Search

Experimental research has been conducted to determine the influence of a free-stream turbulence on zero pressure gradient, fully turbulent boundary layer flow. Convective heat transfer coefficients, boundary layer mean velocity and temperature profile data, and wall static pressure distribution data were obtained for five flow conditions of constant free-stream velocity and free-stream turbulence intensities ranging from approximately 1/4% to 7%. Free-stream multi-component turbulence intensity, longitudinal integral scale, and spectral distributions were obtained for the various turbulence levels. Mean velocity and temperature profile data for the individual boundary layer traverses are presented in this report. (Author), See also Volume 2, AD-A101 096.

Experimental research has been conducted to examine the combined effects of free-stream turbulence and favorable pressure gradients on flat-wall transitional boundary layers. Convective heat transfer coefficients, boundary layer mean velocity and temperature profile data, and wall static pressure distribution data were obtained for four combinations of free-stream turbulence intensity and favorable pressure gradient. Free-stream multi-component turbulence intensity, longitudinal integral scale, and spectral distributions were obtained for the various test cases. Mean velocity and temperature profile data for the individual boundary layer traverses are presented in this report. (Author)

Experimental research has been conducted to determine the influence of free-stream turbulence on zero pressure gradient fully turbulent boundary layer flow. During this period convective heat transfer coefficients, boundary layer mean velocity and temperature profile data, and wall static pressure distribution data were obtained for five flow conditions of constant free-stream velocity and free-stream turbulence intensities ranging from approximately 1/4% to 7%. Free-stream multi-component turbulence intensity, longitudinal integral scale, and spectral distributions were obtained for the various turbulence levels. The test results with 1/4% free-stream turbulence indicate that these data are in excellent agreement with classic two-dimensional low free-stream turbulence, turbulent boundary layer correlations, thus establishing the absolute accuracy of the experiment. The data obtained for the test cases with higher free-stream turbulence indicate that the turbulence has a significant effect on turbulent boundary layer skin friction and heat transfer. It has been shown that these effects are a function of the free-stream turbulence intensity, the turbulence length scale, and the boundary layer momentum thickness Reynolds number. Suggested correlations for the influence of free-stream turbulence on skin friction, heat transfer, and Reynolds analogy factor are given. Also during this period, a boundary layer prediction method has been assessed as to its ability to predict free-stream turbulence effects on flat plate heating and skin friction. Comparisons with the experimental data obtained here showed that adequate predictions can be made for this case with an existing turbulence model. (Author)

This report summarizes the work completed during a preliminary investigation of the applicability of the LAYERGLAZE (trade name) process to the fabrication of copper rotating bands on titanium artillery shell bodies. Two approaches were taken, one involving direct deposition of Cu onto the Ti alloy bodies, and the second involving the use of a compatible interlayer between the Cu and the Ti. Attempts to bound pure Cu to Ti produced deposits of varying densities but none adhered soundly to the Ti substrate. Of the interlayers investigated, Al had the best characteristics, but often melted and mixed when the higher melting Cu was applied to it. Copper alloys were then investigated, in an attempt to improve coupling, increase deposit density, and reduce substrate temperature. Cu-1% Pb appeared to be well-suited for LAYERGLAZE band fabrication, but the structure did contain some cracks. Although additional development work is necessary, significant progress was achieved, and the process exhibits real potential as an alternative approach for depositing rotating bands onto alloy shell casings.

Nitric oxide (NO was measured in the exhaust of three different combustion systems by in situ ultraviolet absorption and by chemiluminescent analysis after gas sampling with several probe designs. The three combustion systems were: (1) a flat flame burner fueled with CH4/N2/O2; (2) a research swirl burner fueled with C3H8/air; and, (3) a modified FT12 combustor operated on Jet A/air. Each combustion system was run at several different conditions so that probe and optical measurements could be obtained over a wide range of exhaust environments encompassing products from lean, stoichiometric, and rich flames, laminar to turbulent flows, and temperatures at centerline from 600 K to 1800 K. The results obtained with the metallic, water-cooled probes of different designs (all expansion-type) agreed with the optical results to within 25%. Some small losses of NO (10-15%) were observed in a lean methane flame at 1800 K with an uncooled stainless steel probe, but under fuel-rich conditions up to 80% NO destruction was observed. Experimental facilities are described, previous results are discussed, and a summary of the major findings of this study is given. (Author), Also available as Rept. nos. NAPC-PE-39C and EPA-460/3-80-015. See also Volume 2, AD-A097 545.

