ALBERT

Description: Basic algorithms for unstructured mesh generation and fluid flow calculation are discussed. In particular the following are addressed: preliminaries of graphs and meshes; duality and data structures; basic graph operations important in CFD (Computational Fluid Dynamics); triangulation methods, including Varonoi diagrams and Delaunay triangulation; maximum principle analysis; finite volume schemes for scalar conservation law equations; finite volume schemes for the Euler and Navier-Stokes equations; and convergence acceleration for steady state calculations.

Description: A 54.95-MHz coherent backscatter radar, an ionosonde and the magnetometer located at Trivandrum in India (8.5 deg N, 77 deg E, 0.5 deg N dip angle) recorded large-amplitude ionospheric fluctuations and magnetic field fluctuations associated with a Pc5 micropulsation event, which occurred during an intense magnetic storm on 24 March 1991 (A(sub p) = 161). Simultaneous 100-n T-level fluctuations are also observed in the H-component at Brorfelde, Denmark (55.6 deg N gm) and at Narsarsuaq, Greenland (70.6 deg N gm). Our study of the above observations shows that the E-W electric field fluctuations in the E- and F-regions and the magnetic field fluctuations at Thumba are dominated by a near-sinusoidal oscillation of 10 min during 1730-1900 IST (1200-1330 UT), the amplitude of the electric field oscillation in the equatorial electrojet (EEJ) is 0.1-0.25 mV/m and it increases with height, while it is about 1.0 mV/m in the F-region, the ground-level H-component oscillation can be accounted for by the ionospheric current oscillation generated by the observed electric field oscillation in the EEJ and the H-component oscillations at Trivandrum and Brofelde are in phase with each other. The observations are interpreted in terms of a compressional cavity mode resonance in the inner magnetosphere and the assoicated ionospheric electric field penetrating from high latitudes to the magnetic equator.

Description: The Coast Ranges of the Cascadia margin are overriding the subducted Juan de Fuca/Gorda plate. We investigate the extent to which the latitudinal change in attributes related to the subduction process. These attributes include the varibale age of the subducted slab that underlies the Coast Ranges and average vertical crustal velocities of the western margin of the Coast Rnages for two markedly different time periods, the last 45 years and the last 100 kyr. These vertical crustal velocities are computed from the resurveying of highway bech marks and from the present elevation of shore platforms that have been uplifted in the late Quaternary, respectively. Topogarphy of the Coast Ranges is in part a function of the age and bouyancy of the underlying subducted plate. This is evident in the fact that the two highest topographic elements of the Coast Rnages, the Klamath Mountains and the Olympic Mountains, are underlain by youngest subducted oceanic crust. The subducted Blanco Fracture Zone in southernmost Oregon is responsible for an age discontinuity of subducted crust under the Klamath Mountains. The norhtern terminus of hte topographically higher Klamaths is offset to the north relative to the position of the underlying Blanco Fracture Zone, teh offset being in the direction of migration of the farcture zone, as dictated by relative plate motions. Vertical crustal velocities at the coast, derived from becnh mark surveys, are as much as an order of magnitude greater than vertical crustal velocities derived from uplifted shore platforms. This uplift rate discrepancy indicates that strain is accumulating on the plate margin, to be released during the next interplate earthquake. In a latitudinal sense, average Coast Rnage topography is relatively high where bench mark-derived, short-term vertical crustal velocities are highest. Becuase the shore platform vertical crustal velocities reflect longer-term, premanent uplift, we infer that a small percentage of the interseismic strain that accumulates as rapid short-term uplift is not recovered by subduction earthquakes but rather contributes to rock uplift of the Coast Ranges. The conjecture that permanent rock uplift is related to interseismic uplift is consistent with the observation that those segments of the subduction zone subject to greater interseismic uplift rates are at approximately the same latitudes as those segments of the Coast Ranges that have higher magnitudes of rock uplift over the long term.

Description: A comprehensive review is presented of the mathematical models used to represent magnetic fields in the Earth's magnetosphere, of the way existing data-based models use these methods and of the associated problems and concepts. The magnetic field has five main components: the internal field, the magnetopause, the ring current, the tail and Birkeland currents. Methods of representing separately each of these are discussed, as is the deformation of magnetic fields; Appendix B traces the connection between deformations and the Cauchy integral. A summary section lists the uses of data-based models and their likely future evolution, and Appendix A supplements the text with a set of problems.

Description: It is shown that to satisfy the general accepted compressible law of the wall derived from the Van Driest transformation, turbulence modeling coefficients must actually be functions of density gradients. The transformed velocity profiles obtained by using standard turbulence model constants have too small a value of the effective von Karman constant kappa in the log-law region (inner layer). Thus, if the model is otherwise accurate, the wake component is overpredicted and the predicted skin friction is lower than the expected value.

Description: We present a map of the coseimic displacement field resulting from the Landers, California, June 28, 1992, earthquake derived using data acquired from an orbiting high-resolution radar system. We achieve results more accurate than previous space studies and similar in accuracy to those obtained by conventional field survey techniques. Data from the ERS 1 synthetic aperture radar instrument acquired in April, July, and August 1992 are used to generate a high-resolution, wide area map of the displacements. The data represent the motion in the direction of the radar line of sight to centimeter level precision of each 30-m resolution element in a 113 km by 90 km image. Our coseismic displacement contour map gives a lobed pattern consistent with theoretical models of the displacement field from the earthquake. Fine structure observed as displacement tiling in regions several kilometers from the fault appears to be the result of local surface fracturing. Comparison of these data with Global Positioning System and electronic distance measurement survey data yield a correlation of 0.96; thus the radar measurements are a means to extend the point measurements acquired by traditional techniques to an area map format. The technique we use is (1) more automatic, (2) more precise, and (3) better validated than previous similar applications of differential radar interferometry. Since we require only remotely sensed satellite data with no additioanl requirements for ancillary information. the technique is well suited for global seismic monitoring and analysis.

Description: The multigrid method has been applied to an existing three-dimensional compressible Euler solver to accelerate the convergence of the implicit symmetric relaxation scheme. This lower-upper symmetric Gauss-Seidel implicit scheme is shown to be an effective multigrid driver in three dimensions. A grid refinement study is performed including the effects of large cell aspect ratio meshes. Performance figures of the present multigrid code on Cray computers including the new C90 are presented. A reduction of three orders of magnitude in the residual for a three-dimensional transonic inviscid flow using 920 k grid points is obtained in less than 4 min on a Cray C90.

Description: The present paper explores the use of large-eddy simulations as a tool for predicting noise from first principles. A high-order numerical scheme is used to perform large-eddy simulations of a supersonic jet flow with emphasis on capturing the time-dependent flow structure representating the sound source. The wavelike nature of this structure under random inflow disturbances is demonstrated. This wavelike structure is then enhanced by taking the inflow disturbances to be purely harmonic. Application of Lighthill's theory to calculate the far-field noise, with the sound source obtained from the calculated time-dependent near field, is demonstrated. Alternative approaches to coupling the near-field sound source to the far-field sound are discussed.

Description: New line parameters for two heavy odd nitrogen molecules HNO3 in the nu(sub 5)/2nu(sub 9) region, and ClONO2 in the nu(sub 4) region are incorporated in the analysis of high resolution i.r. atmospheric spectra. The line parameters are tested and renormalized vs laboratory spectra, and then applied to retrievals from balloon-borne and ground-based solar absorption spectra.

Description: The spectroscopic identification for the HNO3 3 nu(sub 9) - nu(sub 9) band Q branch at 830.4/cm is reported based on 0.01/cm resolution solar occultation spectra of the lower stratosphere recorded by the Atmospheric Trace Molecule Spectroscopy (ATMOS) Fourier transform spectrometer and a recent analysis of this band. Least-squares fits to 0.0025/cm resolution laboratory spectra in the Q branch region indicate an integrated intensity of 0.529 x 10(exp -18)/cm/mol/sq cm at 296 K for this weak band. Stratospheric HNO3 retrievals derived from the ATMOS data are consistent with this value within its estimated uncertainty of about +/- 30%. A set of spectroscopic line parameters suitable for atmospheric studies has been generated.

Description: About 200 i.r. solar spectra recorded at 0.01/cm resolution on 71 days between November 1991 and July 1993 at the Network for the Detection of Stratospheric Change (NDSC) station at Mauna Loa, Hawaii (latitude 19.53 deg N, longitude 155.58 deg W, elevation 3.459 km) have been analyzed with a nonlinear least-squares spectral fitting technique to study temporal variations in the total column of atmospheric ethane (C2H6) above the site. The results were derived from the analysis of the unresolved nu(sub 7) band (P)Q(sub 3) subbranch at 2976.8/cm. A distinct seasonal cycle is observed with a factor of 2 variation, a maximum total column of 1.16 x 10(exp 16) mol/sq cm at the end of winter, and a minimum total column of 0.53 x 10(exp 16) mol/sq cm at the end of summer. Our measurements are compared with previous observations and model predictions.

Description: Quantitative measurements of the wavelength dependence of aerosol extinction in the 750-3400/cm spectral region have been derived from 0.01/cm resolution stratospheric solar occultation spectra recorded by the ATMOS (Atmospheric Trace Molecule Spectroscopy) Fourier transform spectrometer about 9 1/2 months after the Mt Pinatubo volcanic eruption. Strong, broad aerosol features have been identified near 900, 1060, 1190, 1720, and 2900/cm below a tangent height of approximately 30 km. Aerosol extinction measurements derived from approximately 0.05/cm wide microwindows nearly free of telluric line absorption in the ATMOS spectra are compared with transmission calculations derived from aerosol size distribution profiles retrieved from correlative SAGE (Stratospheric Aerosol and Gas Experiment) II visible and near i.r. extinction measurements, seasonal and zonally averaged H2SO4 aerosol weight percentage profiles, and published sulfuric acid optical constants derived from room temperature laboratory measurements. The calculated shapes and positions of the aerosol features are generally consistent with the observations, thereby confirming that the aerosols are predominantly concentrated H2SO4-H2O droplets, but there are significant differences between the measured and calculated wavelength dependences of the aerosol extinction. We attribute these differences as primarily the result of errors in the calculated low temperature H2SO4-H2O optical constants. Errors in both the published room temperature optical constants and the limitations of the Lorentz-Lorenz relation are likely to be important.

Description: We present vertical column abundances of H2O, N2O, HNO3, NO2, O3, HF, HCl, and ClNO3, determined from solar absorption spectra measured by the JPL MkIV interferometer from the NASA DC-8 aircraft. These observations, taken in 1987 and 1992, covered latitudes ranging from 85 deg S to 85 deg N. Although most gases display latitude symmetry, large asymmetries in H2O, HNO3, and O3 are apparent, which can be ascribed to processes enhanced by the colder Antarctic winter temperatures.

Description: Simultaneous in situ measurements of NO2, NO, O3, ClO, pressure and temperature have been made for the first time, presenting a unique opportunity to test our current understanding of the photochemistry of the lower stratospere. Data were collected from several flights of the ER-2 aircraft at mid-latitudes in May 1993 during NASA's Stratospheric Photochemistry, Aerosols and Dynamics Expedition (SPADE). The daytime ratio of NO2/NO remains fairly constant at 19 km with a typical value of 0.68 and standard deviation of +/- 17. The ratio observations are compared with simple steady-state calculations based on laboratory-measured reaction rates and modeled NO2 photolysis rates. At each measurement point the daytime NO2/NO with its measurements uncertainty overlap the results of steady-state caculations and associated uncertainty. Possible sources of error are examined in both model and measurements. It is shown that more accurate laboratory determinations of the NO + 03 reaction rate and of the NO2 cross-sections in the 200-220 K temperature range characteristic of the lower stratosphere would allow for a more robust test of our knowledge of NO(X) phtochemistry by reducing significant sources if uncertainties in the interpretation of statospheric measurements.

Description: The steady state solution of the system of equations consisting of the full Navier-Stokes equations and two turbulence equations has been obtained using a multigrid strategy of unstructured meshes. The flow equations and turbulence equations are solved in a loosely coupled manner. The flow equations are advanced in time using a multistage Runge-Kutta time-stepping scheme with a stability-bound local time step, while turbulence equations are advanced in a point-implicit scheme with a time step which guarantees stability and positivity. Low-Reynolds-number modifications to the original two-equation model are incorporated in a manner which results in well-behaved equations for arbitrarily small wall distances. A variety of aerodynamic flows are solved, initializing all quantities with uniform freestream values. Rapid and uniform convergence rates for the flow and turbulence equations are observed.

Description: The ability to predict short-term variations in the Earth's rotation has gained importance in recent years owing to more precise spacecraft tracking requirements. Universal time (UT1), that component of the Earth's orientation corresponding to the rotation angle, can be measured by number of high-precision space geodetic techniques. A Kalman filter developed at the Jet Propulsion Laboratory (JPL) optimally combines these different data sets and generates a smoothed times series and a set of predictions for UT1, as well as for additional Earth orientation components. These UT1 predictions utilize an empirically derived random walk stochastic model for the length of the day (LOD) and require frequent and up-to-date measurements of either UT1 or LOD to keep errors from quickly accumulating. Recent studies have shown that LOD variations are correlated with changes in the Earth's axial atmospheric angular momentum (AAM) over timescales of several years down to as little as 8 days. AAM estimates and forecasts out to 10 days are routinely available from meteorological analysis centers; these data can supplement geodetic measurements to improve the short-term prediction of LOD and have therefore been incorporated as independent data types in the JPL Kalman filter. We find that AAM and, to a lesser extent, AAM forecast data are extremely helpful in generating accurate near-real-time estimates of UT1 and LOD and in improving short-term predictions of these quantities out to about 10 days.

Description: Disk-shaped current distributions are useful tools in modeling the magnetospheres of Earth and other planets and have also been adapted for modeling the geotail. Such models usually start with an axisymmetric vector potential but may be modified to account for observed asymmetries, variable thickness, and warping in response to an inclined orientation of the planetary dipole axis. Models of this type until now either have lacked the ability to simulate a sharp inner edge of the current and to control accurately its falloff with distance or did not allow a simple analytical representation. Here existing methods will be reviewed, after which a new class of models which overcomes the above deficiencies and also allows the modeling of current disks of finite thickness flanked by current-free regions will be presented.

