Debye temperature. It undergoes a superconducting phase transition with a critical temperature (Tc) at zero magnetic field of 2.13K. The density of states (DOS) at the Fermi energy (EF) at high temperatures obtained from X and S data are 22 and 15 percent higher, respectively, than the value obtained previously from a semi-empirical band-structure calculation. GRA SYNTHESIS, TRANSPORT, AND MICROSTRUCTURAL Jong-Gyu Lee, K. V. Ramanujachary, and M. Greenblatt 1 Jun. 1990 6 p Submitted for publication (Contract N00014-82-K-0317; NSF DMR-87-14072) The n-type superconductor, Nd1.85Ce0.15CuO(4-delta) was prepared by solid state reaction and solution precursor techniques under various conditions and characterized by means of X ray diffraction, magnetic susceptibility, electrical resistivity and SEM measurements. Samples prepared by solution method displayed higher normal state resistance but better superconducting volume fractions than those prepared by conventional ceramic techniques. In addition, the solution precursor technique reduces the sintering temperature from 1150 C (used for solid state route) to 1000 C in achieving the superconducting phase. The solution-route yields particles with relatively uniform size distribution, but poor connectivity between the grains. In contrast, the solid state preparations yield well connected grains but with a larger distribution of sizes. The observed differences in the superconducting properties of samples prepared by different techniques were attributed to the differences in their microstructure. GRA N91-12012# Naval Surface Weapons Center, Silver Spring, MD. Dept. of Research and Technology. SBIR REPORTS ON THE CHEMISTRY OF LITHIUM BATTERY TECHNOLOGY W. P. Kilroy, ed. Nov. 1989 183 p (Contracts N60921-86-C-A457; N60921-88-C-0058; N60921-86-C-0274; N60921-88-C-0057) (AD-A225139; NSWC-MP-89-242) Avail: NTIS HC/MF A09 CSCL 10/3 The following contents are included: Identification of an Improved Mixed Solvent Electrolyte for a Lithium Secondary Battery; Catalyzed Cathodes for Lithium-Thionyl Chloride Batteries; Improved Lithium/Thionyl Chloride Cells Using New Electrolyte Salts; Development of Calcium Primary Cells With Improved Anode Stability and Energy Density. GRA (DE91-000095; SAND-89-0688) Avail: NTIS HC/MF A03 The design intent, design considerations, system use, development, product characteristics, and early production history of the MC3714 Thermal Battery was described. This battery has a required operating life of 146 s above 24.0 V with a constant current load of 0.5 A. It is activated when the MC3830 Actuator initiates the WW42C1 Percussion Primer in the battery. The MC3714 employs the Li(Si)/LICI-CCI/lithiated FeS2 electrochemical system. The battery is a hermetically sealed right circular cylinder with an antirotation ring brazed to the base of the cylinder. The battery is 50 mm long and 38.1 mm in diameter. The mass of the battery is 165 g. The battery was designed and developed to provide the power for the W82 JTA Telemetry System. DOE N91-12014# Universiteit Twente, Enschede (Netherlands). Siep Onneweer 1989 212 P Sponsored by Stichting voor Fundementeel Onderzoek der Materie (ISBN-90-9002692-4; ETN-90-97769; UDC-621.38(043.3)) Copyright Avail: NTIS HC/MF A10 The development of electronic circuit realizations of multiple valued logic is addressed. CMOS circuits were investigated and a current mode CMOS logic, based on a very small set of basic circuit elements, results. The basic elements of current mode CMOS logic, their computer aided design and experimental evaluation and verification are described. Contributions of the study to circuit design, logic design and computer aided logic synthesis are summarized. N91-12015# European Space Agency, Paris (France). DEVELOPMENT AND TEST OF POWER AMPLIFIERS FOR DRIVING LOW-VOLTAGE PIEZO-ACTUATORS IN MULTISTABLE INTERFEROMETERS ESA Hans Naumann and Walter Schmidt (Deutsche Forschungsanstalt fuer Luft- und Raumfahrt, Brunswick, Germany, F.R.) Aug. 1990 44 P Transl. into ENGLISH of Entwicklung und Test von Leistungsverstaerkern fuer Niedervolt-Piezoantriebe zum Einsatz in Multistabilen Interferometern (Brunswick, Fed. Republic of Germany, DFLR), 1989 Original language document was announced as N90-18662 (ESA-TT-1196; DLR-Mitt-89-19; ETN-90-98010) Avail: NTIS HC/MF A03 A power amplifier for capacitive loads such as low voltage piezoelectric translators is described. Various features such as offset control, input attenuation, filtering and hysteresis compensation, required for use in arrangements with multistable interferometers, are discussed. The results relating to optical multistabilty were satisfactory. Up to 95 stable steps were achieved in the optical analog/digital conversion. ESA