Contributions are made to interacting viscous flow theory for symmetric separated flow past blunt airfoil trailing edges. A theoretical and computational technique for predicting turbulent, high Reynolds number, separation and near wake flow is provided. A detailed assessment of a viscous/inviscid iteration and an inverse boundary layer solution procedure is provided through a systematic series of test examples. These involve subsonic laminar and turbulent flows past flat plate and thickened flat plate (blunt trailing-edged) airfoils. In addition, a general approach is outlined for incorporating local strong interaction solutions into an overall viscous solution procedure for isolated airfoils and cascades. (Author)

Experimental facilities used in studying the performance of probes and sampling systems for measuring NO are described. A critical review of the literature on probe measurements of NO sub x is given with emphasis on reported results indicating that probes may perturb the total concentration of NO sub x in a flame. Also, sample line and chemiluminescent analyzer phenomena are reviewed. A model of probe aerodynamics including heat transfer is presented. Kinetics of NO loss are examined and quenching criteria for measuring nitric oxide in flames are given. Sampling probes are described that were designed to preserve NO and are suitable for measurements on small and large combustors. Probes were designed to cool the gases both convectively and aerodynamically. Performance of these probes is compared with model predictions. Concentrations of nitric oxide were measured using several probes for each of three flame environments. The values measured with each probe are compared and related to seed levels of NO. In addition, concentration profiles required to compare probe measurements with optical measurements are provided. (Author), Also available as Rept. nos. NAPC-PE-38C and EPA-460/3-80-014. See also Volume 3, AD-A097 607.

An analysis has been performed on unsteady pressure data measured in the leading edge region of a subsonic cascade oscillating in pitch. The objectives of the investigation were to determine the sensitivity of the freestream velocity disturbance at the inlet plane to variations in interblade phase angle, to compare the behavior of the freestream velocity disturbance as a function of interblade phase angle with that of both the inlet area oscillation and the measured moment response of the cascade, and to determine the influence of the gap-to-chord ratio on the blade loading near the leading edge. In this study, it was found that the behavior of the unsteady freestream flow entering an oscillating subsonic cascade is strongly influenced by the interblade phase angle. A more important finding was that the primary trends in the freestream disturbance match those of both the inlet area oscillation and the moment response of the cascade. Because of this correlation, a hypothesis is presented that describes how the interaction between the inlet area oscillation and the freestream flow might be relevant to the behavior of the cascade moment response and thus to the stability of the cascade motion. Analytical predictions based on this hypothesis are in qualitative agreement with the measured trends. The effect of gap-to-chord ratio on the unsteady blade loading was studied and found to be substantial in two ways: it alters the degree of influence of leading edge dynamic stall on the chordwise load distribution, and it alters the degree of influence of the interblade phase angle on the amplitude of the load response. (Author)

The aims of the present program were to determine both experimentally and analytically the cooling rates experienced by material during laser melting in the deep penetration and shallow penetration metling modes, to determine the microstructures and degrees of microsegregation associated with these cooling rates, to correlate the structure and segregation with weld cracking behavior, and to eventually determine the mechanism involved in weld cracking. A literature survey was performed and two new mathematical models were developed, a two dimensional model for deep penetration welding and a one dimensional model for shallow penetration surface melting. The results of both model calculations were eventually verified by experimental tests and comparisons of results, with excellent agreement being obtained for the deep penetration model and adequate agreement being obtained for the shallow penetration model. In addition a series of three special Ni-Al-Mo alloys was weld-crack tested to determine effects of microstructure on cracking behavior.

Ribbon reinforced composites are attractive as structural materials since relatively isotropic in-plane properties are achieved if the ribbon aspect ratio is sufficiently high. Recent advances in the production of amorphous metal ribbons on a continuos basis have resulted in a ribbon material which exhibits unique properties and is potentially low cost. Commercially available amorphous metal ribbon materials were used in this investigation to reinforce resin matrices. Typical ribbon properties were a density of 8.02 gm/cc, tensile strengths in the range of 2550-2900 MPa, tensile modulus of 165 GPa, and aspect ratios, w/t, up to 380. The longitudinal properties of the composites developed were in agreement with the simple rule of mixtures. The transverse composite properties were dependent on ribbon aspect ratio and the ribbon stacking pattern. The composites developed were compared on a specific property basis to currently used quasi-isotropic graphite/epoxy. It was demonstrated that amorphous metal ribbon reinforced resins are competitive with quasi-isotropic graphite/epoxy on a specific strength basis. However, amorphous metal ribbon reinforced resins are not competitive on a specific modulus basis. Due to the relatively invariant nature of the specific modulus of amorphous metal ribbon materials, no improvement is expected in this regard. It was concluded that these composites would not be in a competitive position for most high performance composite applications, particularly in the aerospace industry.

This report describes the results of an exploratory development program to investigate various system and component aspects of a pulsed FIR heterodyning laser radar. During the course of this developmental program a number of interrelated subsystems required for a heterodyning FIR radar were investigated. The work focused on optically pumped FIR lasers which utilize a CO2 laser pump source. Pulsed FIR and CO2 pump lasers are required for the primary radar transmitter, and CW FIR and CO2 lasers are needed for the laser radar local oscillator. The transversely excited (TE) CO2 laser which serves as the pulsed FIR laser pump is itself a central component of the radar system and the performance characteristics of a recirculating, high pulse repetition frequency (PRF) TE CO2 as a FIR pump were experimentally examined. A number of different pulsed FIR laser resonators were investigated with the objective of minimizing system size. The requirement for a stable FIR local oscillator imposes certain constraints on the CW CO2 pump source and a stable single mode CO2 pump source has been designed and characterized. A sealed-off CW FIR laser with an Invar stabilized cavity was constructed for use as a local oscillator source. A number of detectors for use at submillimeter wavelength were examined and pulsed and CW detection with two micron Schottky diodes were performed. Because of the extreme sensitivity of the CW FIR laser and pump source to temperature fluctuation a technique for stabilizing the CO2 utilizing the photoacoustic effect in the FIR medium was examined and found to be very promising.