Description: An improved model of Earth's gravitational field, Goddard Earth Model T-3 (GEM-T3), has been developed from a combination of satellite tracking, satellite altimeter, and surface gravimetric data. GEM-T3 provides a significant improvement in the modeling of the gravity field at half wavelengths of 400 km and longer. This model, complete to degree and order 50, yields more accurate satellite orbits and an improved geoid representation than previous Goddard Earth Models. GEM-T3 uses altimeter data from GEOS 3 (1975-1976), Seasat (1978) and Geosat (1986-1987). Tracking information used in the solution includes more than 1300 arcs of data encompassing 31 different satellites. The recovery of the long-wavelength components of the solution relies mostly on highly precise satellite laser ranging (SLR) data, but also includes Tracking Network (TRANET) Doppler, optical, and satellite-to-satellite tracking acquired between the ATS 6 and GEOS 3 satellites. The main advances over GEM-T2 (beyond the inclusion of altimeter and surface gravity information which is essential for the resolution of the shorter wavelength geoid) are some improved tracking data analysis approaches and additional SLR data. Although the use of altimeter data has greatly enhanced the modeling of the ocean geoid between 65 deg N and 60 deg S latitudes in GEM-T3, the lack of accurate detailed surface gravimetry leaves poor geoid resolution over many continental regions of great tectonic interest (e.g., Himalayas, Andes). Estimates of polar motion, tracking station coordinates, and long-wavelength ocean tidal terms were also made (accounting for 6330 parameters). GEM-T3 has undergone error calibration using a technique based on subset solutions to produce reliable error estimates. The calibration is based on the condition that the expected mean square deviation of a subset gravity solution from the full set values is predicted by the solutions' error covariances. Data weights are iteratively adjusted until this condition for the error calibration is satisfied. In addition, gravity field tests were performed on strong satellite data sets withheld from the solution (thereby ensuring their independence). In these tests, the performance of the subset models on the withheld observations is compared to error projections based on their calibrated error covariances. These results demonstrate that orbit accuracy projections are reliable for new satellites which were not included in GEM-T3.

Description: We present predictions of the signatures of magnetosheath particle precipitation (in the regions classified as open low-latitude boundary layer, cusp, mantle and polar cap) for periods when the interplanetary magnetic field has a southward component. These are made using the 'pulsating cusp' model of the effects of time-varying magnetic reconnection at the dayside magnetopause. Predictions are made for both low-altitude satellites in the topside ionosphere and for midaltitude spacecraft in the magnetosphere. Low-altitude cusp signatures, which show a continuous ion dispersion signature, reveal 'quasi-steady reconnection' (one limit of the pulsating cusp model), which persists for a period of at least 10 min. We estimate that 'quasi-steady' in this context corresponds to fluctuations in the reconnection rate of a factor of 2 or less. The other limit of the pulsating cusp model explains the instantaneous jumps in the precipitating ion spectrum that have been observed at low altitudes. Such jumps are produced by isolated pulses of reconnection: that is, they are separated by intervals when the reconnection rate is zero. These also generate convecting patches on the magnetopause in which the field lines thread the boundary via a rotational discontinuity separated by more extensive regions of tangential discontinuity. Predictions of the corresponding ion precipitation signatures seen by midaltitude spacecraft are presented. We resolve the apparent contradiction between estimates of the width of the injection region from midaltitude data and the concept of continuous entry of solar wind plasma along open field lines. In addition, we reevaluate the use of pitch angle-energy dispersion to estimate the injection distance.

Description: Simultaneous profiles of aerosol backscatter ratio were measured over Lauder, New Zealand (45 deg S, 170 deg E) on the night of November 24, 1992. Instrumentation comprised two complementary lidar systems and a backscattersonde, to give measurements at wavelengths 351, 490, 532, and 940 nm. The data from the lidars and the backscattersonde were self-consistent, enabling the wavelength dependence of aerosol backscatter to be determined as a function of altitude. This wavelength-dependence is a useful parameter in radiative transfer calculations. In the stratosphere, the average wavelength exponent between 351 and 940 nm was -1.23 +/- 0.1, which was in good agreement with values derived from measured physical properties of aerosols.

Description: High-energy electrons have been measured systematically in a low-altitude (520 x 675 km), nearly polar (inclination = 82 deg) orbit by sensitive instruments onboard the Solar, Anomalous, and Magnetospheric Particle Explorer (SAMPEX). Count rate channels with electron energy thresholds ranging from 0.4 MeV to 3.5 MeV in three different instruments have been used to examine relativistic electron variations as a function of L-shell parameter and time. A long run of essentially continuous data (July 1992 - July 1993) shows substantial acceleration of energetic electrons throughout much of the magnetosphere on rapid time scales. This acceleration appears to be due to solar wind velocity enhancements and is surprisingly large in that the radiation belt 'slot' region often is filled temporarily and electron fluxes are strongly enhanced even at very low L-values (L aprroximately 2). A superposed epoch analysis shows that electron fluxes rise rapidly for 2.5 is approximately less than L is approximately less than 5. These increases occur on a time scale of order 1-2 days and are most abrupt for L-values near 3. The temporal decay rate of the fluxes is dependent on energy and L-value and may be described by J = Ke-t/to with t(sub o) approximately equals 5-10 days. Thus, these results suggest that the Earth's magnetosphere is a cosmic electron accelerator of substantial strength and efficiency.

Description: The analysis of balloon envelopes by the finite element (FE) method is plagued by convergence problems. A pratical FE analysis approach is based on the fact that in thin shells with non-zero Gaussian curvature the membrane solution component is essentially decoupled from the bending solution component. A proxy-problem is solved by using a small artificial bending stiffness that assures convergence without significantly affecting the membrane solution component. This approach has been previously validated on slightly overpressurized balloon envelopes. Extensions of this approach to more difficult problems in the structural analysis of balloon envelopes are presented. The convergence forcing modelling measures are discussed. Implications of the findings of the analysis results to future balloon designs are also discussed.

Description: Rocket data have been used to evaluate the characteristics of precipitating relativistic electrons and their effects on the electrodynamic structure of the middle atmosphere. These data were obtained at Poker Flat, Alaska, on May 13 and 14, 1990, during a midday, highly relativistic electron (HRE) precipitation event. Solid state detectors were used to measure the electron fluxes and their energy spectra. An X ray scintillator was included on each flight to measure bremsstrahlung X rays produced by energetic electrons impacting on the upper atmosphere. However, these were found the be of negligible importance for this particular event. The energy deposition by the electrons has been determined from the flux measurements and compared with in situ measurements of the atmospheric electrical response. The electrodynamic measurements were obtained by the same rockets and additionally on May 13, with an accompanying rocket. The impact flux was highly irregular, containing short-lived bursts of relativistic electrons, mainly with energies below 0.5 MeV and with fluxes most enhanced between pitch angles of 0 deg - 20 deg. Although the geostationary counterpart of this measured event was considered to be of relatively low intensity and hardness, energy deposition peaked near 75 km with fluxes approaching an ion pair production rate in excess of 100/cu cm s. This exceeds peak fluxes in relativistic electron precipitation (REP) events as observed by us in numerous rocket soundings since 1976. Conductivity measurements from a blunt probe showed that negative electrical conductivities exceeded positive conductivities down to 50 km or lower, consistent with steady ionization by precipitating electrons above 1 MeV. These findings imply that the electrons from the outer radiation zone can modulate the electrical properties of the middle atmosphere to altitudes below 50 km. During the decline and activity minimum of the current solar cycle, we anticipate the occurence of similar events but with fluxes 1-2 orders of magnitude above that reported here, based on studies of earlier solar cycles (e.g., Baker et al., 1993).

Description: Nitrous oxide (N2O) measured on board the ER-2 aircraft during the Airborne Arctic Stratospheric Expedition 2 (AASE 2) has been used to monitor descent of air inside the Arctic vortex between October 1991 and March 1992. Monthly mean N2O fields are calculated from the flight data and then compared with mean fields calculated from the high-resolution Geophysical Fluid Dynamics Laboratory general circulation model SKYHI in order to evaluate the model's simulation of the polar vortex. From late fall through winter the model vortex evolves in much the same way as the 1991-1992 vortex, with N2O gradients at the edge becoming progressively steeper. The October to March trends in N2O profiles inside the vortex are used to verify daily net heating rates in the vortex that were computed from clear sky radiative heating rates and National Meteorological Center temperature observations. The computed heating rates successfully estimate the descent of vortex air from December through February but suggest that before December, air at high latitudes may not be isolated from the midlatitudes. SKYHI heating rates are in good agreement with the computed rates but tend to be slightly higher (i.e., less cooling) due to meteorological differences between SKYHI and the 1991-1992 winter. Three ER-2 flights measured N2O just north of the subtropical jet. These low-midlatitude profiles show only slight differences from the high-midlatitude profiles (45 deg - 60 deg N), indicating strong meridional mixing in the midlatitude 'surf zone.' Mean midwinter N2O profiles inside and outside the vortex calculated from AASE 2 data are shown to be nearly identical to 1989 AASE profiles, pointing to the N2O/potential temperature relationship as an excellent marker for vortex air.

Description: The combined Nimbus 7 solar backscattered ultraviolet (SBUV) and NOAA 11 SBUV/2 ozone data, covering a period of more than a solar cycle (about 15 years), are used to study the UV response of ozone in the stratosphere. The study shows that about 2% change in total column ozone and about 5-7% change in ozone mixing ratio in the upper stratosphere (0.7 to 2 hPa) may be attributed to the change in the solar UV flux over a solar cycle. In the upper stratosphere, where photochemical processes are expected to play a major role, the measured solar cycle variation of ozone is significantly larger than inferred either from the photochemical models or from the ozone response to the 27-day solar UV modulation. For example, the observed solar cycle related change in ozone mixing ratio at 2 hPa is about 1% for 1% change in the solar UV flux near 200 nm. The inferred change in ozone from either the photochemical models or from the 27-day ozone-UV response is about a factor of 2-3 lower than this value.

Description: Vertical profiles of N2O and NO(y) taken by the ER-2 outside the vortex are used to construct average vertical profiles of F(NO(y)) = NO(y)/(A-N2O), where A is the tropospheric content of N2O three years prior to the measurements. The southern hemisphere had less nitrous oxide in the range 400 less than Theta less than 470 K, by up to 25% relative to the northern hemisphere. F(NO(y)) is the ratio of NOy produced to N2O lost in a stratospheric air mass since entry from the troposphere. The profiles of F(NO(y)) have the following characteristics: (1) Relative to 1991-1992, a year without denitrification inside or outside the vortex, the northern hemisphere in 1988-1989 showed denitrification outside the vortex ranging up to 25% and averaging 17% above Theta = 425 K. (2) Relative to the northern hemisphere in 1991-1992, the southern hemisphere in 1987 showed denitrification outside the vortex ranging up to 32% and averaging 20% above Theta = 400 K. (3) Below Theta = 400 K the southern hemisphere showed enhancements of F(NO(y)) relative to the northern hemisphere in 1991-1992 ranging up to 200% at Theta = 375 K, outside the vortex. Corresponding profiles of residual water, R(H2O) = H2O - 2(1.6 - CH4), are considered and shown to be consistent with those of F(NO(y)) in the sense that they show deficits outside the Antarctic vortex, which was both dehydrated and denitrified, but not outside the 1988-1989 Arctic vortex, which was denitrified but not dehydrated. R(H2O) is the water content of stratospheric air with the contribution from methane oxidation subtracted. Comparison of F(NO(y)) and R(H2O) below 400 K outside the Antarctic vortex leads to the suggetion that dehydration in the Antarctic vortex occurs by the sedimentation of ice crystals large enough to fall out of the stratosphere, whereas denitrification occurs mainly on mixed nitric acid-water crystals which evaporate below the base of the vortex at Theta = 400 K but above the tropopause.

Description: During January and August 1985, the scanning radiometers of the Earth Radiation Budget Experiment(ERBE) aboard the Earth Radiation Budget Satellite (ERBS) and the NOAA-9 satellite were operated in along-track scanning modes. Along-track scanning permits the study of many measurement problems. It provides the data for developing a limb-darkening model for a single site over a short period of time and also permits the indentification of the scene from data taken at smaller nadir angles. The earth-emitted radiation measured by the scanners has been analyzed to produce limb-darkening models for a variety of scene types. Limb-darkening models relate the radiance in any given direction to the radiant flux. The scene types were computed using measurements within 10 deg of zenith. The models have values near zenith of 1.02-1.09. The typical zenith values of the model are 1.06 for both day and night for ERBS, and for NOAA-9, 1.06 for day and 1.05 for night. Mean models are formed for the ERBS and NOAA-9 results and are found to differ less than 1%, the ERBS results being the higher. The models vary about 1% with latitude near zenith and agree with earlier models that were used to analyze ERBE data typically to 2%.

Description: One of the oldest mysteries in geomagnetism is the linkage between solar and geomagnetic activity. In investigating the causes of geomagnetic storms occurring during solar maximum, the following topics are discussed: solar phenomena; types of solar wind; magnetic reconnection and magnetic storms; an interplanetary example; and future space physics missions.

Description: Atmospheric mass loading produces a primarily vertical displacement of the Earth's crust. This displacement is correlated with surface pressure and is large enough to be detected by very long baseline interferometry (VLBI) measurements. Using the measured surface pressure at VLBI stations, we have estimated the atmospheric loading term for each station location directly from VLBI data acquired from 1979 to 1992. Our estimates of the vertical sensitivity to change in pressure range from 0 to -0.6 mm/mbar depending on the station. These estimates agree with inverted barometer model calculations (Manabe et al., 1991; vanDam and Herring, 1994) of the vertical displacement sensitivity computed by convolving actual pressure distributions with loading Green's functions. The pressure sensitivity tends to be smaller for stations near the coast, which is consistent with the inverted barometer hypothesis. Applying this estimated pressure loading correction in standard VLBI geodetic analysis improves the repeatability of estimated lengths of 25 out of 37 baselines that were measured at least 50 times. In a root-sum-square (rss) sense, the improvement generally increases with baseline length at a rate of about 0.3 to 0.6 ppb depending on whether the baseline stations are close to the coast. For the 5998-km baseline from Westford, Massachusetts, to Wettzell, Germany, the rss improvement is about 3.6 mm out of 11.0 mm. The average rss reduction of the vertical scatter for inland stations ranges from 2.7 to 5.4 mm.

Description: A theoretical model is used to describe and investigate the effects of simultaneous crystallization, radiation loss, and entrainment of cooler material on the temperature of a well-mixed core of an active aa lava flow. Entrainment of crust, levee debris, and base material into the interior of active flows has been observed, but the degree of assimilation and the thermal consequences are difficult to quantify. The rate of entrainment can be constrained by supplementing the theoretical model with information on the crystallization along the path of the flow and estimation of the radiative loss from the flow interior. Application of the model is demonstrated with the 1984 Mauna Loa flow, which was erupted about 30 C undercooled. Without any entrainment of cooler material, the high crystallization rates would have driven temperatures in the core wall above temperatures measured by thermocouple and estimated from glass geothermometry. One plausible scenario for this flow, which agrees with available temperature and crystallinity measurements, has a high initial rate of entrainment during the first 8 hours of travel (a mass ratio of entrained material to fluid core of about 15% if the average temperature of the entrained material was 600 C), which counterbalances the latent heat from approximately 40% crystallization. In this scenario, the model suggests an additional 5% crystallization and a 5% entrainment mass ratio over the subsequent 16-hour period. Measurements of crystallization, radiative losses, and entrainment factors are necessary for understanding the detailed thermal histories of active lava flows.

Description: Numerical results obtained with direct simulation Monte Carlo and Navier-Stokes methods are presented for a Mach-20 nitrogen flow about a 70-deg blunted cone. The flow conditions simuulated are those that can be obtained in existing low-density hypersonic wind tunnels. Three sets of flow conditions are considered with freestream Knudsen numbers ranging from 0.03 to 0.001. The focus is on the wake structure: how the wake structure changes as a function of rarefaction, what the afterbody levels of heating are, and to what limits the continuum models are realistic as rarefaction in the wake is progressively increased. Calculations are made with and without an afterbody sting. Results for the after body sting are emphasizes in anticipation of an experimental study for the current flow conditions and model configuration. The Navier-Stokes calculations were made with and without slip boundary conditions. Comparisons of the results obtained with the two simulation methodologies are made for both flowfield structure and surface quantities.