N91-12016# Air Force Systems Command, Wright-Patterson AFB, OH. INTERLOCK DEVICE Yu. A. Ananyev, V. F. Borisov, V. M. Irtuganov, V. P. Kalinin, and V. S. Popov 12 Jun. 1990 8 p Transl. into ENGLISH of Russian Patent no. 421084, 25 Apr. 1972 p 1-2 (AD-A225261; FTD-ID(RS)R-0410-90) Avail: NTIS HC/MF A02 CSCL 10/2 The invention pertains to pulsed power supplies and can be used in protecting flash tubes from surges in laser pumping systems. There exists an interlock device containing (parallel-connected to each other) an inductive storage, a load in the form of m branches, each of which consists of n series-connected gas filled flash tubes, an interlock discharger with a firing unit and a current supply connected across a current commutator to the inductive storage, a control pulse generator, and a flash tube firing element. In this device, the interlock discharger is started by means of a circuit for comparing the voltage in the load with the reference voltage, which does not enable the tubes to be reliably protected against overloading when some of them fail because the surge effects at the output of the inductive storage are smoothed out due to mutual shunting of the parallel load branches and the resistance drop in flash tubes with rise in current. GRA N91-12017# Naval Postgraduate School, Monterey, CA. VARIABLE-FREQUENCY PULSE-WIDTH-MODULATION (PWM) FOR ZERO VOLTAGE SWITCHING IN A BOOST dc-dc REGULATOR M.S. Thesis Daniel S. Hunter Mar. 1990 68 p (AD-A225303) Avail: NTIS HC/MF A04 CSCL 10/2 A technique for operating a pulse width modulated (PWM) dc-dc regulator in the boost mode while switching the MOSFET when the drain-to-source potential is near zero volts was developed and is described. This is accomplished by using frequency-modulation in addition to pulse width modulation. Zero-voltage switching will provide power converter designers an alternative for designing high frequency converters with minimal transient turn-on losses, the predominant form of converter losses experienced in high frequency operation. High frequency operation will result in smaller reactive components, which produce higher power density converters, as well as increasing the transient response of the regulated converter. In addition to allowing for high frequency operation, the design exhibits many desirable power switch properties, such as limiting the peak voltage to the output voltage level and operating with the minimum possible current levels for a given power requirement. A circuit built and tested utilizing zero-voltage switching in a regulated boost converter verified the principles of operation for yielding a high efficiency, high frequency converters. GRA N91-12018# Naval Postgraduate School, Monterey, CA. CIRCUIT MODELS FOR INDUCTIVE STRIPS IN FIN-LINE M.S. Thesis Georgios Karaminas Dec. 1989 108 p (AD-A225309) Avail: NTIS HC/MF A06 CSCL 09/1 This thesis describes a computer aided design (CAD) compatible circuit model for an inductive strip centered in WR(90) fin-line with W/b less than 1, E, 1. The circuit model is shown to predict strip scattering coefficients which agree closely with those computed numerically using the spectral domain method. GRA N91-12019# Carnegie-Mellon Univ., Pittsburgh, PA. Dept. of Electrical and Computer Engineering. OPTIMAL PHASE-ONLY FILTERS Final Report, Apr. 1988 - B. V. Kumar and Z. Bahri Apr. 1990 100 p (AD-A225312; RADC-TR-90-17) Avail: NTIS HC/MF A05 CSCL 12/9 Phase-only Filters (POFs) and Binary Phase-only Filters (BPOFs) were studied. It is proved that the conventional POF introduced originally by Horner and Gianino is indeed the best among all POFS from the viewpoint of maximizing the Signal-to-Noise Ratio (SNR). The importance is illustrated of selecting the optimal support function for the POF and show with the help of example images that output SNR can be improved considerably (by about 5 dB) by using optimal POFS (OPOFs). An efficient algorithm is described for designing OPOFs. Because of their suitability for binary Spatial Light Modulators (SLMs) such as the Magneto Optic SLM (MOSLM), binary POFS (BPOFs) are of much interest. The optimal BPOF (OBPOF) is introduced and efficient algorithms are presented for designing OBPOFs. Several simulation results based on the use of 32 X 32 images are included to verify the performance of OBPOFS. GRA N91-12020# Naval Postgraduate School, Monterey, CA. PHYSICAL PROCESSES IN HOLLOW CATHODE DISCHARGE M.S. Thesis Hwang-Jin Han Dec. 1989 61 p (AD-A225329) Avail: NTIS HC/MF A04 CSCL 09/1 The hollow cathode is an effective source of dense, low energy plasma. Hollow cathodes find use in ion beam sources for laboratory and space applications. They can also be used independently for satellite charge control, and ion beam