This report describes the essential features of electron-beam controlled discharges of the type common to rare gas halide and mercury-halide lasers. Particular attention is directed toward the XeCl (308 nm) laser using HCl as the halogen donor, and the HgBr(502 nm) laser using HgBr2. The basic kinetic processes controlling the behavior of these lasers are discussed, with special emphasis on electron collisions with HCl and HgBr2, respectively. (Author)

Species detectivity is limited in conventional CARS with aligned laser polarizations by the presence of a background, nonresonant electronic contribution to the CARS signal. At low species densities, a modulated CARS spectrum results which can be used to perform concentration measurements from the spectral shape, a unique feature of CARS. When the species concentration becomes very low, the 'signal' from the specie of interest, i.e. the modulation, essentially disappears into the nonresonant background and the specie is no longer detectable. By proper orientation of the polarizations of the pump and Stokes laser fields and the CARS signal polarization analyzer, the nonresonant background can be suppressed leaving the pure CARS spectrum of the specie being probed. Unfortunately, a factor of sixteen reduction in resonant mode signal accompanies the elimination of the background. In this report, investigations of background suppression from CO in flames are described using polarization sensitive CARS with spectrally broadband Stokes waves. The studies were conducted in flat CO-air diffusion flames employing nonplanar, crossed-beam phase matching, i.e. folded BOXCARS. The relative concentration sensitivity of CARS with and without background suppression was compared at both high and low temperatures.

This report describes the results of a two-year investigation to develop a modeling strategy to predict the final shape of components manufactured by the Hot Isostatic Pressing (HIP) process. When a powder metal is subjected to the high pressures at high temperatures there is a permanent reduction in the macroscopic volume by approximately 30 percent. The volume reduction, which is not included in classical creep and plasticity theories, necessitates the modification of the classical theories. The major part of this report concerns development of a constitutive model for powder metals. A model for powder metals subjected to an external hydrostatic pressure has been developed and compares well with experimental results. The exact form oof the theory for a multi-axial stress state was determined by a series of mechanical tests. The constitutive model has been incorporated into the MACR nonlinear finite element computer program and analysis of an independent experiment using the modified code has been performed and verified the accuracy of the theory. Once the theory had been verified it was applied to the HIP of a disk and a sensitivity study was performed to determine an efficient modeling strategy. The results of the program are discussed in detail. (Author)

This study sought to formulate a working theory with which the performance of two-phase nozzles can be analyzed parametrically over a wide range of flow conditions and fluid properties. As a basis for formulating the theory, the technical status of two-phase nozzles was assessed, the governing flow parameters were identified, and the flow domains classified with the previously available theory. A working theory of two-phase nozzles suitable for parametric analyzes and optimization was then formulated, incorporating the significant interphase effects of velocity slip, droplet breakup, and droplet heat transfer. Liquid/gas combinations for two-phase nozzles, and suitable combinations for marine propulsion and other selected applications were discussed. Single-phase (or homogeneous) theory was reviewed from the viewpoints of energy conversion in a nozzle and was compared with models applicable to separate phases in terms of their flow domains. Finally, a working theory of two-phase nozzles was formulated based on the governing flow parameters and on six basic equations that described the two-phase flow. From this theory, a computer model was developed which includes the significant interphase effects of velocity slip, droplet beakup, and heat transfer and which is suitable for parametric analysis and optimization. Preliminary parametric calculations based on use of this model were compared with results from previous two-phase nozzle models and with available experimental data.

This executive summary report presents the major results of a comprehensive study program to evaluate the technological and economic feasibility of utilizing open- and closed-cycle gas turbines to provide lightweight propulsion power for future Navy surface ship applications. The level of technology considered in this study is that judged by the Contractor to be available during the 1990's. Naval ship types which could benefit from implementing lightweight propulsion systems were identified and characterized, and applicable propulsion system configurations and power levels compatible with component capabilities were selected. Extensive parametric analyses were made of the performance, weight, and cost characteristics for the baseline propulsion systems, from which the total propulsion-system-plus-fuel weight was then estimated. The payload capabilities and endurance limitations with open- and closed-cycle gas turbine propulsion systems for selected ship types are presented., See also AD-A040 032 and AD-A062 746.

Total and partial cross sections for mercuric bromide (HgBr2) electron attachment and ionization reactions of importance in laser discharges have been measured in electron beam experiments. Corresponding rate coefficients have been determined independently in a pulsed, electron swarm apparatus. These studies have determined that the electron impact ionization cross section of HgBr2 is very large, increasing from a threshold at 10.6 eV to a value of 2 x 10 to the minus 15th power sq cm at 70 eV, with HgBr2 ion being the principal ion formed at low energies. Beam measurements also have revealed only one electron loss process, dissociative attachment. This attachment reaction which produces Br ion has a threshold energy of 3.1 eV and a peak cross section of 1 x 10 to the minus 17th power sq cm at 3.7 eV. From swarm measurements, an attachment rate coefficient of 1.5 x 10 to the minus 10th power cu cm/sec at a few electron volts average electron energy has been determined. This value is consistent with cross section values determined from beam measurements and with values inferred from kinetic modeling studies of electron beam controlled HgBr2 laser discharges.