Description: The data from the infrared telescope (IRT), which was flown on space shuttle Challenger Spacelab 2 mission (July 1985), were originally reported by Koch et al. (1987) as originating from near orbital emissions, primarily H2O. In this study, analysis of this data was extended to determine the collisional cross sections for the excitation of the low lying vibrational levels of H2O, present in the orbiter cloud, by atmospheric O(3P). The evaluation of the contribution to the measured signal from solar excitation and ram O excitation of outgassing H2O permits the determination of the H2O column density and the excitation cross section of the (101) level at an O(3P) velocity of approximately 7.75 km/s. Contributions to the radiation in the 1.7-3.0 micron band by transitions from the (100), (001), and multiquantum excited levels are discussed. The findings of the study are (1) the IRT data for the 4.5-9.5 micron and the nighttime data for the 1.7-3.0 micron sensors are consistent with being explained by collision excitation of H2O by O(3P), (2) diurnal variations of 4.5-9.5 micron intensities follow the model predicted O density for a full orbit, (3) daytime increases in the H2O cloud density were not evident, (4) the cross sections for the collisional excitation process are derived and compared to values computated by Johnson (1986) and Redmon et al. (1986), (5) theoretical investigation suggests greater than 60% of the radiation from H2O is a result of multiphoton emission resulting from collisional multiquanta excitation, and (6) the large daytime increase in the 1.7-3.0 micron intensity data suggests that O(+) may likely be instrumental in producing excited H2O(+) through charge exchange.

Description: The devolatilization of calcium sulfate, which is present in the target rock of the Chicxulub, Mexico impact structure, and dispersal in the stratosphere of the resultant sulfuric acid aerosol have been suggested as a possible mechanism for the Cretaceous-Tertiary extinctions. We measured the amount of SO2 produced from two shock-induced devolatilization reactions of calcium sulfate up to 42 GPa in the laboratory. We found both to proceed to a much lower extent than calculated by equilibrium thermodynamic calculations. Reaction products are found to be approx. 10(exp -2) times those calculated for equilibrium. Upon modeling the quantity of sulfur oxides degassed into the atmosphere from shock devolatilization of CaSO4 in the Chicxulub lithographic section, the resulting 9 x 10(exp 16) to 6 x 10(exp 17) g (in sulfur mass) is lower by a factor of 10-100 than previous upper limit estimates, the related environmental stress arising from the resultant global cooling and fallout of acid rain is insufficient to explain the widespread K-T extinctions.

Description: Energetic atomic (O(+1) and N(+1)) and molecular (O2(+1), NO(+1), and N2(+1)) ions of ionospheric origin were observed in Earth's magnetotail at X approximately -146 R(sub E) during two plasma sheet sunward/tailward flow-reversal events measured by instruments on the GEOTAIL spacecraft. These events were associated with concurrent ground-measured geomagnetic disturbance intensification at auroral-and mid-latitudes (Kp = 7(-)). Energetic ions in the sunward-component and tailward flows were from both the solar wind and ionosphere. Plasma and energetic ions participated in the flows. During tailward flow, ionospheric origin ion abundance ratios at approximately 200-900 km/s in the rest frame were N(+1)/O(+1) = approximately 25-30% and ((O2(+1), NO(+1), and N2(+1))/O(+1) = approximately 1-2%. We argue that tailward flow most likely initiated approximately 80-100 R(sub E) tailward of Earth and molecular ions were in the plasma sheet prior to geomagnetic intensification onset.

Description: Wave forms of BEN (Broadband Electrostatic Noise) in the geomagnetic tail were first detected by the Wave Form Capture reciever on the GEOTAIL spacecraft. The results show that most of the BEN in the plasma sheet boundary layer (PSBL) are not continuous broadband noise but are composed of a series of solitary pulses having a special form which we term 'Electrostatic Solitary Waves (ESW)'. A nonlinear BGK potential model is proposed as the generation mechanism for the ESW based upon a simple particle simulation which considers the highly nonlinear evolution of the electron beam instability. The wave forms produced by this simulation are very similar to those observed by GEOTAIL and suggest that the nonlinear dynamics of the electron beam play an essential role in the generation of ESW.

Description: We report on the observations of a number of quasi-dc electric field events associated with large-scale atmospheric weather formations. The observations were made by the electric field experiment onboard the San Marco D satellite, operational in an equatorial orbit from May to December 1988. Several theoretical studies suggest that electric fields generated by thunderstorms are present at high altitudes in the ionosphere. In spite of such favorable predictions, weather-related events are not often observed since they are relatively weak. We shall report here on a set of likely E field candidates for atmospheric-ionospheric causality, these being observed over the Indonesian Basin, northern South America, and the west coast of Africa; all known sites of atmospheric activity. As we shall demonstrate, individual events often be traced to specific active weather features. For example, a number of events were associated with spacecraft passages near Hurricane Joan in mid-October 1988. As a statistical set, the events appear to coincide with the most active regions of atmospheric weather.

Description: A radiation model, together with National Meteorological Center temperature observations, was used to compute daily net heating rates in the northern hemisphere (NH) for the Arctic late fall and winter periods of both 1988-1989 and 1991-1992 and in the southern hemisphere (SH) for the Antarctic fall and winters of 1987 and 1992. The heating rates were interpolated to potential temperature (theta) surfaces between 400 K and 2000 K and averaged within the polar vortex, the boundary of which was determined by the maximum gradient in potential vorticity. The averaged heating rates were used in a one-dimensional vortex interior descent model to compute the change in potential temperature with time of air parcels initialized at various theta values, as well as to compute the descent in log pressure coodinates. In the NH vortex, air parcels which were initialized at 18 km on November 1, descended about 6 km by March 21, while air initially at 25 km descended 9 km in the same time period. this represents an average descent rate in the lower stratosphere of 1.3 to 2 km per month. Air initialized at 50 km descended 27 km between November 1 and March 21. In the SH vortex, parcels initialized at 18 km on March 1, descended 3 km, while air at 25 km descended 5-7 km by the end of October. This is equivalent to an average descent in the lower stratosphere of 0.4 to 0.9 km per month during this 8-month period. Air initialized at 52 km descended 26-29 km between March 1 and October 31. In both the NH and the SH, computed descent rates increased markedly with height. The descent for the NH winter of 1992-1993 and the SH winter of 1992 computed with a three-dimensional trajectory model using the same radiation code was within 1 to 2 km of that calculated by the one-dimensional model, thus validating the vortex averaging procedure. The computed descent rates generally agree well with observations of long-lived tracers, thus validating the radiative transfer model.

Description: The map of the coseismic displacement field generated by interferometric processing of synthetic aperture radar (SAR) images taken before and after the June 28, 1992, Landers earthquake sequence brings new insights into the nature of deformation caused by these earthquakes. We use the interferometric map generated by Massonnet et al. (1993) to analyze the surface displacement field in the vicinity of the fault trace. Complexities in the fringe pattern near the fault reflect short-wavelength variations of the surface rupture and slip distribution, and attest to large displacement gradients. Along two sections of the fault, characteristic fringe patterns can be recognized, contrasting in density and direction with patterns observed away from the rupture. In order to understand the observed fringe patterns, we compute synthetic interferograms in three simple cases: (1) rigid-body rotations about a vertical axis, (2) about a horizontal axis (tilt), and (3) distributed, simple shear. The orientation and spatial separation of interferometric fringes predicted by these models help constrain near-field deformation and rupture parameters. Where the Kickapoo fault connects with the Homestead Valley fault, the interferogram shows a clear pattern of parallel N20 deg W fringes separated by about 160 m. This pattern and vertical offsets measured along the Kickapoo fault suggest that the block between this fault and the Johnson Valley fault may have been tilted, down to the west. A 5-km block lifted by 1 m on one side would be tilted by an angle of 0.01 deg (190 microrad), producing fringes separated by about 160 m, parallel to the tilt axis. Such a tilt, parallel to a N20 deg W direction, would account for the gradual, northward increase of the vertical slip component observed along the Kickapoo fault. This tilt may also explain the 1 m of reverse slip observed along the 'slip gap' section of the Homestead Valley break. Between the southern end of the Johnson Valley fault and the Eureka Peak fault, where no surface rupture has been mapped, the dense pattern of fringes implies distributed shear, probably resulting from fault slip at depth. The density and direction of the fringes in the gap are consistent with a right-lateral slip of 1.2-3.8 m on a blind fault locked above the depth of 1.5-2 km. Such observations of small wavelength features in the SAR interferogram bring new insights into the near-field displacement gradient and thus on response of the uppermost crust to seismic rupture.

Description: The acceleration of protons in a dynamically evolving magnetotail is investigated by tracing particles in the fields obtained from a three-dimensional resistive magnetohydrodynamic (MHD) simulation. The MHD simulation, representing plasmoid formation and ejection through a near-Earth reconnection process, leads to cross-tail electric fields of up to approximately 4 mV/m with integrated voltages across the tail of up to approximately 200 kV. Energization of particles takes place over a wide range along the tail, due to the large spatial extent of the increased electric field together with the finite cross-tail extent of the electric field region. Such accelerated particles appear earthward of the neutral line over a significant portion of the closed field line region inside of the separatrix, not just in the vicinity of the separatrix. Two different acceleration processes are identified: a 'quasi-potential' acceleration, due to particle motion in the direction of the cross-tail electric field, and a 'quasi-betatron' effect, which consists of multiple energy gains from repeated crossings of the acceleration region, mostly on Speiser-type orbits, in the spatially varying induced electric field. The major source region for accelerated particles in the hundreds of keV range is the central plasma sheet at the dawn flank outside the reconnection site. Since this source plasma is already hot and dense, its moderate energization by a factor of approximately 2 may be sufficient to explain the observed increases in the energetic particle fluxes. Particles from the tail are the source of beams at the plasma sheet/lobe boundary. The temporal increase in the energetic particle fluxes, estimated from the increase in energy gain, occurs on a fast timescale of a few minutes, coincident with a strong increase in B(sub z), despite the fact that the inner boundary ('injection boundary') of the distribution of energized particles is fairly smooth.

Description: A high-precision radiometric satellite tracking system, Doppler Orbitography and Radio-positioning Integrated by Satellite system (DORIS), has recently been developed by the French space agency, Centre National d'Etudes Spatiales (CNES). DORIS was designed to provide tracking support for missions such as the joint United States/French TOPEX/Poseidon. As part of the flight testing process, a DORIS package was flown on the French SPOT 2 satellite. A substantial quantity of geodetic quality tracking data was obtained on SPOT 2 from an extensive international DORIS tracking network. These data were analyzed to assess their accuracy and to evaluate the gravitational modeling enhancements provided by these data in combination with the Goddard Earth Model-T3 (GEM-T3) gravitational model. These observations have noise levels of 0.4 to 0.5 mm/s, with few residual systematic effects. Although the SPOT 2 satellite experiences high atmospheric drag forces, the precision and global coverage of the DORIS tracking data have enabled more extensive orbit parameterization to mitigate these effects. As a result, the SPOT 2 orbital errors have been reduced to an estimated radial accuracy in the 10-20 cm RMS range. The addition of these data, which encompass many regions heretofore lacking in precision satellite tracking, has significantly improved GEM-T3 and allowed greatly improved orbit accuracies for Sun-synchronous satellites like SPOT 2 (such as ERS 1 and EOS). Comparison of the ensuing gravity model with other contemporary fields (GRIM-4C2, TEG2B, and OSU91A) provides a means to assess the current state of knowledge of the Earth's gravity field. Thus, the DORIS experiment on SPOT 2 has provided a strong basis for evaluating this new orbit tracking technology and has demonstrated the important contribution of the DORIS network to the success of the TOPEX/Poseidon mission.

Description: The presence of anisotropic plasma distributions, trapped at the Earth's magnetic equator, has consequences for the electric field structure which must exist in equilibrium along the magnetic field line. Data from SCATHA and Dynamics Explorer 1 indicate that the core ion distributions at the magnetic equator can be well described as bi-Maxwellian distributions, with a perpendicular temperature an order of magnitude larger than the parallel temperature. A collisionless model is developed for the variation in plasma parameters, following the forms developed by Whipple (1977). If the core electron anisotropy is low, the resulting electric field of approximately 0.1 microV/m is pointed away from the equator. Under these conditions the self-consistent electric field will not overcome the effects of magnetic trapping. The resulting potential distribution results in a local maximum in total plasma density at the equator. Only when the electron distribution is primarily field-aligned can there be a density minimum at the equator. Comparisons are made between this model and the observed variations in DE 1 plasma parameters with latitude.

Description: The transformation validity question utilizing resulting data from direct numerical simulations (DNS) of supersonic, isothermal cold wall channel flow was investigated. The DNS results stood for a wide scope of parameter and were suitable for the purpose of examining the generality of Van Driest transformation. The Van Driest law of the wall can be obtained from the inner-layer similarity arguments. It was demonstrated that the Van Driest transformation cannot be incorporated to collapse the sublayer and log-layer velocity profiles simultaneously. Velocity and temperature predictions according to the preceding composite mixing-length model were presented. Despite satisfactory congruity with the DNS data, the model must be perceived as an engineering guide and not as a rigorous analysis.

Description: A Monte Carlo solution technique has been formulated to predict the radiative heat transfer in three-dimensional, inhomogeneous participating media which exhibit spectrally dependent emission and absorption and anisotropic scattering. Details of the technique and selected numerical sensitivities are discussed. The technique was applied to a problem involving a medium composed of a gas mixture of carbon dioxide and nitrogen and suspended carbon particles. A homogeneous medium was modeled to examine the effect of total pressure and carbon-particle concentration on radiative heat transfer. Variation in total pressure, over the range studied, had minimal effect on the amount of heat radiated to the enclosure walls and on the radiative-flux distribution within the medium. Increases in the carbon particle concentration produced significantly higher heat fluxes at the boundaries and altered the radiative flux distribution. The technique was then applied to an inhomogeneous medium to examine effects of specific temperature and carbon particle concentration distributions on radiative heat transfer. For the inhomogeneous conditions examined, the largest radiative flux divergence occurs near the center of the medium and the regions near some enclosure walls act as energy sinks.

Description: A combined 3-dimensional circulation model and aerosol microphysical/transport model is used to simulate the dispersion of the Mt. Pinatubo volcanic cloud in the stratosphere for the first few months following the eruption. Radiative heating of the cloud due to upwelling infrared radiation from the troposphere is shown to be an important factor affecting the transport. Without cloud heating, cloud material stays mostly north of the equator, whereas with cloud heating, the cloud is transported southward across the equator within the first two weeks following the eruption. Generally the simulations agree with Total Ozone Mapping Spectrometer (TOMS), Advanced Very High Resolution Radiometer (AVHRR), and Stratospheric Aerosol and Gas Experiment (SAGE) satellite observations showing the latitude distribution of cloud material to be between about 20 deg S and 30 deg N within the first few months. Temperature perturbations in the stratosphere induced by the aerosol heating are generally 1-4 K, in the range of those observed.