neutralization. A heaterless hollow cathode design was tested with argon gas used as a propellant. The device properties, that is, the emission currents as a function of discharge current, propellant flow rate and other physical parameters were investigated. Starting behavior was a main point of the investigation. The results of these experiments were compared with studies of the conventional hollow cathode. GRA Republic of China), no. 6, 1984 p 11-23 (AD-A225344; FTD-ID(RS)R-0306-90) Avail: NTIS HC/MF A03 CSCL 20/6 As the use of lasers in the military and in the civilian economy increases with each passing day, it is often necessary for the human eye or sensitive instruments to observe weak lasers, such as the return waves of laser radar and laser communications signals; but it is also necessary to provide protection against damage to the eye from the strong lasers of enemy laser weapons. For this reason, it is necessary to have a kind of automatic optical self-protecting switch. Based upon a study of the transmitting and scattering characteristics of multilayer dielectric optical waveguides, a practical computer program is set up for designing a type of auto-self-protecting optical switch with a computer model by using the nonlinear property of dielectric layers and the plasma behavior of metal substrates. This technique can be used to protect the human eye and sensitive detectors from damage caused by strong laser beams. GRA INTERFACE-STATE MEASUREMENTS OF GaAs SCHOTTKY A recent series of measurements has shown that the barrier height of Schottky diodes on GaAs may change under long term biasing conditions. The aging has been found to occur under reverse bias conditions with a logarithmic dependence on time. This study has been concerned with looking at the interface-state configuration of these diodes using admittance techniques to see if interface-state generation/destruction plays any role in this aging process. Two groups of diodes, Au/W/GaAs and Au/Pt/Ti/GaAs, were studied. Both of these diode types are contained within power MESFET structures. As has been reported elsewhere, the Au/W/GaAs, diodes exhibit pronounced aging effects, while the Au/Pt/Ti/GaAs diodes show this effect only slightly. Two variations of the admittance technique were tried. The first variation, which looked for conductance variations as a function of frequency, was not successful, presumably due to the relatively large values of conductance already present in these structures making small variations impossible to see. The second variation looked at forward bias capacitance using a specially modified bridge circuit which enabled the large conductance signal to be nulled out. In order to obtain the interface-state densities from the forward bias capacitance data several assumptions had to be made so that existing theory could be applied. The first of these was that the responsible species are donor interface-states which reside in the upper half of the bandgap. N91-12023# GRA Messerschmitt-Boelkow-Blohm G.m.b.H., Ottobrunn INTERFERENCES [DIE STROMEINSPEISUNG AUF great deal of the interference activity. The drawbacks of the classical methods being outlined, a new process for the simulation of external high frequency disturbances was developed: the Bulk Current Injection Test (BCIT). Its principles are reported, it is shown that for the determination of an improvement factor the method is very useful, as well as for relative measurements. But the BCIT method takes a lot of time, even using computers. ESA N91-12024# Technische Univ., Delft (Netherlands). FUNDAMENTAL ASPECTS AND DESIGN OF FM UPCONVERSION RECEIVER FRONT-END WITH ON-CHIP SAW FILTERS Ph.D. Thesis Paulus Thomas Maria vanZeijl 1990 151 p (Contract STW-DEL-44-0592) (ETN-90-98136) Avail: NTIS HC/MF A08 The characteristics of FM (Frequency Modulation) receivers, including tuned radio frequency and single conversion receivers, are described. The modeling of SAW (Surface Acoustic Wave) delay lines, SAW transversal filters and the behavior of SAW resonators and SAW resonator filters are discussed. The design of the FM upconversion receiver front end is described and a new class of balanced dual loop amplifiers is discussed. ESA N91-12025# Helsinki Univ. of Technology, Espoo (Finland). Radio Lab. MILLIMETER-WAVE SCHOTTKY MIXER RECEIVERS FOR APPLICATIONS IN ANTENNA MEASUREMENTS AND RADIO ASTRONOMY Ph.D. Thesis Arto Olavi Lehto Jun. 1990 21 p Sponsored by Jenny ja Antti Wihurin Rahasto and Tekniikan Edistaermissaatio (Rept-S-185; ISBN-951-22-0271-9; ISSN-0355-5712; ETN-90-98230) Avail: NTIS HC/MF A03 Whisker contacted Schottky mixer receivers and some of their applications are addressed. As an introduction to more detailed study, for which references