A demonstration of the use of microprocessors for implementing linear quadratic Gaussian (LQG) control was conducted. The demonstration consisted of simulating linear system dynamics on an analog computer and implementing LQG control and estimation dynamics on a microprocessor. Two cases were studied -- a single input second order system and a four input fifth order system. The second order system was controlled using an Intel 8080 8-bit microprocessor and fifth order system was controlled using a 16-bit Digital Equipment Corp LSI 11/2 microprocessor. Key requirements addressed include microprocessor requirements for: (1) word size; (2) computational capability, including arithmetic and input/output operations; and (3) memory requirements. The requirements were compared against predicted requirements made using previously developed analytic techniques. The implementation involved developing general purpose algorithms required for implementing LQG control and estimation. These algorithms consisted of matrix/vector multiplication, vector addition and input/output service routines. The same algorithms were employed in both the second order and the fifth order demonstration.

A conceptual design of a diagnostic system using coherent anti-Stokes Raman Spectroscopy (CARS) to determine time resolved temperature and species concentrations in the Army Mechanics and Materials Research Center ballistic compressor has been formulated. Performance calculations have been carried out for hydrogen which is an ideal diagnostic species for this application. Results are also presented of computer code calculations of collision narrowed spectra of nitrogen. Several pump laser systems were compared, and a high repetition frequency Nd:YAG laser was determined to be superior. Population saturation, stimulated Raman gain, and optical breakdown are not expected to pose severe limitations on laser intensity or measurement signal-to-noise ratios. Consideration is given to problems pertinent to operation of the ballistic compressor including compressor recoil, window deflections and generation of CARS from the test section windows. Finally, the possibility of using pulsed stimulated Raman gain spectroscopy (SRGS) as an alternative to CARS at the high densities typical of the ballistic compressor has been identified. SRGS appears to offer adequate measurement ability and possesses certain advantages over CARS. (Author)

This research develops techniques and computational schemes for a priori prediction of the kinetics and the product distribution of atomdiatom reactions. This involves (1) improved calculation techniques for potential energy surfaces for the reactants and products; (2) quantum and classical mechanical calculations of the dynamics of the reaction; and (3) detailed calculations for systems with high potential as electronic transition lasers.

The operation of a N2(+) ion waveguide laser (4278 A) excited by charge transfer from He2(+) created in a capacitively coupled self-sustained discharge is described. In addition, the results of an experimental diagnostic program and a computer modeling program, conducted to explain the general operations of the discharge pumped He-N2 system, are presented. The results of this program indicate upper laser level electron quenching and lower level pumping may be seriously limiting the achievable laser efficiency of this system.

For the equivalent compacted ring resonator, beam injection can help maintain bare resonator power coupling at the unperturbed value. For small misalignments, injection improves bare beam quality, but it does not improve quality degradation due to astigmatism. For the annular ring resonator, beam injection maintains both bare resonator power coupling and beam quality at the unperturbed values for very small misalignments. And, as expected, this resonator is much more sensitive to misalignment than its equivalent compacted resonator. Additionally, the reverse wave analysis has shown the unstable ring/linear stable MO injection configuration to offer satisfactory reverse wave suppression. Based on bare resonator analyses injection locking shows promise. To determine its potential, loaded resonator analyses for high-power devices must be performed. mode discrimination, possible parasitics, sensitivity to misalignment,and scale-ability to high power are potential problems that need in-depth investigation. Additionally, the technical issues associated with cavity length equalization to less than lambda; matching of the MO and HEL spectra; and the stability, mode matching and power requirements for the MO need to be addressed experimentally so the real hardware requirements can be established., Prepared in cooperation with Pratt and Whitney Aircraft, Government Products Div.

Coating compositions were developed to protect the directionally solidified gamma/gamma prime-alpha alloy (Ni-31w/oMo-6.2w/oAl) during cyclic oxidation and sulfidation in burner rig tests. The coating composition Ni-13.8w/oCo-15.8w/oCr-8.2w/oAl-16.9w/oPt-0.1w/oY was specifically identified in this study. For the family of alloys represented by gamma prime-alpha, the addition of 17 percent platinum to the NiCoCrAlY composition was insufficient to prevent cracking of the coating. Coatings containing 37 w/o platinum did not crack in the 2100 F cyclic oxidation tests. Chromium used as an intervening layer between gamma prime-alpha and the overlay coating reacted with the substrate to form brittle phases which cracked during thermal cycling. The substitution of nickel for chromium as an intervening layer resulted in void formation (Kirkendall porosity) leading to eventual coating delamination. (Author)

A general theory is presented for high Reynolds number steady boundary layer separation, with application here to the near wake region of blunt subsonic trailing edges. The theory is based on the concept of a thin viscous layer displacing and thus interacting with an inviscid flow. The governing equations are derived in a shear layer oriented coordinate system and subsequently simplified through use of Prandtl's Transposition Theorem. These are put in final form through introduction of turbulent Levy Lees type variables to minimize growth of the viscous layer in the computational domain. Solutions were first obtained for asymptotically large Reynolds number through use of Triple Deck scaling laws. A numerical algorithm was developed using an inverse viscous layer approach coupled with a direct inviscid flow solver. The method was found to be fast, stable and accurate through a comparative assessment with previously published flat plate trailing edge solutions. Blunt trailing edge separated flows were then studied providing the first such steady, high Reynolds number solutions ever produced. A numerical algorithm was also written for solving the finite Reynolds number form of the governing equations. The approach was a direct extension of the Triple Deck Solver. Good agreement has been achieved with recently published flat plate trailing edge solutions at a Reynolds number of 100,000. (Author)