Description: We use geodetic data from Very Long Baseline Interferometry (VLBI) to determine the pre- and postseismic velocities of two sites. We then place limits on variations in interseismic strain buildup. The 1987 and 1988 Gulf of Alaska earthquakes (each Ms = 7.6) broke the Pacific plate interior. During the earthquakes the Cape Yakataga site moved 78 mm toward southwest. During the 1989 Loma Prieta earthquake (Ms = 7.1) the Fort Ord site moved 48 mm toward north. Baselines (a) from Fairbanks to Cape Yakataga and (b) from Mojave to Fort Ord change at nearly the same rate before and after the earthquakes. Postseismic transients, which we determine from differences between post- and preseismic rates, are minor: at Cape Yakataga the transient is 3 +/- 4 mm in a postseismic interval of 23 months, and at Fort Ord the transient is 6 +/- 5 mm in 21 months. The slip beneath the Loma Prieta rupture needed to generate the Fort Ord transient is 0.22 +/- 0.19 m, one-tenth the coseismic slip (2 m). We analyze elastic lithosphere-viscous asthenosphere models to determine that the characteristic time describing exponential decay in deep fault slip is longer than 6 years. The VLBI measurements are consistent with uniform interseismic strain buildup. They disagree with fast postseismic rates caused by an asthenosphere with very low viscosity.

Description: The AEPI (Atmospheric Emissions Photometric Imager) experiment on the ATLAS-1 shuttle mission (launched March 24, 1992) imaged the earth night airglow emission in O2 Atmospheric (0,0) bands, at 762.0 nm. Earthward views of O2 A bands show structure from gravity waves which exhibit extended horizontal structure with horizontal wavelenghts on the order of 50-100 km. These observations of the O2 A (0,0) bands are particularly interesting since in this wavelength the lower atmosphere absorbs all the earth-reflected emissions and most of the spectrally diffuse backgrounds. Herein we present observations of gravity waves using a topside airglow imaging technique.

Description: Fujii et al. (1994) obtained characteristics of the electrodynamic parameters, that is, field-aligned currents, electric fields, and electron precipitation, which are associted with auroral substorm events in the nighttime sector, through a unique analysis that places the ionospheric measurements of these parameters into the context of a generic substorm determined from global auroral images. In this paper we investigate in considerably more detail the characteristics of the field-aligned currents using data from the same set of passes as the previuos study. We show for the first time that the net upward field-aligned currents throughout the surge and surge horn are sufficient to account for most if not all of the converging currents of the auroral electrojets. Current densities are largest in the surge and surge horn. Current region continuity does not appear to exist across the substorm bulge region. Much of the auroral substorm field-aligned current is composed of filamentary currents and finite current segments at large angles to each other. The westward electrojet may contain large gradients in intensity both in local time and latitude due to sets of localized field-aligned currents. The net downward current for several hours to the west of the surge is insufficient to account for the eastward electrojet, consistent with the concept that this electrojet originates primarily on the dayside. Our pattern of field-aligned currents associated with the surge has common features and also differs significantly from the patterns previously derived from data from radars and ground-based magnetometer arrays. Our pattern is considerably more complex, probably due to the much higher resolution in latitude of the satellite data. It is also larger in area, since our average substorm is much larger than those pertaining to the previous patterns, giving a substorm wedge considerably wider than that obtained from the radar and array data.

Description: Numerical results obtained with direct simulation Monte Carlo and Navier-Stokes methods are presented for a Mach-20 nitrogen flow about a 70-deg blunted cone. The flow conditions simulated are those that can be obtained in existing low-density hypersonic wind tunnels. Three sets of flow conditions are considered with freestream Knudsen numbers ranging from 0.03 to 0.001. The focus is on the wake structure: how the wake structure changes as a function of rare faction, what the afterbody levels of heating are, and to what limits the continuum models are realistic as rarefunction in the wake is progressively increased. Calculations are made with and without an afterbody sting. Results for the afterbody sting are emphasized in anticipation of an experimental study for the current flow conditions and model configuration. The Navier-Stokes calculations were made with and without slip boundary conditions. Comparisons of the results obtained with the two simulation methodologies are made for both flowfield structure and surface quantities.

Description: An analytical model of the bidirectional reflectance for optically semi-infinite plant canopies has been extended to describe the reflectance of finite depth canopies contributions from the underlying soil. The model depends on 10 independent parameters describing vegetation and soil optical and structural properties. The model is inverted with a nonlinear minimization routine using directional reflectance data for lawn (leaf area index (LAI) is equal to 9.9), soybeans (LAI, 2.9) and simulated reflectance data (LAI, 1.0) from a numerical bidirectional reflectance distribution function (BRDF) model (Myneni et al., 1988). While the ten-parameter model results in relatively low rms differences for the BRDF, most of the retrieved parameters exhibit poor stability. The most stable parameter was the single-scattering albedo of the vegetation. Canopy albedo could be derived with an accuracy of less than 5% relative error in the visible and less than 1% in the near-infrared. Sensitivity were performed to determine which of the 10 parameters were most important and to assess the effects of Gaussian noise on the parameter retrievals. Out of the 10 parameters, three were identified which described most of the BRDF variability. At low LAI values the most influential parameters were the single-scattering albedos (both soil and vegetation) and LAI, while at higher LAI values (greater than 2.5) these shifted to the two scattering phase function parameters for vegetation and the single-scattering albedo of the vegetation. The three-parameter model, formed by fixing the seven least significant parameters, gave higher rms values but was less sensitive to noise in the BRDF than the full ten-parameter model. A full hemispherical reflectance data set for lawn was then interpolated to yield BRDF values corresponding to advanced very high resolution radiometer (AVHRR) scan geometries collected over a period of nine days. The resulting parameters and BRDFs are similar to those for the full sampling geometry, suggesting that the limited geometry of AVHRR measurements might be used to reliably retrieve BRDF and canopy albedo with this model.

Description: The Simplified Shuttle Payload Thermal Analyzer program (SSPTA) was developed to aid in the evaluation of thermal design concepts of instruments to be flown in the Space Shuttle cargo bay. SSPTA consists of a collection of programs that are currently used in the thermal analysis of spacecraft and have been modified for quick, preliminary analysis of payloads. SSPTA includes a reduced math model of the Shuttle cargo bay to simplify use of the program for payload analysis. One of the prime objectives in developing SSPTA was to create a program which was easy to use. With SSPTA, the user required input is simple and the user is free from many of the concerns of computer usage such as disk space handling, tape usage, and complicated program control. Although SSPTA was designed primarily to analyze Shuttle payloads, it can easily be used to perform thermal analysis in other situations. SSPTA is comprised of a system of data files called 'bins', a master program, and a set of thermal subprograms. The bin system is a collection of disk files which contain data required by or computed by the thermal subprograms. SSPTA currently has the capability of handling 50 bins. The master program serves primarily as a manager for the bin system and its interaction with the thermal subprograms. Input to the master program consists of simple user commands which direct the data manipulation procedures, prepare the data for these procedures, and call the appropriate thermal subprograms. The subprograms of SSPTA are all based on programs which have been used extensively in the analysis of orbiting spacecraft and space hardware. Subprogram CONSHAD uses the user supplied geometric radiation model to compute black body view factors, shadow factors, and a description of the surface model. The subprogram WORKSHEET uses the surface model description, optical property data, and node assignment data to prepare input for SCRIPTF. Subprogram SCRIPTF computes the inverses of the infrared (IR) and ultraviolet (UV) radiation transfer equations; it also computes the radiation coupling between nodes in the thermal model. Subprogram ORBITAL uses the shadow tables to compute incident flux intensities on each surface in the geometric model. Subprogram ABSORB uses these flux intensities combined with the IR and UV inverses to compute the IR and UV fluxes absorbed by each surface. The radiation couplings from SCRIPTF and the absorbed fluxes from ABSORB are used by subprogram TTA to compute the temperature and power balance for each node in the thermal model. Output consists of tabulated data from each of the subprograms executed during a particular analysis. Due to the modular form of SSPTA, analyses may be run in whole or in part, and new subprograms may be added by the user. SSPTA is written in FORTRAN for use on a DEC VAX-11/780. SSPTA was originally developed in 1977 for use on IBM 370 series computers. This version is an update which was ported to the VAX in 1980.

Description: Remote Interactive Particle-tracing (RIP) is a distributed-graphics program which computes particle traces for computational fluid dynamics (CFD) solution data sets. A particle trace is a line which shows the path a massless particle in a fluid will take; it is a visual image of where the fluid is going. The program is able to compute and display particle traces at a speed of about one trace per second because it runs on two machines concurrently. The data used by the program is contained in two files. The solution file contains data on density, momentum and energy quantities of a flow field at discrete points in three-dimensional space, while the grid file contains the physical coordinates of each of the discrete points. RIP requires two computers. A local graphics workstation interfaces with the user for program control and graphics manipulation, and a remote machine interfaces with the solution data set and performs time-intensive computations. The program utilizes two machines in a distributed mode for two reasons. First, the data to be used by the program is usually generated on the supercomputer. RIP avoids having to convert and transfer the data, eliminating any memory limitations of the local machine. Second, as computing the particle traces can be computationally expensive, RIP utilizes the power of the supercomputer for this task. Although the remote site code was developed on a CRAY, it is possible to port this to any supercomputer class machine with a UNIX-like operating system. Integration of a velocity field from a starting physical location produces the particle trace. The remote machine computes the particle traces using the particle-tracing subroutines from PLOT3D/AMES, a CFD post-processing graphics program available from COSMIC (ARC-12779). These routines use a second-order predictor-corrector method to integrate the velocity field. Then the remote program sends graphics tokens to the local machine via a remote-graphics library. The local machine interprets the graphics tokens and draws the particle traces. The program is menu driven. RIP is implemented on the silicon graphics IRIS 3000 (local workstation) with an IRIX operating system and on the CRAY2 (remote station) with a UNICOS 1.0 or 2.0 operating system. The IRIS 4D can be used in place of the IRIS 3000. The program is written in C (67%) and FORTRAN 77 (43%) and has an IRIS memory requirement of 4 MB. The remote and local stations must use the same user ID. PLOT3D/AMES unformatted data sets are required for the remote machine. The program was developed in 1988.

Description: This software package includes two programs, the KPD12 and the KPD12P. Both programs utilizes the vortex-blob method to simulate flow around solid bodies, in an unbounded domain using the KPD12, with periodicity in one direction using the KPD12P. The main advantage of the vortex-blob method is the ability to handle situations involving arbitrary shapes including multiple bodies. The user just supplies points on the solid boundaries; there is no grid. The KPD12 program has worked successfully on bluff bodies, stalled wings, and multiple-element airfoils. The KPD12P program has been used successfully on high-solidity separated cascades and on cases of rotating stall in cascades of thin airfoils. However, they do not capture subtle viscous effects such as incipient separation and friction drag. The KPD12 and the KPD12P programs apply the vortex-blob method to time-dependent, high-Reynolds-number flows around solid bodies. Both programs solve the two-dimensional incompressible Navier-Stokes equations, neglecting the viscous effects away from the walls. By creating new vortices along the wall at every time step, they treat the no-penetration and no-slip boundary conditions while using an influence matrix. The code automatically controls the number of vortices. Furthermore, the code has the option of treating the boundary layers by simple integral methods to determine the separation points. The KPD12 outputs forces, moments, and pressure distributions on the bodies. The KPD12P also outputs the turning angle and loss of total pressure. The source code is in Cray FORTRAN and contains a few calls to Cray vector functions which are vectorized with the Cray compiler. However, substitutes for these vector functions are provided. The code is set up to plot the bodies, vortex positions, and streamlines using the DISSPLA graphics software. The software requires a mainframe computer with at least 589k of memory available running under COS 1.16. KPD12 was developed in 1988.

Description: The Steady State Thermal Analysis Program (STEADY) provides the thermal designer with a quick and convenient method for calculating heat loads and temperatures. STEADY can be used on small nodal networks for conceptual or preliminary thermal design and analysis. STEADY will accept up to 20 nodes of fixed or variable temperature, with constant or temperature-dependent thermal conductivities, and any set of consistent units. In a steady state thermal network, the heat balance on each variable temperature node must sum to zero. The general heat transfer equations are solved with a Newton-Raphson technique and refined by a fourth order quartic solution. Input data includes the number of nodes, number of boundary nodes, the fixed temperatures at all boundary nodes, initial temperature guesses for variable nodes, impressed heat loads, conduction and radiation coefficients, and control parameters such as convergence criteria, maximum iterations, and damping factors. The output is stored in a print file and tabulates final temperatures and heat flows for all nodes. STEADY is menu driven and allows the user to save files for future modification. STEADY is written in FORTRAN 77 (Ryan McFarland's RMFORTRAN) for interactive execution and has been implemented on the IBM PC computer series under DOS with a central memory requirement of approximately 92K of 8 bit bytes using a math coprocessor, and 103K bytes without the coprocessor. This program was developed in 1987.

Description: The Composite Plate Buckling Analysis Program (COMPPAP) was written to help engineers determine buckling loads of orthotropic (or isotropic) irregularly shaped plates without requiring hand calculations from design curves or extensive finite element modeling. COMPPAP is a one element finite element program that utilizes high-order displacement functions. The high order of the displacement functions enables the user to produce results more accurate than traditional h-finite elements. This program uses these high-order displacement functions to perform a plane stress analysis of a general plate followed by a buckling calculation based on the stresses found in the plane stress solution. The current version assumes a flat plate (constant thickness) subject to a constant edge load (normal or shear) on one or more edges. COMPPAP uses the power method to find the eigenvalues of the buckling problem. The power method provides an efficient solution when only one eigenvalue is desired. Once the eigenvalue is found, the eigenvector, which corresponds to the plate buckling mode shape, results as a by-product. A positive feature of the power method is that the dominant eigenvalue is the first found, which is this case is the plate buckling load. The reported eigenvalue expresses a load factor to induce plate buckling. COMPPAP is written in ANSI FORTRAN 77. Two machine versions are available from COSMIC: a PC version (MSC-22428), which is for IBM PC 386 series and higher computers and compatibles running MS-DOS; and a UNIX version (MSC-22286). The distribution medium for both machine versions includes source code for both single and double precision versions of COMPPAP. The PC version includes source code which has been optimized for implementation within DOS memory constraints as well as sample executables for both the single and double precision versions of COMPPAP. The double precision versions of COMPPAP have been successfully implemented on an IBM PC 386 compatible running MS-DOS, a Sun4 series computer running SunOS, an HP-9000 series computer running HP-UX, and a CRAY X-MP series computer running UNICOS. COMPPAP requires 1Mb of RAM and the BLAS and LINPACK math libraries, which are included on the distribution medium. The COMPPAP documentation provides instructions for using the commercial post-processing package PATRAN for graphical interpretation of COMPPAP output. The UNIX version includes two electronic versions of the documentation: one in LaTex format and one in PostScript format. The standard distribution medium for the PC version (MSC-22428) is a 5.25 inch 1.2Mb MS-DOS format diskette. The standard distribution medium for the UNIX version (MSC-22286) is a .25 inch streaming magnetic tape cartridge (Sun QIC-24) in UNIX tar format. For the UNIX version, alternate distribution media and formats are available upon request. COMPPAP was developed in 1992.