are given, the noise sources and construction of Schottky mixers and receivers are explained. A summary of the other studies is given: an embedding network for a single ended whisker contacted millimeter wave mixer; measurement of the amplitude and phase patterns of two corrugated feed horns; a method of measuring absorption losses and beam efficiencies of large reflector antennas at millimeter wave frequencies; the construction of of 40 to 50 GHz cooled Schottky mixer front end; a 75 to 95 GHz cooled Schottky mixer receiver. ESA 34 FLUID MECHANICS AND HEAT TRANSFER Includes boundary layers; hydrodynamics; fluidics; mass transfer; and ablation cooling. For related information see also 02 Aerodynamics and 77 Thermodynamics and Statistical Physics. N91-12026 Clarkson Univ., Potsdam, NY. A NUMERICAL STUDY OF TURBULENT PIPE FLOW Ph.D. David Masser Irwin 1990 106 p Avail: Univ. Microfilms Order No. DA9025574 Turbulence in pressure driven pipe flow is numerically simulated from the three-dimensional incompressible Navier-Stokes equations written in rotation form. The velocity and pressure-head fields are represented by truncated Fourier series azimuthally and axially and by Chebyshev series in the radial direction. A fractional time stepping method separates the equations into three parts which evaluate the cross-product, pressure, and viscous termis separately by determining intermediate fields and calculating the correction due to each. Pseudo-spectral operations and fast Fourier transforms are incorporated to evaluate the cross-product. Eigenvector/eigenvalue solutions of the Poisson-like equations involved in the pressure and viscous parts are computed in a pre-processing code and coupled with streamlined matrix multiplications used in production runs. These methods and techniques provide for accuracy and efficiency. Various 2-D and 3-D linear tests of the simulation have conformed to calculated analytical results. Demonstrations of transitions to turbulence by examining 3-D growth in the presence of a 2-D finite amplitude perturbation were in agreement with previous published numerical computations of a similar nature. The calculations have determined that various statistical properties of the generated velocity fields compare favorably with available experimental data for fully developed turbulent flow. Dissert. Abstr. VISUALIZATION AND VELOCITY MEASUREMENT OF Avail: Univ. Microfilms Order No. DA9023160 Modeling of internal flow fields with hot, compressible fluids and sometimes combustion using cold flow techniques is discussed. The flow in a gas generator was modeled using cold air. The experimental set up was designed and fabricated to simulate the unsteady flow with different configurations of inlet tubes. Tests were run for flow visualization and measurement of axial velocity at different frequencies ranging from 5 to 12 Hz. Flow visualization showed that the incoming flow was a complex jet flow confined to a cylindrical enclosure, while the outgoing flow resembled the venting of a pressurized vessel. The pictures show a complex flow pattern due to the angling of the jet towards the wall for the bent tube configurations and straightened flows with straight tube and other configurations with straighteners. Velocity measurements were made at an inlet Re of 8.1 x 10(exp 4) based on maximum velocity and inlet diameter. Phase averaged mean velocities were observed to be well defined during charging and diminished during venting inside the cylinder. For the straight tube inlet comparison with a steady flow measurement of sudden expansion flow showed a qualitative similarity of the mean axial velocity distribution and centerline velocity decay during the charging phases. For the bent tube inlet case the contour plots showed the flow tendency towards the wall. Two cells were seen in the contours for the 8 and 12 Hz cases. The deviation of the point of occurrence of maximum velocity in a radial profile was found to be about 6.5 degrees. Entrance velocity profiles showed symmetry for the straight tube inlet but were skewed for the bent tube inlet. Contour plots of the phase averaged axial turbulence intensity for bent tube cases showed higher values in the core and near the wall in the region of impingement. Axial turbulence intensity measured for the straight tube case showed features as observed in an axisymmetric sudden expansion flow. Dissert. Abstr. N91-12028 Michigan Univ., Ann Arbor. Avail: Univ. Microfilms Order No. DA9023631 A theoretical and experimental study of the dense-spray region of pressure-atomized nonevaporating sprays is described, emphasizing flows in the wind-induced and atomization breakup regimes. Large scale water jets were injected vertically downward