The detailed nature of the three-dimensional flow downstream of an isolated compressor rotor has been studied experimentally in a large scale rotating rig. Radial-circumferential arrays of pneumatic data have been acquired in the rotating frame for reference at values of (C sub x/U sub m) of 0.65, 0.75, 0.85 and 0.95 at planes 10, 30, 50 and 110% of blade axial chord aft of the rotor. The data consists of the total and static pressures and the yaw and pitch angles. From this data the relative velocity vector has been completely defined. From the data it can be seen that the radial flow in the wake increases dramatically as the rotor loading is increased (by going to lower values of (C sub x/U sub m)). At high values of (C sub x/U sub m) low total pressure fluid accumulates at the hub causing corner stall. As (C sub x/U sub m) is reduced the radial flow carries the low total pressure fluid out to the tip, causing the hub corner stall to be inhibited, i.e. reduced in its streamwise extent aft of the blade trailing edge. By this process non-strip theory effects are seen to have a strong impact on the nature of rotor performance. (Author)

Collisional transfer of population between rotational levels of DF was measured using an i r double resonance method. The data were used to calculate rotational rates. Pressure broadening of DF absorption was determined for the fundamental lines P6 through P18. Total molecular collisional rates were obtained from the absorption data and compared with the total of individual collisional rates for velocity, vibrational and rotational relaxation. The results were compared with two kinetic rate models. Observation were also made of rotational lasing and vibration to rotation transfer in DF. (Author)

This report describes the analytical and experimental results of a study to determine the feasibility of using molecular seeds to couple CO2 laser radiation and CO laser radiation to the working fluid, H2, of a high performance rocket thruster. The gas mixtures studied experimentally were H2/H2O, H2/D2O, and H2/NH3 at CO2 laser wavelengths. The promise of CO as an absorber of CO laser radiation was established analytically. To attain the high temperatures, 1000-6000K, desired, these mixtures were placed in a cell positioned in a diagnostic apparatus which determined the spatial variation of the mixture absorption coefficient and temperature about the axis of a hot, dense plasma sustained by focuesed high power CW CO2 laser radiation. The measurements on these mixtures indicated their performance as couplers to exceed expectations created by the analytical study. In addition, it was found that in the absence of plasma, the temperature of the mixtures could be raised to temperatures as high as 2000-4000K simply by absorption by the mixture of laser energy at the high power beam focal spot to create the observed laser sustained flame. Further work is needed to determine the conditions under which this flame can be sustained in the laser energized rocket thruster. (Author)

Techniques are described for prolonging the duration of stable e-beam controlled excimer laser discharges, including: temporal tailoring of either the discharge voltage or the ionization source, and kinetics modification by way of additives. Theoretical and experimental results are presented for KrF* discharges. XeCl(B) formation processes are examined for conditions typical of a discharge excited laser using HCl as the chlorine donor. It is shown that vibrational excitation of HCl followed by dissociative attachment is a primary step in the reaction sequence resulting in cl(=). XeCl(B) formation is the result of a three-body Xe(+) - Cl(=) recombination reaction. Experimental results are presented which demonstrate efficient (about 2%) XeCl laser operation in an e-beam assisted discharge in which over 75% of the energy was deposited by the discharge. (Author)

Glass-ceramic composites reinforced with SiC fibers have been fabricated that exhibit good strength properties from RT to 1000 C. In particular, a lithium-aluminum-silicate (LAS) matrix reinforced with SiC yarn has been shown to possess bend strengths of over 690 MPa (100 ksi) from RT to 800 C. However, the TiO2 nucleating agent present in the usual LAS materials can cause matrix crystallization during hot-pressing and fiber-matrix interactions, making the fabrication of consistently high strength composites extremely difficult. The substitution of ZrO2 for the TiO2 appears to have eliminated this problem. Composites fabricated from SiC monofilament with a magnesium-aluminum-silicate (MAS) matrix have also exhibited high strengths from RT to 1000 C. Again, however, consistent fabrication of reproducibly high strength composites has been difficult due to lack of wetting and bonding between the SiC fibers and the matrix. (Author)

This report presents the technical characteristics of alternative closed-cycle gas turbine propulsion systems incorporating various electrical and mechanical transmissions, and reviews the critical technologies and development requirements for these propulsion systems as part of a comprehensive study program which assesses the technological and economic feasibility of utilizing open- and closed-cycle gas turbines to provide lightweight propulsion power for future Navy capital ship applications. The level of technology considered is that judged by the Contractor to be available for use during the 1990's.