Description: The Composite Plate Buckling Analysis Program (COMPPAP) was written to help engineers determine buckling loads of orthotropic (or isotropic) irregularly shaped plates without requiring hand calculations from design curves or extensive finite element modeling. COMPPAP is a one element finite element program that utilizes high-order displacement functions. The high order of the displacement functions enables the user to produce results more accurate than traditional h-finite elements. This program uses these high-order displacement functions to perform a plane stress analysis of a general plate followed by a buckling calculation based on the stresses found in the plane stress solution. The current version assumes a flat plate (constant thickness) subject to a constant edge load (normal or shear) on one or more edges. COMPPAP uses the power method to find the eigenvalues of the buckling problem. The power method provides an efficient solution when only one eigenvalue is desired. Once the eigenvalue is found, the eigenvector, which corresponds to the plate buckling mode shape, results as a by-product. A positive feature of the power method is that the dominant eigenvalue is the first found, which is this case is the plate buckling load. The reported eigenvalue expresses a load factor to induce plate buckling. COMPPAP is written in ANSI FORTRAN 77. Two machine versions are available from COSMIC: a PC version (MSC-22428), which is for IBM PC 386 series and higher computers and compatibles running MS-DOS; and a UNIX version (MSC-22286). The distribution medium for both machine versions includes source code for both single and double precision versions of COMPPAP. The PC version includes source code which has been optimized for implementation within DOS memory constraints as well as sample executables for both the single and double precision versions of COMPPAP. The double precision versions of COMPPAP have been successfully implemented on an IBM PC 386 compatible running MS-DOS, a Sun4 series computer running SunOS, an HP-9000 series computer running HP-UX, and a CRAY X-MP series computer running UNICOS. COMPPAP requires 1Mb of RAM and the BLAS and LINPACK math libraries, which are included on the distribution medium. The COMPPAP documentation provides instructions for using the commercial post-processing package PATRAN for graphical interpretation of COMPPAP output. The UNIX version includes two electronic versions of the documentation: one in LaTex format and one in PostScript format. The standard distribution medium for the PC version (MSC-22428) is a 5.25 inch 1.2Mb MS-DOS format diskette. The standard distribution medium for the UNIX version (MSC-22286) is a .25 inch streaming magnetic tape cartridge (Sun QIC-24) in UNIX tar format. For the UNIX version, alternate distribution media and formats are available upon request. COMPPAP was developed in 1992.

Description: Structural flaws and cracks may grow under fatigue inducing loads and, upon reaching a critical size, cause structural failure to occur. The growth of these flaws and cracks may occur at load levels well below the ultimate load bearing capability of the structure. The Fatigue Crack Growth Computer Program, NASA/FLAGRO, was developed as an aid in predicting the growth of pre-existing flaws and cracks in structural components of space systems. The earlier version of the program, FLAGRO4, was the primary analysis tool used by Rockwell International and the Shuttle subcontractors for fracture control analysis on the Space Shuttle. NASA/FLAGRO is an enhanced version of the program and incorporates state-of-the-art improvements in both fracture mechanics and computer technology. NASA/FLAGRO provides the fracture mechanics analyst with a computerized method of evaluating the "safe crack growth life" capabilities of structural components. NASA/FLAGRO could also be used to evaluate the damage tolerance aspects of a given structural design. The propagation of an existing crack is governed by the stress field in the vicinity of the crack tip. The stress intensity factor is defined in terms of the relationship between the stress field magnitude and the crack size. The propagation of the crack becomes catastrophic when the local stress intensity factor reaches the fracture toughness of the material. NASA/FLAGRO predicts crack growth using a two-dimensional model which predicts growth independently in two directions based on the calculation of stress intensity factors. The analyst can choose to use either a crack growth rate equation or a nonlinear interpolation routine based on tabular data. The growth rate equation is a modified Forman equation which can be converted to a Paris or Walker equation by substituting different values into the exponent. This equation provides accuracy and versatility and can be fit to data using standard least squares methods. Stress-intensity factor numerical values can be computed for making comparisons or checks of solutions. NASA/FLAGRO can check for failure of a part-through crack in the mode of a through crack when net ligament yielding occurs. NASA/FLAGRO has a number of special subroutines and files which provide enhanced capabilities and easy entry of data. These include crack case solutions, cyclic load spectrums, nondestructive examination initial flaw sizes, table interpolation, and material properties. The materials properties files are divided into two types, a user defined file and a fixed file. Data is entered and stored in the user defined file during program execution, while the fixed file contains already coded-in property value data for many different materials. Prompted input from CRT terminals consists of initial crack definition (which can be defined automatically), rate solution type, flaw type and geometry, material properties (if they are not in the built-in tables of material data), load spectrum data (if not included in the loads spectrum file), and design limit stress levels. NASA/FLAGRO output includes an echo of the input with any error or warning messages, the final crack size, whether or not critical crack size has been reached for the specified stress level, and a life history profile of the crack propagation. NASA/FLAGRO is modularly designed to facilitate revisions and operation on minicomputers. The program was implemented on a DEC VAX 11/780 with the VMS operating system. NASA/FLAGRO is written in FORTRAN77 and has a memory requirement of 1.4 MB. The program was developed in 1986.

Description: The Thermal Radiation Analyzer System, TRASYS, is a computer software system with generalized capability to solve the radiation related aspects of thermal analysis problems. TRASYS computes the total thermal radiation environment for a spacecraft in orbit. The software calculates internode radiation interchange data as well as incident and absorbed heat rate data originating from environmental radiant heat sources. TRASYS provides data of both types in a format directly usable by such thermal analyzer programs as SINDA/FLUINT (available from COSMIC, program number MSC-21528). One primary feature of TRASYS is that it allows users to write their own driver programs to organize and direct the preprocessor and processor library routines in solving specific thermal radiation problems. The preprocessor first reads and converts the user's geometry input data into the form used by the processor library routines. Then, the preprocessor accepts the user's driving logic, written in the TRASYS modified FORTRAN language. In many cases, the user has a choice of routines to solve a given problem. Users may also provide their own routines where desirable. In particular, the user may write output routines to provide for an interface between TRASYS and any thermal analyzer program using the R-C network concept. Input to the TRASYS program consists of Options and Edit data, Model data, and Logic Flow and Operations data. Options and Edit data provide for basic program control and user edit capability. The Model data describe the problem in terms of geometry and other properties. This information includes surface geometry data, documentation data, nodal data, block coordinate system data, form factor data, and flux data. Logic Flow and Operations data house the user's driver logic, including the sequence of subroutine calls and the subroutine library. Output from TRASYS consists of two basic types of data: internode radiation interchange data, and incident and absorbed heat rate data. The flexible structure of TRASYS allows considerable freedom in the definition and choice of solution method for a thermal radiation problem. The program's flexible structure has also allowed TRASYS to retain the same basic input structure as the authors update it in order to keep up with changing requirements. Among its other important features are the following: 1) up to 3200 node problem size capability with shadowing by intervening opaque or semi-transparent surfaces; 2) choice of diffuse, specular, or diffuse/specular radiant interchange solutions; 3) a restart capability that minimizes recomputing; 4) macroinstructions that automatically provide the executive logic for orbit generation that optimizes the use of previously completed computations; 5) a time variable geometry package that provides automatic pointing of the various parts of an articulated spacecraft and an automatic look-back feature that eliminates redundant form factor calculations; 6) capability to specify submodel names to identify sets of surfaces or components as an entity; and 7) subroutines to perform functions which save and recall the internodal and/or space form factors in subsequent steps for nodes with fixed geometry during a variable geometry run. There are two machine versions of TRASYS v27: a DEC VAX version and a Cray UNICOS version. Both versions require installation of the NASADIG library (MSC-21801 for DEC VAX or COS-10049 for CRAY), which is available from COSMIC either separately or bundled with TRASYS. The NASADIG (NASA Device Independent Graphics Library) plot package provides a pictorial representation of input geometry, orbital/orientation parameters, and heating rate output as a function of time. NASADIG supports Tektronix terminals. The CRAY version of TRASYS v27 is written in FORTRAN 77 for batch or interactive execution and has been implemented on CRAY X-MP and CRAY Y-MP series computers running UNICOS. The standard distribution medium for MSC-21959 (CRAY version without NASADIG) is a 1600 BPI 9-track magnetic tape in UNIX tar format. The standard distribution medium for COS-10040 (CRAY version with NASADIG) is a set of two 6250 BPI 9-track magnetic tapes in UNIX tar format. Alternate distribution media and formats are available upon request. The DEC VAX version of TRASYS v27 is written in FORTRAN 77 for batch execution (only the plotting driver program is interactive) and has been implemented on a DEC VAX 8650 computer under VMS. Since the source codes for MSC-21030 and COS-10026 are in VAX/VMS text library files and DEC Command Language files, COSMIC will only provide these programs in the following formats: MSC-21030, TRASYS (DEC VAX version without NASADIG) is available on a 1600 BPI 9-track magnetic tape in VAX BACKUP format (standard distribution medium) or in VAX BACKUP format on a TK50 tape cartridge; COS-10026, TRASYS (DEC VAX version with NASADIG), is available in VAX BACKUP format on a set of three 6250 BPI 9-track magnetic tapes (standard distribution medium) or a set of three TK50 tape cartridges in VAX BACKUP format. TRASYS was last updated in 1993.

Description: TDIGG is a fast and versatile program for generating two-dimensional computational grids for use with finite-difference flow-solvers. Both algebraic and elliptic grid generation systems are included. The method for grid generation by algebraic transformation is based on an interpolation algorithm and the elliptic grid generation is established by solving the partial differential equation (PDE). Non-uniform grid distributions are carried out using a hyperbolic tangent stretching function. For algebraic grid systems, interpolations in one direction (univariate) and two directions (bivariate) are considered. These interpolations are associated with linear or cubic Lagrangian/Hermite/Bezier polynomial functions. The algebraic grids can subsequently be smoothed using an elliptic solver. For elliptic grid systems, the PDE can be in the form of Laplace (zero forcing function) or Poisson. The forcing functions in the Poisson equation come from the boundary or the entire domain of the initial algebraic grids. A graphics interface procedure using the Silicon Graphics (GL) Library is included to allow users to visualize the grid variations at each iteration. This will allow users to interactively modify the grid to match their applications. TDIGG is written in FORTRAN 77 for Silicon Graphics IRIS series computers running IRIX. This package requires either MIT's X Window System, Version 11 Revision 4 or SGI (Motif) Window System. A sample executable is provided on the distribution medium. It requires 148K of RAM for execution. The standard distribution medium is a .25 inch streaming magnetic IRIX tape cartridge in UNIX tar format. This program was developed in 1992.

Description: The Systems Improved Numerical Fluids Analysis Code, SINFAC, consists of additional routines added to the April 1983 revision of SINDA, a general thermal analyzer program. The purpose of the additional routines is to allow for the modeling of active heat transfer loops. The modeler can simulate the steady-state and pseudo-transient operations of 16 different heat transfer loop components including radiators, evaporators, condensers, mechanical pumps, reservoirs and many types of valves and fittings. In addition, the program contains a property analysis routine that can be used to compute the thermodynamic properties of 20 different refrigerants. SINFAC can simulate the response to transient boundary conditions. SINFAC was first developed as a method for computing the steady-state performance of two phase systems. It was then modified using CNFRWD, SINDA's explicit time-integration scheme, to accommodate transient thermal models. However, SINFAC cannot simulate pressure drops due to time-dependent fluid acceleration, transient boil-out, or transient fill-up, except in the accumulator. SINFAC also requires the user to be familiar with SINDA. The solution procedure used by SINFAC is similar to that which an engineer would use to solve a system manually. The solution to a system requires the determination of all of the outlet conditions of each component such as the flow rate, pressure, and enthalpy. To obtain these values, the user first estimates the inlet conditions to the first component of the system, then computes the outlet conditions from the data supplied by the manufacturer of the first component. The user then estimates the temperature at the outlet of the third component and computes the corresponding flow resistance of the second component. With the flow resistance of the second component, the user computes the conditions down stream, namely the inlet conditions of the third. The computations follow for the rest of the system, back to the first component. On the first pass, the user finds that the calculated outlet conditions of the last component do not match the estimated inlet conditions of the first. The user then modifies the estimated inlet conditions of the first component in an attempt to match the calculated values. The user estimated values are called State Variables. The differences between the user estimated values and calculated values are called the Error Variables. The procedure systematically changes the State Variables until all of the Error Variables are less than the user-specified iteration limits. The solution procedure is referred to as SCX. It consists of two phases, the Systems phase and the Controller phase. The X is to imply experimental. SCX computes each next set of State Variables in two phases. In the first phase, SCX fixes the controller positions and modifies the other State Variables by the Newton-Raphson method. This first phase is the Systems phase. Once the Newton-Raphson method has solved the problem for the fixed controller positions, SCX next calculates new controller positions based on Newton's method while treating each sensor-controller pair independently but allowing all to change in one iteration. This phase is the Controller phase. SINFAC is available by license for a period of ten (10) years to approved licensees. The licenced program product includes the source code for the additional routines to SINDA, the SINDA object code, command procedures, sample data and supporting documentation. Additional documentation may be purchased at the price below. SINFAC was created for use on a DEC VAX under VMS. Source code is written in FORTRAN 77, requires 180k of memory, and should be fully transportable. The program was developed in 1988.

Description: MSIS is an empirical model of the thermosphere based on Mass Spectrometer and Incoherent Scatter data. It provides a description of atmospheric temperature, density, and composition for altitudes higher than 85 kilometers. There are coefficients in MSIS to account for yearly and daily variations, geodetic latitude and longitude, altitude, and solar activity. Variations due to magnetic storms are represented by three-hour magnetic ap indices. The MSIS model enables a more timely prediction of aeronomic densities for specific events, such as rocket flights. The database for this model is a comprehensive summary of rocket flight, satellite, incoherent scatter radar, grenade, and falling sphere measurements. Subsets of data were formed by random selection after sorting on altitude, latitude, time of day, etc. Curve fitting was done with four to five thousand data points at a time. The resulting coefficients are presented in subroutines which calculate thermospheric composition and temperature for a user-supplied position and time. MSIS is written in FORTRAN 77 for use with batch or interactive programs and has been implemented on a DEC VAX series computer operating under VMS 4.3 with a central memory requirement of approximately 25K of 8 bit bytes. MSIS is based on a 1977 thermosphere model and was last updated in 1987 to reflect the CIRA 1986 Neutral Thermosphere Model.