into still air at atmospheric pressure. Injectors producing fully-developed turbulent pipe flow and slug flow exit conditions were considered. The applicability of the locally-homogeneous flow (LHF) approximation to predict flow properties in the near-injector region was also evaluated. It was found that mixing was strongly influenced by the degree of flow development at the injector exit as well as the breakup regime. The structure of the flow in the atomization breakup regime consisted of a liquid core surrounded by a drop-containing mixing layer with the inner portion of the mixing layer containing large irregularly-shaped liquid elements and drops. Even though the mean liquid volume fraction was high near the axis, the gas-containing region is relatively dilute at each instant, suggesting that drop formation and secondary drop breakup dominate the process in the mixing layer-not drop collisions. The velocities of large drops are generally much larger than small drops and the gas throughout the mixing layer providing direct evidence of significant separated flow effects. Conditions required for secondary drop breakup were found to exist mainly near the liquid surface. The LHF analysis gave encouraging predictions of dense-spray properties in the near-injector region for atomization breakup, including representation of flow effects wherever liquid volume fractions were greater than approximately 0.2. It was effective at high liquid volume fractions. The LHF analysis underestimates the velocity of the larger drops in the mixing layer and overestimates the rate of development of the spray in the more dilute spray regions. Dissert. Abstr. N91-12029 Illinois Univ. at Urbana-Champaign, Savoy. Experiments were conducted using pressure measurements, schlieren photography, and velocity measurements with a two-component laser Doppler velocimeter (LDV) system. These diagnostic systems were developed for use with the mixing layer wind-tunnel facility. Many conditions were evaluated, and seven mixing layer cases were fully examined, with relative Mach numbers ranging from 0.40 to 1.97. The spatial development and similarity of the mixing layers were examined, as well as the entrainment process and the effects of particle dynamics. Analyses to predict the mean density profiles, mean transverse velocity profiles, shape of the kinematic Reynolds stress profiles, and entrainment mass fraction of a fully developed, compressible mixing layer were also developed. From the schlieren photographs, no organized, large-scale structures were observed to dominate the mixing layers. The development of the mixing layers required a Reynolds number (based on the freestream velocity difference and local mixing layer thickness) on the order of 10(exp 5). In the fully developed regions of the mixing layers, it was found that transverse turbulence intensities and normalized kinematic Reynolds stresses decreased with increasing relative Mach number like the decreases measured in normalized growth rate, while the streamwise turbulence intensities and kinematic Reynolds stress correlation coefficients remained approximately constant. By examining the LDV velocity measurements from particles with different response characteristics, it was shown that particle dynamics effects were not a problem with these measurements. Also, by measuring velocity profiles on both sides of the wind-tunnel midplane, it was found that the flow fields were reasonably two-dimensional. Dissert. Abstr. Avail: Univ. Microfilms Order No. DA9023678 The interaction of two-dimensional vortex flows with a free surface and the interaction of surface waves with a vortex generated current are studied numerically using a combined vortex/boundary integral technique. The interaction is controlled by a Froude number, based on the vortical motion, and the geometric parameters describe the initial vortex configuration. The head-on collision of a vortex pair with the free surface and the large amplitude Kelvin-Helmholtz instability of a submerged shear-layer below the free surface are studied in considerable detail. Here the vorticity is modeled as point vortices, vortex sheets, and finite area vortex regions. In the large Froude number limit, the surface motion follows the vortical flow, but depends only weakly on the actual value of the Froude number. For low Froude numbers, the free surface remains almost flat, and the disturbances caused by the vortical flow decrease rapidly with Froude number. The physical circumstance, in which the surface wave is generated by a time dependent pressure load, is studied numerically with the multipole expansion technique (MET). For the strong interaction, which depends on both Froude number and the initial depth of |