This report describes the first year of progress of a two year investigation to develop a modeling strategy to predict the final shape of components manufactured by the Hot Isostatic Pressing (HIP) process. When a powder metal is subjected to the high pressures at high temperatures there is a permanent reduction in the macroscopic volume by approximately 30 percent. The volume reduction, which is not included in classical creep and plasticity theories, necessitates the modification of the classical theories. The major part of this report concerns the development of a constitutive model for powder metals. A model for powder Metals subjected to an external hydrostatic pressure has been developed and compares well with experimental results. The exact form of the theory depends on the results of mechanical tests presently being performed. The constitutive model is being incorporated into the MARC Analysis Research Corporation (MARC) nonlinear finite element computer program and experiments analyzes using the modified code will be performed to verify the accuracy of the theory. Once the theory has been verified it will be applied to the HIP of a disk where a sensitivity study will determine an efficient modeling strategy. The results and future plans of the program are discussed in detail. (Author)

The application of reaction sintered Si3N4 energy absorbing surface layers to hot-pressed Si3N4 was investigated. The surface layer was formed by in-place nitridation of silicon powder. It was found that reaction sintered Si3N4 layers of 1 mm thickness, fabricated from either -100, +200, -200, or -325 mesh Si powder and nitrided in 96% N2/4% H2 so that approximately 20-25 vol % unnitrided Si remained in the layer, resulted in a sevenfold iNCrease in ballistic impact resistance of a 0.64 cm thick hot-pressed Si3N4 substrate from RT to 1370 C. Both Nc-132 Si3N4, with mgO additive, and NCX-34 Si3N4, with Y2O3 additive, were evaluated as substrate material. The finer grain size -200 and -325 mesh nitrided Si layers were found to be preferable for their smoothness and relatively high density. It was found that nitriding in N2/H2 mixtures, rather than pure N2, resulted in a microstructure that did not substantially degrade the strength of the hot-pressed Si3N4 substrate.

An experimental investigation was conducted to determine the airflow characteristics in the vicinity of a model helicopter to determine the aerodynamic interference at representative rocket trajectories and wind sensor locations of an Army AH-1G helicopter operating in low speed forward flight. Laser velocimetry and flow visualization techniques were applied in a wind tunnel to provide flow velocity and wake geometry data for use in a rocket aeroballistics analysis, correlation with theory, and for a systematic determination of the influence of the helicopter components (rotor, fuselage, and wing) and significant operating condition parameters (flight speed, flight direction, ground effect, gross weight, and aircraft/rocket attitude) on the airflow influencing rocket trajectories. In addition, the flow velocities at potential locations for a fire control wind sensor mounted on the helicopter were determined to evaluate potential wind sensor locations for accurately measuring flow velocities at the low flight speeds required in accordance with current terrain flying/rocket firing tactics.

The objective of this research investigation is to develop and evaluate analytical procedures for establishing microprocessor requirements for implementing modern control logic. Key issues in microprocessor implementation of modern control logic include: (1) accuracy, (2) computational capability, and (3) memory requirements. These requirements must be established for the interface as well as for the microprocessor control code. (Author)

This paper reports the results of an analysis of basic kinetic and plasma processes in fast-pulse (about 100 nsec) electric discharges containing mixtures of the mercuric-bromide molecule HgBr2, and N2 in a Ne background. Formation of the laser molecule HgBr (B 2 Sigma(+)) is shown to occur as result of dissociative excitation transfer following quenching of N2 (A 3 Sigma(+) sub u) by HgBr2. (Author)

Experimental research has been conducted to determine the influence of free-stream turbulence on zero pressure gradient, incompressible, fully turbulent boundary layer flow. During this period convective heat transfer coefficients, boundary layer mean velocity and temperature profile and wall static pressure distribution data were obtained for two flow conditions of constant free-stream velocity and low free-stream turbulence intensity and for one flow condition of constant free-stream velocity and higher free-stream turbulence. Documentation of the free-stream turbulence for these flows is currently in progress. The conclusion reached from the low free-stream turbulence test results is that these data are in excellent agreement with classic two-dimensional, low free-stream turbulence, turbulent boundary layer correlations, thus establishing the absolute accuracy of the experiment. The data obtained for the higher free-stream turbulence test case indicates that free-stream turbulence does have a significant effect on fully turbulent boundary layer skin friction and heat transfer. A quantitative assessment of this influence will emerge as data is obtained for additional free-stream turbulence levels and as the turbulence distributions are documented.

Ceramic bodies having compositions in the Beta', Beta'-15R, and Beta'-0' phase fields in the Si3N4 -SiO2-Al2O3-SiO2 system which are prepared using fine grained constitutents cannot be sintered to high density, since little or no liquid forms during normal firing schedules. Beta'-X phase bodies, however, sinter to near theoretical density as a result of the presence of a liquid phase under equilibrium conditions at temperatures above about 1750 C. One such body, composition 20e/oA1, 20e/oO, was characterized in terms of room temperature and 1370 C flexural strength, 1370 C creep, and 1000 and 1400 C static oxidation behavior, and Mach 0.8 flame erosion at 1270 C. Attempts were made to supply a transient liquid to effect sintering of Beta', Beta'-15R, and Beta'-O' bodies by controlling particle size distributions of different constituent phases. There was evidence that in fact transient liquids were induced at temperatures above 1750 C, but solid phase particle bridging effects prevented gross shrinkage. It seems highly likely that the application of pressure at temperature to such formulations could yield the desired result--dense Beta' based materials that contain neither residual glass nor X phase. (Author)