Description: The Thermal Radiation Analyzer System, TRASYS, is a computer software system with generalized capability to solve the radiation related aspects of thermal analysis problems. TRASYS computes the total thermal radiation environment for a spacecraft in orbit. The software calculates internode radiation interchange data as well as incident and absorbed heat rate data originating from environmental radiant heat sources. TRASYS provides data of both types in a format directly usable by such thermal analyzer programs as SINDA/FLUINT (available from COSMIC, program number MSC-21528). One primary feature of TRASYS is that it allows users to write their own driver programs to organize and direct the preprocessor and processor library routines in solving specific thermal radiation problems. The preprocessor first reads and converts the user's geometry input data into the form used by the processor library routines. Then, the preprocessor accepts the user's driving logic, written in the TRASYS modified FORTRAN language. In many cases, the user has a choice of routines to solve a given problem. Users may also provide their own routines where desirable. In particular, the user may write output routines to provide for an interface between TRASYS and any thermal analyzer program using the R-C network concept. Input to the TRASYS program consists of Options and Edit data, Model data, and Logic Flow and Operations data. Options and Edit data provide for basic program control and user edit capability. The Model data describe the problem in terms of geometry and other properties. This information includes surface geometry data, documentation data, nodal data, block coordinate system data, form factor data, and flux data. Logic Flow and Operations data house the user's driver logic, including the sequence of subroutine calls and the subroutine library. Output from TRASYS consists of two basic types of data: internode radiation interchange data, and incident and absorbed heat rate data. The flexible structure of TRASYS allows considerable freedom in the definition and choice of solution method for a thermal radiation problem. The program's flexible structure has also allowed TRASYS to retain the same basic input structure as the authors update it in order to keep up with changing requirements. Among its other important features are the following: 1) up to 3200 node problem size capability with shadowing by intervening opaque or semi-transparent surfaces; 2) choice of diffuse, specular, or diffuse/specular radiant interchange solutions; 3) a restart capability that minimizes recomputing; 4) macroinstructions that automatically provide the executive logic for orbit generation that optimizes the use of previously completed computations; 5) a time variable geometry package that provides automatic pointing of the various parts of an articulated spacecraft and an automatic look-back feature that eliminates redundant form factor calculations; 6) capability to specify submodel names to identify sets of surfaces or components as an entity; and 7) subroutines to perform functions which save and recall the internodal and/or space form factors in subsequent steps for nodes with fixed geometry during a variable geometry run. There are two machine versions of TRASYS v27: a DEC VAX version and a Cray UNICOS version. Both versions require installation of the NASADIG library (MSC-21801 for DEC VAX or COS-10049 for CRAY), which is available from COSMIC either separately or bundled with TRASYS. The NASADIG (NASA Device Independent Graphics Library) plot package provides a pictorial representation of input geometry, orbital/orientation parameters, and heating rate output as a function of time. NASADIG supports Tektronix terminals. The CRAY version of TRASYS v27 is written in FORTRAN 77 for batch or interactive execution and has been implemented on CRAY X-MP and CRAY Y-MP series computers running UNICOS. The standard distribution medium for MSC-21959 (CRAY version without NASADIG) is a 1600 BPI 9-track magnetic tape in UNIX tar format. The standard distribution medium for COS-10040 (CRAY version with NASADIG) is a set of two 6250 BPI 9-track magnetic tapes in UNIX tar format. Alternate distribution media and formats are available upon request. The DEC VAX version of TRASYS v27 is written in FORTRAN 77 for batch execution (only the plotting driver program is interactive) and has been implemented on a DEC VAX 8650 computer under VMS. Since the source codes for MSC-21030 and COS-10026 are in VAX/VMS text library files and DEC Command Language files, COSMIC will only provide these programs in the following formats: MSC-21030, TRASYS (DEC VAX version without NASADIG) is available on a 1600 BPI 9-track magnetic tape in VAX BACKUP format (standard distribution medium) or in VAX BACKUP format on a TK50 tape cartridge; COS-10026, TRASYS (DEC VAX version with NASADIG), is available in VAX BACKUP format on a set of three 6250 BPI 9-track magnetic tapes (standard distribution medium) or a set of three TK50 tape cartridges in VAX BACKUP format. TRASYS was last updated in 1993.

Description: INS3D computes steady-state solutions to the incompressible Navier-Stokes equations. The INS3D approach utilizes pseudo-compressibility combined with an approximate factorization scheme. This computational fluid dynamics (CFD) code has been verified on problems such as flow through a channel, flow over a backwardfacing step and flow over a circular cylinder. Three dimensional cases include flow over an ogive cylinder, flow through a rectangular duct, wind tunnel inlet flow, cylinder-wall juncture flow and flow through multiple posts mounted between two plates. INS3D uses a pseudo-compressibility approach in which a time derivative of pressure is added to the continuity equation, which together with the momentum equations form a set of four equations with pressure and velocity as the dependent variables. The equations' coordinates are transformed for general three dimensional applications. The equations are advanced in time by the implicit, non-iterative, approximately-factored, finite-difference scheme of Beam and Warming. The numerical stability of the scheme depends on the use of higher-order smoothing terms to damp out higher-frequency oscillations caused by second-order central differencing. The artificial compressibility introduces pressure (sound) waves of finite speed (whereas the speed of sound would be infinite in an incompressible fluid). As the solution converges, these pressure waves die out, causing the derivation of pressure with respect to time to approach zero. Thus, continuity is satisfied for the incompressible fluid in the steady state. Computational efficiency is achieved using a diagonal algorithm. A block tri-diagonal option is also available. When a steady-state solution is reached, the modified continuity equation will satisfy the divergence-free velocity field condition. INS3D is capable of handling several different types of boundaries encountered in numerical simulations, including solid-surface, inflow and outflow, and far-field boundaries. Three machine versions of INS3D are available. INS3D for the CRAY is written in CRAY FORTRAN for execution on a CRAY X-MP under COS, INS3D for the IBM is written in FORTRAN 77 for execution on an IBM 3090 under the VM or MVS operating system, and INS3D for DEC RISC-based systems is written in RISC FORTRAN for execution on a DEC workstation running RISC ULTRIX 3.1 or later. The CRAY version has a central memory requirement of 730279 words. The central memory requirement for the IBM is 150Mb. The memory requirement for the DEC RISC ULTRIX version is 3Mb of main memory. INS3D was developed in 1987. The port to the IBM was done in 1990. The port to the DECstation 3100 was done in 1991. CRAY is a registered trademark of Cray Research Inc. IBM is a registered trademark of International Business Machines. DEC, DECstation, and ULTRIX are trademarks of the Digital Equipment Corporation.

Description: SAGE, Self Adaptive Grid codE, is a flexible tool for adapting and restructuring both 2D and 3D grids. Solution-adaptive grid methods are useful tools for efficient and accurate flow predictions. In supersonic and hypersonic flows, strong gradient regions such as shocks, contact discontinuities, shear layers, etc., require careful distribution of grid points to minimize grid error and produce accurate flow-field predictions. SAGE helps the user obtain more accurate solutions by intelligently redistributing (i.e. adapting) the original grid points based on an initial or interim flow-field solution. The user then computes a new solution using the adapted grid as input to the flow solver. The adaptive-grid methodology poses the problem in an algebraic, unidirectional manner for multi-dimensional adaptations. The procedure is analogous to applying tension and torsion spring forces proportional to the local flow gradient at every grid point and finding the equilibrium position of the resulting system of grid points. The multi-dimensional problem of grid adaption is split into a series of one-dimensional problems along the computational coordinate lines. The reduced one dimensional problem then requires a tridiagonal solver to find the location of grid points along a coordinate line. Multi-directional adaption is achieved by the sequential application of the method in each coordinate direction. The tension forces direct the redistribution of points to the strong gradient region. To maintain smoothness and a measure of orthogonality of grid lines, torsional forces are introduced that relate information between the family of lines adjacent to one another. The smoothness and orthogonality constraints are direction-dependent, since they relate only the coordinate lines that are being adapted to the neighboring lines that have already been adapted. Therefore the solutions are non-unique and depend on the order and direction of adaption. Non-uniqueness of the adapted grid is acceptable since it makes possible an overall and local error reduction through grid redistribution. SAGE includes the ability to modify the adaption techniques in boundary regions, which substantially improves the flexibility of the adaptive scheme. The vectorial approach used in the analysis also provides flexibility. The user has complete choice of adaption direction and order of sequential adaptions without concern for the computational data structure. Multiple passes are available with no restraint on stepping directions; for each adaptive pass the user can choose a completely new set of adaptive parameters. This facility, combined with the capability of edge boundary control, enables the code to individually adapt multi-dimensional multiple grids. Zonal grids can be adapted while maintaining continuity along the common boundaries. For patched grids, the multiple-pass capability enables complete adaption. SAGE is written in FORTRAN 77 and is intended to be machine independent; however, it requires a FORTRAN compiler which supports NAMELIST input. It has been successfully implemented on Sun series computers, SGI IRIS's, DEC MicroVAX computers, HP series computers, the Cray YMP, and IBM PC compatibles. Source code is provided, but no sample input and output files are provided. The code reads three datafiles: one that contains the initial grid coordinates (x,y,z), one that contains corresponding flow-field variables, and one that contains the user control parameters. It is assumed that the first two datasets are formatted as defined in the plotting software package PLOT3D. Several machine versions of PLOT3D are available from COSMIC. The amount of main memory is dependent on the size of the matrix. The standard distribution medium for SAGE is a 5.25 inch 360K MS-DOS format diskette. It is also available on a .25 inch streaming magnetic tape cartridge in UNIX tar format or on a 9-track 1600 BPI ASCII CARD IMAGE format magnetic tape. SAGE was developed in 1989, first released as a 2D version in 1991 and updated to 3D in 1993.

Description: Delaminations near the outer surface of a laminate are susceptible to local buckling and buckling-induced delamination propagation when the laminate is subjected to transverse impact loading. This results in a loss of stiffness and strength. TRBUCKL is an unique dynamic delamination buckling and delamination propagation analysis capability that can be incorporated into the structural analysis program, NASTRAN. This capability will aid engineers in the design of structures incorporating composite laminates. The capability consists of: (1) a modification of the direct time integration solution sequence which provides a new analysis algorithm that can be used to predict delamination buckling in a laminate subjected to dynamic loading; and (2) a new method of modeling the composite laminate using plate bending elements and multipoint constraints. The capability now exists to predict the time at which the onset of dynamic delamination buckling occurs, the dynamic buckling mode shape, and the dynamic delamination strain energy release rate. A procedure file for NASTRAN, TRBUCKL predicts both impact induced buckling in composite laminates with initial delaminations and the strain energy release rate due to extension of the delamination. In addition, the file is useful in calculating the dynamic delamination strain energy release rate for a composite laminate under impact loading. This procedure simplifies the simulation of progressive crack extension. TRBUCKL has been incorporated into COSMIC NASTRAN. TRBUCKL is a DMAP Alter for NASTRAN. It is intended for use only with the COSMIC NASTRAN Direct Transient Analysis (RF 9) solution sequence. The program is available as a listing only. TRBUCKL was developed in 1987.

Description: INS3D computes steady-state solutions to the incompressible Navier-Stokes equations. The INS3D approach utilizes pseudo-compressibility combined with an approximate factorization scheme. This computational fluid dynamics (CFD) code has been verified on problems such as flow through a channel, flow over a backwardfacing step and flow over a circular cylinder. Three dimensional cases include flow over an ogive cylinder, flow through a rectangular duct, wind tunnel inlet flow, cylinder-wall juncture flow and flow through multiple posts mounted between two plates. INS3D uses a pseudo-compressibility approach in which a time derivative of pressure is added to the continuity equation, which together with the momentum equations form a set of four equations with pressure and velocity as the dependent variables. The equations' coordinates are transformed for general three dimensional applications. The equations are advanced in time by the implicit, non-iterative, approximately-factored, finite-difference scheme of Beam and Warming. The numerical stability of the scheme depends on the use of higher-order smoothing terms to damp out higher-frequency oscillations caused by second-order central differencing. The artificial compressibility introduces pressure (sound) waves of finite speed (whereas the speed of sound would be infinite in an incompressible fluid). As the solution converges, these pressure waves die out, causing the derivation of pressure with respect to time to approach zero. Thus, continuity is satisfied for the incompressible fluid in the steady state. Computational efficiency is achieved using a diagonal algorithm. A block tri-diagonal option is also available. When a steady-state solution is reached, the modified continuity equation will satisfy the divergence-free velocity field condition. INS3D is capable of handling several different types of boundaries encountered in numerical simulations, including solid-surface, inflow and outflow, and far-field boundaries. Three machine versions of INS3D are available. INS3D for the CRAY is written in CRAY FORTRAN for execution on a CRAY X-MP under COS, INS3D for the IBM is written in FORTRAN 77 for execution on an IBM 3090 under the VM or MVS operating system, and INS3D for DEC RISC-based systems is written in RISC FORTRAN for execution on a DEC workstation running RISC ULTRIX 3.1 or later. The CRAY version has a central memory requirement of 730279 words. The central memory requirement for the IBM is 150Mb. The memory requirement for the DEC RISC ULTRIX version is 3Mb of main memory. INS3D was developed in 1987. The port to the IBM was done in 1990. The port to the DECstation 3100 was done in 1991. CRAY is a registered trademark of Cray Research Inc. IBM is a registered trademark of International Business Machines. DEC, DECstation, and ULTRIX are trademarks of the Digital Equipment Corporation.

Description: The ability to treat arbitrary boundary shapes is one of the most desirable characteristics of a method for generating grids, including those about airfoils. In a grid used for computing aerodynamic flow over an airfoil, or any other body shape, the surface of the body is usually treated as an inner boundary and often cannot be easily represented as an analytic function. The GRAPE computer program was developed to incorporate a method for generating two-dimensional finite-difference grids about airfoils and other shapes by the use of the Poisson differential equation. GRAPE can be used with any boundary shape, even one specified by tabulated points and including a limited number of sharp corners. The GRAPE program has been developed to be numerically stable and computationally fast. GRAPE can provide the aerodynamic analyst with an efficient and consistent means of grid generation. The GRAPE procedure generates a grid between an inner and an outer boundary by utilizing an iterative procedure to solve the Poisson differential equation subject to geometrical restraints. In this method, the inhomogeneous terms of the equation are automatically chosen such that two important effects are imposed on the grid. The first effect is control of the spacing between mesh points along mesh lines intersecting the boundaries. The second effect is control of the angles with which mesh lines intersect the boundaries. Along with the iterative solution to Poisson's equation, a technique of coarse-fine sequencing is employed to accelerate numerical convergence. GRAPE program control cards and input data are entered via the NAMELIST feature. Each variable has a default value such that user supplied data is kept to a minimum. Basic input data consists of the boundary specification, mesh point spacings on the boundaries, and mesh line angles at the boundaries. Output consists of a dataset containing the grid data and, if requested, a plot of the generated mesh. The GRAPE program is written in FORTRAN IV for batch execution and has been implemented on a CDC 6000 series computer with a central memory requirement of approximately 135K (octal) of 60 bit words. For plotted output the commercially available DISSPLA graphics software package is required. The GRAPE program was developed in 1980.