This report describes an experimental and theoretical study of nonlinear tapped delay lines for use as an electronically programmable amplitude and phase weighted transversal filter. The Surface Acoustic Wave (SAW) delay line device studied here, which is called the Programmable Frequency Filter Module (PFFM), is a hybrid, integrated circuit device which utilizes SAW taped delay lines and attached microelectronic diode and resistor arrays. The PFFM is based on the UTRC PSK Diode Correlator which underwent basic development as a 128 tap Phase Shift Key (PSK) signal encoder and correlator under NESC Contract N00014-75-C-0978. While the signal processing capabilities of the single acoustic channel PSK Diode Correlator rely on discrete tap programming in 0 and 180 degree relative phase increments, the dual acoustic channel PFFM described here utilizes continuous tap programming in both amplitude and phase. Each acoustic channel of the device consists of a P diode-tapped signal processing region and two input transducer ports. The input signal and CW reference voltages are applied to the input ports at center frequency F(1) and f(2), respectively. A low power, wide bandwidth nonlinear mixing process occurs at each diode tap, and the difference frequency signals from all taps are summed at a common electrode comprising the filter output port. Rapid tap amplitude an biphase coding is achieved by controlling magnitude and polarity applied to the diode mixer located at each delay line tapping transducer.

Stabilized comb spectra generators using SAW Technology have been developed for operation at 333 MHz and higher frquencies. These signal sources and an additional CW source at 1984 MHz are being developed to provide precision signals for a rapid frequency hopping frequency synthesizer. The comb spectrum generator is based upon the Mode Locked SAW Oscillator (MLSO), a regenerative rf pulse oscillator. Stabilization, i.e., synchronization to a 1 MHz reference signal, is accomplished through injection locking of the MLSO at the fundamental rf frequency. The stabilization signal is provided by means of a Phase Locked Loop oscillator that is assembled from commercially available IC components. The first complete prototype consists of two four inch square circuit boards and consumes about five watts of power. This first prototype operating at 333 MHz generates the desired spectral lines (9 lines spaced 3 MHz apart) and the signal characteristics closely satisfies the goals of the program both for low FM noise and spurious signals. The judicious use of SAW filters was important for meeting the program goals. These SAW filters together with the IC components made possible a small compact package that can be even further reduced in size with further development. Preliminary results with the 828 MHz MLSO and the 1984 MHz CW source are also promising. (Author)

Silicon carbide and alumina fiber reinforced aluminum matrix composites were evaluated and compared by tensile testing, three point bend flexural testing, elevated temperature exposure, thermal cycling, and instrumented pendulum impact. It was found that the silicon carbide/aluminum provided the higher level of axial tensile strength while the alumina/aluminum was superior in transverse tension. The difference in transverse strength was mainly due to the poor fiber-matrix bond existent in the silicon carbide/aluminum as compared to the very high strength bond in the alumina/aluminum composite. A very large difference in composite impact performance was found to be affected by the differences in fiber and matrix properties. In addition, measured composite impact energy dissipation capability was related to specimen test geometry as well as the relative levels of composite tensile and shear strengths. Both composite materials were also shown to suffer significantly due to either thermal cycling (SiC/aluminum) or constant elevated temperature exposure (alumina/aluminum). (Author)

A research program has been conducted towards developing an analysis to predict viscous flow in a cascade blade passage. Two approaches have been investigated. The first concerns obtaining solutions to the compressible Navier-Stokes equations of motion, and the second considers the formulation and numerical solution of viscid/inviscid interaction models for localized flow regions to predict several of the viscous phenomena of importance in turbomachinery applications. In the present effort an existing computer code, based on an alternating direction, implicit time-marching solution to a general coordinate representation of the Navier-Stokes equations has been modified to be applicable to the blade passage problem. In addition, a simple external flow field, i.e., flow past a circular cylinder, has been calculated in order to gain experience in working with this code and to assess its reliability and efficiency. In the second part of the present effort an interacting boundary layer model, incorporating asymptotic triple deck concepts, has been constructed to predict the flow in the trailing edge region of an airfoil.

The application of the electrochemical compatibility test method to metals exposed to nitrogen tetroxide (N2O4) is described. It was determined that N N2O4 is not sufficiently conductive to permit electrochemical measurements to be carried out. It was necessary to add water in the concentration range of ten to eleven value percent in order to obtain compatibility data. This concentration corresponds to the high density acid phase of the N2O4 - H2O system. Above one to two value percent water a small portion of the high density acid phase is formed which can produce localized corrosion. It is suggested that corrosion in the high acid phase should be considered for further study even though the overall water concentration is quite low. (Author)

This report presents the results of conceptual design studies and reliability analyses as part of a comprehensive study program to evaluate the technological and economic feasibility of utilizing open- and closed-cycle gas turbines integrated with fossil or nuclear heat sources for providing advanced lightweight propulsion power for future Navy capital ship applications. The level of technology considered is that judged by the Contractor to be available during the 1990's., Includes Rept. no. UTRC/R77-952972-1 dated Nov 77.