Description: Predictions of fatigue crack growth behavior can be made with the Fatigue Crack Growth Structural Analysis (FASTRAN II) computer program. As cyclic loads are applied to a selected crack configuration with an initial crack size, FASTRAN II predicts crack growth as a function of cyclic load history until either a desired crack size is reached or failure occurs. FASTRAN II is based on plasticity-induced crack-closure behavior of cracks in metallic materials and accounts for load-interaction effects, such as retardation and acceleration, under variable-amplitude loading. The closure model is based on the Dugdale model with modifications to allow plastically deformed material to be left along the crack surfaces as the crack grows. Plane stress and plane strain conditions, as well as conditions between these two, can be simulated in FASTRAN II by using a constraint factor on tensile yielding at the crack front to approximately account for three-dimensional stress states. FASTRAN II contains seventeen predefined crack configurations (standard laboratory fatigue crack growth rate specimens and many common crack configurations found in structures); and the user can define one additional crack configuration. The baseline crack growth rate properties (effective stress-intensity factor against crack growth rate) may be given in either equation or tabular form. For three-dimensional crack configurations, such as surface cracks or corner cracks at holes or notches, the fatigue crack growth rate properties may be different in the crack depth and crack length directions. Final failure of the cracked structure can be modelled with fracture toughness properties using either linear-elastic fracture mechanics (brittle materials), a two-parameter fracture criterion (brittle to ductile materials), or plastic collapse (extremely ductile materials). The crack configurations in FASTRAN II can be subjected to either constant-amplitude, variable-amplitude or spectrum loading. The applied loads may be either tensile or compressive. Several standardized aircraft flight-load histories, such as TWIST, Mini-TWIST, FALSTAFF, Inverted FALSTAFF, Felix and Gaussian, are included as options. FASTRAN II also includes two other methods that will help the user input spectrum load histories. The two methods are: (1) a list of stress points, and (2) a flight-by-flight history of stress points. Examples are provided in the user manual. Developed as a research program, FASTRAN II has successfully predicted crack growth in many metallic materials under various aircraft spectrum loading. A computer program DKEFF which is a part of the FASTRAN II package was also developed to analyze crack growth rate data from laboratory specimens to obtain the effective stress-intensity factor against crack growth rate relations used in FASTRAN II. FASTRAN II is written in standard FORTRAN 77. It has been successfully compiled and implemented on Sun4 series computers running SunOS and on IBM PC compatibles running MS-DOS using the Lahey F77L FORTRAN compiler. Sample input and output data are included with the FASTRAN II package. The UNIX version requires 660K of RAM for execution. The standard distribution medium for the UNIX version (LAR-14865) is a .25 inch streaming magnetic tape cartridge in UNIX tar format. It is also available on a 3.5 inch diskette in UNIX tar format. The standard distribution medium for the MS-DOS version (LAR-14944) is a 5.25 inch 360K MS-DOS format diskette. The contents of the diskette are compressed using the PKWARE archiving tools. The utility to unarchive the files, PKUNZIP.EXE, is included. The program was developed in 1984 and revised in 1992. Sun4 and SunOS are trademarks of Sun Microsystems, Inc. IBM PC is a trademark of International Business Machines Corp. MS-DOS is a trademark of Microsoft, Inc. F77L is a trademark of the Lahey Computer Systems, Inc. UNIX is a registered trademark of AT&T Bell Laboratories. PKWARE and PKUNZIP are trademarks of PKWare, Inc.

Description: The Panel Analysis and Sizing Code (PASCO) was developed for the buckling and vibration analysis and sizing of prismatic structures having an arbitrary cross section. PASCO is primarily intended for analyzing and sizing stiffened panels made of laminated orthotropic materials and is of particular value in analyzing and sizing filamentary composite structures. When used in the analysis mode, PASCO calculates laminate stiffnesses, lamina stress and strains (including the effects of temperature and panel bending), buckling loads, vibration frequencies, and overall panel stiffness. When used in the sizing mode, PASCO adjusts sizing variables to provide a low-mass panel design that carries a set of specified loadings without exceeding buckling or material strength allowables and that meets other design requirements such as upper and lower bounds on sizing variables, upper and lower bounds on overall bending, extensional and shear stiffnesses, and lower bounds on vibration frequencies. Although emphasis in PASCO is placed on flat panels having several identical bays, the only restriction on configuration modeling is that the structure is assumed to be prismatic. In addition, it is assumed that loads and temperatures do not vary along the length of a panel. Because of their wide application in aerospace structures, stiffened panels are readily handled by PASCO. The panel cross section may be composed of an arbitrary assemblage of thin, flat, rectangular plate elements that are connected together along their longitudinal edges. Each plate element consists of a balanced symmetric laminate of any number of layers of orthotropic material. Any group of element widths, layer thicknesses, and layer orientation angles may be selected as sizing variables. Substructuring is available to increase the efficiency of the analysis and to simplify the modeling of complicated structures. The Macintosh version of PASCO includes an interactive, graphic preprocessor called MacPASCO. The main objective of MacPASCO is to make the use of PASCO faster, simpler, and less error-prone. By using a graphical user interface (GUI), MacPASCO simplifies the specification of panel geometry and reduces user input errors, thus making the modeling and analysis of panel designs more efficient. The user draws the initial structural geometry on the computer screen, then uses a combination of graphic and text inputs to: refine the structural geometry, specify information required for analysis such as panel load conditions, and define design variables and constraints for minimum-mass optimization. Composite panel design is an ideal application because the graphical user interface can: serve as a visual aid, eliminate the tedious aspects of text-based input, and eliminate many sources of input errors. The current version of MacPASCO does not implement all the modeling features of PASCO, but has been found to be sufficient for many users. Many difficulties common to text-based inputs are avoided because MacPASCO uses a GUI. First, the graphic displays eliminate syntax errors, like misplaced commas and incorrect command names, because there is no text-based syntax. Second, graphic displays allow the user to see the geometry as it is created and immediately detect and correct any errors. Third, MacPASCO's drawing tools have been designed to avoid modeling errors. Fourth, the graphic displays make revisions to existing structural designs much easier and less error-prone by eliminating the need for the user to conceptualize the text input as geometry. The user can work directly with the geometry displayed on the screen. Finally, MacPASCO automatically generates the correct PASCO input from the geometry displayed on the screen. This input file can be used with any machine version of PASCO to actually perform the analysis and sizing and to output results. The DEC VAX version of PASCO is written in FORTRAN IV for batch execution and has been implemented on a DEC VAX series computer. The Macintosh version of PASCO was developed for Macintosh II series computers with at least 2Mb of RAM running MPW Pascal 3.0 and Language Systems FORTRAN 2.0 under the MPW programming environment. It includes MPW compatible makefiles for compiling the source code. The Macintosh version uses input files compatible with versions of PASCO running on different platforms. MacPASCO is written in Macintosh Programmers Workbench 3.0, MPW Pascal 3.0, and MacAPP 2.0. The Pascal source code is included on the distribution diskette. MacAPP is a development library which is not included. MacPASCO requires a Mac Plus, SE/30, or MacII, IIx, IIcx, IIci, or IIfx running System 6.0 or greater. MacPASCO is System 7.0 compatible. A minimum of 2Mb of RAM is required for execution. The Macintosh version of PASCO is distributed on four 3.5 inch 800K Macintosh format diskettes. The DEC VAX version is distributed on a 9-track 1600 BPI magnetic tape. The PASCO program was developed in 1981, adapted to the DEC VAX in 1983 and to the Macintosh in 1991. MacPASCO was released in 1992.

Description: The Thermal Radiation Analyzer System, TRASYS, is a computer software system with generalized capability to solve the radiation related aspects of thermal analysis problems. TRASYS computes the total thermal radiation environment for a spacecraft in orbit. The software calculates internode radiation interchange data as well as incident and absorbed heat rate data originating from environmental radiant heat sources. TRASYS provides data of both types in a format directly usable by such thermal analyzer programs as SINDA/FLUINT (available from COSMIC, program number MSC-21528). One primary feature of TRASYS is that it allows users to write their own driver programs to organize and direct the preprocessor and processor library routines in solving specific thermal radiation problems. The preprocessor first reads and converts the user's geometry input data into the form used by the processor library routines. Then, the preprocessor accepts the user's driving logic, written in the TRASYS modified FORTRAN language. In many cases, the user has a choice of routines to solve a given problem. Users may also provide their own routines where desirable. In particular, the user may write output routines to provide for an interface between TRASYS and any thermal analyzer program using the R-C network concept. Input to the TRASYS program consists of Options and Edit data, Model data, and Logic Flow and Operations data. Options and Edit data provide for basic program control and user edit capability. The Model data describe the problem in terms of geometry and other properties. This information includes surface geometry data, documentation data, nodal data, block coordinate system data, form factor data, and flux data. Logic Flow and Operations data house the user's driver logic, including the sequence of subroutine calls and the subroutine library. Output from TRASYS consists of two basic types of data: internode radiation interchange data, and incident and absorbed heat rate data. The flexible structure of TRASYS allows considerable freedom in the definition and choice of solution method for a thermal radiation problem. The program's flexible structure has also allowed TRASYS to retain the same basic input structure as the authors update it in order to keep up with changing requirements. Among its other important features are the following: 1) up to 3200 node problem size capability with shadowing by intervening opaque or semi-transparent surfaces; 2) choice of diffuse, specular, or diffuse/specular radiant interchange solutions; 3) a restart capability that minimizes recomputing; 4) macroinstructions that automatically provide the executive logic for orbit generation that optimizes the use of previously completed computations; 5) a time variable geometry package that provides automatic pointing of the various parts of an articulated spacecraft and an automatic look-back feature that eliminates redundant form factor calculations; 6) capability to specify submodel names to identify sets of surfaces or components as an entity; and 7) subroutines to perform functions which save and recall the internodal and/or space form factors in subsequent steps for nodes with fixed geometry during a variable geometry run. There are two machine versions of TRASYS v27: a DEC VAX version and a Cray UNICOS version. Both versions require installation of the NASADIG library (MSC-21801 for DEC VAX or COS-10049 for CRAY), which is available from COSMIC either separately or bundled with TRASYS. The NASADIG (NASA Device Independent Graphics Library) plot package provides a pictorial representation of input geometry, orbital/orientation parameters, and heating rate output as a function of time. NASADIG supports Tektronix terminals. The CRAY version of TRASYS v27 is written in FORTRAN 77 for batch or interactive execution and has been implemented on CRAY X-MP and CRAY Y-MP series computers running UNICOS. The standard distribution medium for MSC-21959 (CRAY version without NASADIG) is a 1600 BPI 9-track magnetic tape in UNIX tar format. The standard distribution medium for COS-10040 (CRAY version with NASADIG) is a set of two 6250 BPI 9-track magnetic tapes in UNIX tar format. Alternate distribution media and formats are available upon request. The DEC VAX version of TRASYS v27 is written in FORTRAN 77 for batch execution (only the plotting driver program is interactive) and has been implemented on a DEC VAX 8650 computer under VMS. Since the source codes for MSC-21030 and COS-10026 are in VAX/VMS text library files and DEC Command Language files, COSMIC will only provide these programs in the following formats: MSC-21030, TRASYS (DEC VAX version without NASADIG) is available on a 1600 BPI 9-track magnetic tape in VAX BACKUP format (standard distribution medium) or in VAX BACKUP format on a TK50 tape cartridge; COS-10026, TRASYS (DEC VAX version with NASADIG), is available in VAX BACKUP format on a set of three 6250 BPI 9-track magnetic tapes (standard distribution medium) or a set of three TK50 tape cartridges in VAX BACKUP format. TRASYS was last updated in 1993.

Description: INS3D computes steady-state solutions to the incompressible Navier-Stokes equations. The INS3D approach utilizes pseudo-compressibility combined with an approximate factorization scheme. This computational fluid dynamics (CFD) code has been verified on problems such as flow through a channel, flow over a backwardfacing step and flow over a circular cylinder. Three dimensional cases include flow over an ogive cylinder, flow through a rectangular duct, wind tunnel inlet flow, cylinder-wall juncture flow and flow through multiple posts mounted between two plates. INS3D uses a pseudo-compressibility approach in which a time derivative of pressure is added to the continuity equation, which together with the momentum equations form a set of four equations with pressure and velocity as the dependent variables. The equations' coordinates are transformed for general three dimensional applications. The equations are advanced in time by the implicit, non-iterative, approximately-factored, finite-difference scheme of Beam and Warming. The numerical stability of the scheme depends on the use of higher-order smoothing terms to damp out higher-frequency oscillations caused by second-order central differencing. The artificial compressibility introduces pressure (sound) waves of finite speed (whereas the speed of sound would be infinite in an incompressible fluid). As the solution converges, these pressure waves die out, causing the derivation of pressure with respect to time to approach zero. Thus, continuity is satisfied for the incompressible fluid in the steady state. Computational efficiency is achieved using a diagonal algorithm. A block tri-diagonal option is also available. When a steady-state solution is reached, the modified continuity equation will satisfy the divergence-free velocity field condition. INS3D is capable of handling several different types of boundaries encountered in numerical simulations, including solid-surface, inflow and outflow, and far-field boundaries. Three machine versions of INS3D are available. INS3D for the CRAY is written in CRAY FORTRAN for execution on a CRAY X-MP under COS, INS3D for the IBM is written in FORTRAN 77 for execution on an IBM 3090 under the VM or MVS operating system, and INS3D for DEC RISC-based systems is written in RISC FORTRAN for execution on a DEC workstation running RISC ULTRIX 3.1 or later. The CRAY version has a central memory requirement of 730279 words. The central memory requirement for the IBM is 150Mb. The memory requirement for the DEC RISC ULTRIX version is 3Mb of main memory. INS3D was developed in 1987. The port to the IBM was done in 1990. The port to the DECstation 3100 was done in 1991. CRAY is a registered trademark of Cray Research Inc. IBM is a registered trademark of International Business Machines. DEC, DECstation, and ULTRIX are trademarks of the Digital Equipment Corporation.

Description: PLAN2D is a FORTRAN computer program for the plastic analysis of planar rigid frame structures. Given a structure and loading pattern as input, PLAN2D calculates the ultimate load that the structure can sustain before collapse. Element moments and plastic hinge rotations are calculated for the ultimate load. The location of hinges required for a collapse mechanism to form are also determined. The program proceeds in an iterative series of linear elastic analyses. After each iteration the resulting elastic moments in each member are compared to the reserve plastic moment capacity of that member. The member or members that have moments closest to their reserve capacity will determine the minimum load factor and the site where the next hinge is to be inserted. Next, hinges are inserted and the structural stiffness matrix is reformulated. This cycle is repeated until the structure becomes unstable. At this point the ultimate collapse load is calculated by accumulating the minimum load factor from each previous iteration and multiplying them by the original input loads. PLAN2D is based on the program STAN, originally written by Dr. E.L. Wilson at U.C. Berkeley. PLAN2D has several limitations: 1) Although PLAN2D will detect unloading of hinges it does not contain the capability to remove hinges; 2) PLAN2D does not allow the user to input different positive and negative moment capacities and 3) PLAN2D does not consider the interaction between axial and plastic moment capacity. Axial yielding and buckling is ignored as is the reduction in moment capacity due to axial load. PLAN2D is written in FORTRAN and is machine independent. It has been tested on an IBM PC and a DEC MicroVAX. The program was developed in 1988.