Two new fibers have been introduced into the program. Both fibers are SiC with one in a large diameter(140 micron) monofilament form and the other a multifilament yarn. This second fiber has only recently been developed and is produced from an organometallic precursor. The new SiC yarn was characterized as to its surface chemistry, mechanical properties, fracture morphology, and compared with the characteristics of the large diameter monofilament. Powder metallurgical and casting techniques were explored for composite fabrication and it was found that the PM approach offered the greatest success. Lost cost, high rate processing techniques were also successfully used to produce high strength composities using both high and low bonding pressures. In the later case the matrix was densified while partially molten. Composite axial strength, transverse strength, and impact energy were measured. The SiC monofilament reinforced composites offered the highest levels of strength and impact resistance; however, the strength of the fiber-matrix bond severely limited composite transverse strength. (Author)

Synthesis procedures for nonlinear stochastic feedback control of gas turbine engines were developed and evaluated. Modern estimation and control procedures based upon separating the stochastic and deterministic aspects of the control problem were employed. The resulting closed-loop control consists of nonlinear deterministic feedback control logic designed using piecewise-linear/piecewise-optimal techniques, and an estimator designed using nonlinear filtering logic. Engine variables estimated from noise-corrupted measurements are fed back through the deterministic control logic to generate commanded inputs to the engine. Mode-switching logic was developed to provide smooth transition between small-signal regulation and large-signal transient modes of estimator/controller operation., See also Rept. no. UTRC-R77-942577-6 dated Mar 77, AD-A040 901.

The oxidation/corrosion characteristics of directionally solidified eutectics including gamma prime-gamma plus alpha alloys are sufficiently well documented to state that oxidation and hot corrosion resistant coatings will be required over a wide range of temperatures. The specific problem with respect to coating the gamma/gamma-alpha family of alloys is not simply coating chemistry. Rather the major obstacle here involves mating the anisotropic thermal expansion behavior of the substrate with that of the coating so that the latter will not crack under thermal cycling. The approach taken in this study includes minimizing the differences in thermal expansion between the protective coating and the eutectic substrate by selective alloying of the coating and increasing the strain capacity of the coating.

Single crystal AlN films were grown by sputtering on single crystal (0001) oriented sapphire substrates. The substrate size was 25.4 x 9.0 x 2.0 mm. This size was chosen as a convenient size for the fabrication of UHF SAW frequency filters. The AlN was synthesized by reactive rf diode sputtering of a pure aluminum target in an ammonia atmosphere. Growth conditions were established which yielded clear, uniform, single crystal AlN films. Films were evaluated through microscopic examinations, interference spectrophotometric thickness measurement and electron channeling and X-ray diffractometer crystallographic analysis. Films ranged from 2 micrometers to 2.3 micrometers in thickness and were uniform to less than 0.1 micrometers across the long dimension of the substrate. Films were single crystal with the (0001) A1N direction parallel to the (0001) Al2O3 direction. Films were clear and smooth replicating the surface of the substrate. Twelve samples were delivered in accordance with the requirements of this program. Surface acoustic wave interdigital electrode transducers and delay lines were fabricated on samples identical to those delivered. Measurements of delay line performance at 201, 477, and 848 MHz were made to provide an estimate of the basic constants for surface acoustic wave excitation and wave propagation. Relatively low acoustic propagation loss was observed for all AlN on sapphire delay lines studied. (Author)

Film cooling has been studied on the rotor blade of a large scale (low speed) model of a high pressure turbine first stage. Film coolant was discharged from single holes on the pressure and suction surfaces on the airfoil. For each blowing site the coolant-to-free stream mass flux ratio and density ratio were varied from 0.5 to 1.5 and from 1.0 and 4.0 respectively. Both surface flow visualization and local film cooling effectiveness data were obtained. Film coolant effectiveness data is presented in the form of effectiveness profiles and contours downstream of each hole. The observation was made that although it can have a strong radial component, the trajectory of the film coolant was very insensitive to coolant flow conditions. The existence of the radial component of the film coolant trajectory was found to have a strong impact on the nature of the effectiveness distributions. The data have been compared with data taken by other investigators on flat surfaces and in plane cascades. Agreement between the flat plate data and the suction surface data was reasonably good. However, the pressure surface results showed a much faster decay of the effectiveness than did the flat plate data due to effects thought to be related to both curvature and radial flow. (Author)

A method for computing three-dimensional flow over an ogival body at angle of attack is described. An approximate set of governing equations is given for viscous flows which have a primary flow direction. A two-level second-order accurate marching procedure is presented for general equations. With this procedure, a three-dimensional turbulent flow can be solved in any coordinate system by marching along the assumed primary flow direction. General tube-like coordinates are developed for a class of geometries applicable to flows between tubular surfaces. The coordinates are then specialized to the flow field bounded between an ogival body at angle of attack and its bow shock. Unlike the ogival body surface, the bow shock surface is not known in advance of the solution but instead must be computed as the solution develops. One marching step of the solution process is broken down into several steps. First, the bow shock surface is discretely extended by an iteration of explicit local solutions. The bow shock surface is then smoothly extended to provide a best fit to the discrete shock data. Tube-like coordinates are generated and finally the second order numerical scheme is applied to advance the fully viscous solution to the next station. In addition, some preliminary computational results were obtained. Specifically, the code was applied to subsonic boundary layers, purely supersonic flow with shocks, and mixed subsonic-supersonic flow. (Author)