Description: Predictions of fatigue crack growth behavior can be made with the Fatigue Crack Growth Structural Analysis (FASTRAN II) computer program. As cyclic loads are applied to a selected crack configuration with an initial crack size, FASTRAN II predicts crack growth as a function of cyclic load history until either a desired crack size is reached or failure occurs. FASTRAN II is based on plasticity-induced crack-closure behavior of cracks in metallic materials and accounts for load-interaction effects, such as retardation and acceleration, under variable-amplitude loading. The closure model is based on the Dugdale model with modifications to allow plastically deformed material to be left along the crack surfaces as the crack grows. Plane stress and plane strain conditions, as well as conditions between these two, can be simulated in FASTRAN II by using a constraint factor on tensile yielding at the crack front to approximately account for three-dimensional stress states. FASTRAN II contains seventeen predefined crack configurations (standard laboratory fatigue crack growth rate specimens and many common crack configurations found in structures); and the user can define one additional crack configuration. The baseline crack growth rate properties (effective stress-intensity factor against crack growth rate) may be given in either equation or tabular form. For three-dimensional crack configurations, such as surface cracks or corner cracks at holes or notches, the fatigue crack growth rate properties may be different in the crack depth and crack length directions. Final failure of the cracked structure can be modelled with fracture toughness properties using either linear-elastic fracture mechanics (brittle materials), a two-parameter fracture criterion (brittle to ductile materials), or plastic collapse (extremely ductile materials). The crack configurations in FASTRAN II can be subjected to either constant-amplitude, variable-amplitude or spectrum loading. The applied loads may be either tensile or compressive. Several standardized aircraft flight-load histories, such as TWIST, Mini-TWIST, FALSTAFF, Inverted FALSTAFF, Felix and Gaussian, are included as options. FASTRAN II also includes two other methods that will help the user input spectrum load histories. The two methods are: (1) a list of stress points, and (2) a flight-by-flight history of stress points. Examples are provided in the user manual. Developed as a research program, FASTRAN II has successfully predicted crack growth in many metallic materials under various aircraft spectrum loading. A computer program DKEFF which is a part of the FASTRAN II package was also developed to analyze crack growth rate data from laboratory specimens to obtain the effective stress-intensity factor against crack growth rate relations used in FASTRAN II. FASTRAN II is written in standard FORTRAN 77. It has been successfully compiled and implemented on Sun4 series computers running SunOS and on IBM PC compatibles running MS-DOS using the Lahey F77L FORTRAN compiler. Sample input and output data are included with the FASTRAN II package. The UNIX version requires 660K of RAM for execution. The standard distribution medium for the UNIX version (LAR-14865) is a .25 inch streaming magnetic tape cartridge in UNIX tar format. It is also available on a 3.5 inch diskette in UNIX tar format. The standard distribution medium for the MS-DOS version (LAR-14944) is a 5.25 inch 360K MS-DOS format diskette. The contents of the diskette are compressed using the PKWARE archiving tools. The utility to unarchive the files, PKUNZIP.EXE, is included. The program was developed in 1984 and revised in 1992. Sun4 and SunOS are trademarks of Sun Microsystems, Inc. IBM PC is a trademark of International Business Machines Corp. MS-DOS is a trademark of Microsoft, Inc. F77L is a trademark of the Lahey Computer Systems, Inc. UNIX is a registered trademark of AT&T Bell Laboratories. PKWARE and PKUNZIP are trademarks of PKWare, Inc.

Description: The ability to treat arbitrary boundary shapes is one of the most desirable characteristics of a method for generating grids. 3DGRAPE is designed to make computational grids in or about almost any shape. These grids are generated by the solution of Poisson's differential equations in three dimensions. The program automatically finds its own values for inhomogeneous terms which give near-orthogonality and controlled grid cell height at boundaries. Grids generated by 3DGRAPE have been applied to both viscous and inviscid aerodynamic problems, and to problems in other fluid-dynamic areas. 3DGRAPE uses zones to solve the problem of warping one cube into the physical domain in real-world computational fluid dynamics problems. In a zonal approach, a physical domain is divided into regions, each of which maps into its own computational cube. It is believed that even the most complicated physical region can be divided into zones, and since it is possible to warp a cube into each zone, a grid generator which is oriented to zones and allows communication across zonal boundaries (where appropriate) solves the problem of topological complexity. 3DGRAPE expects to read in already-distributed x,y,z coordinates on the bodies of interest, coordinates which will remain fixed during the entire grid-generation process. The 3DGRAPE code makes no attempt to fit given body shapes and redistribute points thereon. Body-fitting is a formidable problem in itself. The user must either be working with some simple analytical body shape, upon which a simple analytical distribution can be easily effected, or must have available some sophisticated stand-alone body-fitting software. 3DGRAPE does not require the user to supply the block-to-block boundaries nor the shapes of the distribution of points. 3DGRAPE will typically supply those block-to-block boundaries simply as surfaces in the elliptic grid. Thus at block-to-block boundaries the following conditions are obtained: (1) grids lines will match up as they approach the block-to-block boundary from either side, (2) grid lines will cross the boundary with no slope discontinuity, (3) the spacing of points along the line piercing the boundary will be continuous, (4) the shape of the boundary will be consistent with the surrounding grid, and (5) the distribution of points on the boundary will be reasonable in view of the surrounding grid. 3DGRAPE offers a powerful building-block approach to complex 3-D grid generation, but is a low-level tool. Users may build each face of each block as they wish, from a wide variety of resources. 3DGRAPE uses point-successive-over-relaxation (point-SOR) to solve the Poisson equations. This method is slow, although it does vectorize nicely. Any number of sophisticated graphics programs may be used on the stored output file of 3DGRAPE though it lacks interactive graphics. Versatility was a prominent consideration in developing the code. The block structure allows a great latitude in the problems it can treat. As the acronym implies, this program should be able to handle just about any physical region into which a computational cube or cubes can be warped. 3DGRAPE was written in FORTRAN 77 and should be machine independent. It was originally developed on a Cray under COS and tested on a MicroVAX 3200 under VMS 5.1.

Description: The SNIP program is a FORTRAN computer code that generates NASTRAN structural model thermal loads when given SINDA (or similar thermal model) temperature results. SNIP correlates thermal nodes to structural elements to interface SINDA finite difference thermal models with NASTRAN finite element structural models. Node-to-element correlation includes determining which SINDA nodes should be related to each NASTRAN element and calculating a weighing factor for temperatures associated with each element-related thermal node. SNIP provides structural model thermal loads that accurately reflect thermal model results while reducing the time required to interface thermal and structural models as compared to other methods. SNIP uses thermal model geometry to search the three-dimensional space around each structural element for the nearest thermal nodes. Thermal model geometry is the combination of standard thermal model temperature results from SINDA and structural model geometry from NASTRAN. Thermal and structural models must both be defined in the same, single Cartesian coordinate system. The thermal nodes located nearest each element are used to determine element temperature for thermal distortion and stress analysis. The program shapes the three-dimensional search region while the user controls the size. With these region specifications, the numerical coding of thermal nodes, and the structural element numbers; the code can provide for the separation of substructures during correlation. The input to SNIP contains a file of thermal model temperature results and a physical location of each thermal node in three-dimensional space, combined in a SNIP-unique format. The input also contains a standard NASTRAN input deck for a model made up of plate, shell, beam, and bar elements. SNIP supports the CTRIA, CQUAD, CBAR, and CBEAM elements of NASTRAN. The user adjusts the input parameters in the source code which control the node-to-element correlation. The program outputs NASTRAN element temperature load cards for each element and NASTRAN case control cards for each temperature load set. SNIP also outputs a list of elements that contains the numbers of the SINDA nodes related to each NASTRAN element and the weight that is given to each node in temperature calculations. SNIP is written in ANSI standard FORTRAN 77. The PC version requires a PC FORTRAN compiler and has compiled successfully using Lahey FORTRAN v. 3.0. A core memory of 300k is recommended. The program was developed in 1987.

Description: This algorithm has been developed for calculating both the quantity of compressor bleed flow required to cool a turbine and the resulting decrease in efficiency due to cooling air injected into the gas stream. Because of the trend toward higher turbine inlet temperatures, it is important to accurately predict the required cooling flow. This program is intended for use with axial flow, air-breathing jet propulsion engines with a variety of airfoil cooling configurations. The algorithm results have compared extremely well with figures given by major engine manufacturers for given bulk metal temperatures and cooling configurations. The program calculates the required cooling flow and corresponding decrease in stage efficiency for each row of airfoils throughout the turbine. These values are combined with the thermodynamic efficiency of the uncooled turbine to predict the total bleed airflow required and the altered turbine efficiency. There are ten airfoil cooling configurations and the algorithm allows a different option for each row of cooled airfoils. Materials technology is incorporated and requires the date of the first year of service for the turbine stator vane and rotor blade. The user must specify pressure, temperatures, and gas flows into the turbine. This program is written in FORTRAN IV for batch execution and has been implemented on an IBM 3080 series computer with a central memory requirement of approximately 61K of 8 bit bytes. This program was developed in 1980.

Description: A fast algorithm has been developed for accurately generating boundary-conforming, three-dimensional consecutively refined computational grids applicable to arbitrary wing-body and axial turbomachinery geometries. This algorithm has been incorporated into the GRID3O computer program. The method employed in GRID3O is based on using an analytic function to generate two-dimensional grids on a number of coaxial axisymmetric surfaces positioned between the centerbody and the outer radial boundary. These grids are of the O-type and are characterized by quasi-orthogonality, geometric periodicity, and an adequate resolution throughout the flow field. Because the built-in nonorthogonal coordinate stretching and shearing cause the grid lines leaving the blade or wing trailing-edge to end at downstream infinity, use of the generated grid simplifies the numerical treatment of three-dimensional trailing vortex sheets. The GRID3O program is written in FORTRAN IV for batch execution and has been implemented on an IBM 370 series computer with a central memory requirement of approximately 450K of 8 bit bytes. The GRID3O program was developed in 1981.

Description: As turbine-engine core operating conditions become more severe, designers must develop more effective means of cooling blades and vanes. In order to design reliable, cooled turbine blades, advanced transient thermal calculation techniques are required. The TACT1 computer program was developed to perform transient and steady-state heat-transfer and coolant-flow analyses for cooled blades, given the outside hot-gas boundary condition, the coolant inlet conditions, the geometry of the blade shell, and the cooling configuration. TACT1 can analyze turbine blades, or vanes, equipped with a central coolant-plenum insert from which coolant-air impinges on the inner surface of the blade shell. Coolant-side heat-transfer coefficients are calculated with the heat transfer mode at each station being user specified as either impingement with crossflow, forced convection channel flow, or forced convection over pin fins. A limited capability to handle film cooling is also available in the program. The TACT1 program solves for the blade temperature distribution using a transient energy equation for each node. The nodal energy balances are linearized, one-dimensional, heat-conduction equations which are applied at the wall-outer-surface node, at the junction of the cladding and the metal node, and at the wall-inner-surface node. At the mid-metal node a linear, three-dimensional, heat-conduction equation is used. Similarly, the coolant pressure distribution is determined by solving the set of transfer momentum equations for the one-dimensional flow between adjacent fluid nodes. In the coolant channel, energy and momentum equations for one-dimensional compressible flow, including friction and heat transfer, are used for the elemental channel length between two coolant nodes. The TACT1 program first obtains a steady-state solution using iterative calculations to obtain convergence of stable temperatures, pressures, coolant-flow split, and overall coolant mass balance. Transient calculations are based on the steady-state solutions obtained. Input to the TACT1 program includes a geometrical description of the blade and insert, the nodal spacing to be used, and the boundary conditions describing the outside hot-gas and the coolant-inlet conditions. The program output includes the value of nodal temperatures and pressures at each iteration. The final solution output includes the temperature at each coolant node, and the coolant flow rates and Reynolds numbers. This program is written in FORTRAN IV for batch execution and has been implemented on an IBM 360 computer with a central memory requirement of approximately 480K of 8 bit bytes. The TACT1 program was developed in 1978.

Description: This is a finite-difference program for calculating the viscous compressible boundary layer flow over either planar or axisymmetric surfaces. The flow may be initially laminar and progress through a transitional zone to a fully turbulent flow, or it may remain laminar, depending on the imposed boundary conditions, laws of viscosity, and numerical solution of the momentum and energy equations. The flow may also be forced into a turbulent flow at a chosen spot by the data input. The input may contain factors of arbitrary Reynolds number, free-stream Mach number, free stream turbulence, wall heating or cooling, longitudinal wall curvature, wall suction or blowing, and wall roughness. The solution may start from an initial Falkner-Skan similarity profile, an approximate equilibrium turbulent profile, or an initial arbitrary input profile. This program has been implemented on the IBM 7094/7044 Direct Couple System. This program is written in FORTRAN IV and was developed in 1974.

Description: A computer program has been developed for the design of sharp-edged throat supersonic nozzles where losses are accounted for by correcting the ideal nozzle geometry for boundary layer displacement thickness. The ideal nozzle is designed by the method of characteristics to produce uniform parallel flow at the nozzle exit in the smallest possible distance. Boundary-layer parameters (displacement and momentum thicknesses) are calculated for the ideal nozzle, and the final nozzle geometry is obtained by adding the displacement thickness to the ideal nozzle coordinates. The boundary layer parameters are also used to calculate the aftermixing conditions downstream of the nozzle assuming the flow mixes to a uniform state. The computer program input consists essentially of the nozzle-exit Mach number, specific-heat ratio, nozzle angle, throat half-height, nozzle subsonic section coordinates and corresponding pressure ratios, total temperature and pressure, gas constant, and initial momentum or displacement thickness. The program gas properties are set up for air; for other gases, changes are required to the program. The computer program output consists of the corrected nozzle coordinates, the principal boundary-layer parameters, and the aftermixing conditions. This program has been implemented on the IBM 7094/7044 Direct Couple System.

Description: A computer program has been developed which analyzes by means of mathematical models the temperature profiles in the contents of a filled propellant tank. In designing space vehicles using cryogenic liquid propellants, it is necessary to know how heat transferred from the tank walls and heat absorbed internally affect the temperature distribution with the tank contents. The mathematical flow model is based on results from small-scale experiments. The results showed that when a subcooled fluid is subject to both nonuniform internal heating and wall heating, two distinct temperature regions are developed. In the lower region, the fluid is thoroughly mixed and maintains a uniform temperature profile. In the upper region, a stratified layer develops, and a temperature gradient is formed from the accumulation of warm fluid from the boundary layer along the tank walls; it also indicated that the temperature profiles in the stratified layer exhibited similarity. This concept was developed primarily for internal heating caused by nuclear radiation. However, the theory and computer program are applicable for any form of internal or bulk heating. This program is written in FORTRAN IV for batch execution and has been implemented on the IBM 7094. This program was developed in 1970.

Description: This program solves the two-dimensional, compressible laminar or turbulent boundary-layer equations in an arbitrary pressure gradient. Cohen and Reshotko's method is used for the laminar boundary layer, Sasman and Cresci's method for the turbulent boundary layer, and the Schlichting-Ulrich-Granville method to predict transition. Transition may also be forced at any point by the user. Separation, if it occurs, is predicted for both laminar and turbulent flow. The user may begin values for displacement thickness and momentum thickness in either laminar or turbulent flow. This program was implemented on the IBM 7094.

Description: NPLOT is an interactive computer graphics program for plotting undeformed and deformed NASTRAN finite element models (FEMs). Although there are many commercial codes already available for plotting FEMs, these have limited use due to their cost, speed, and lack of features to view BAR elements. NPLOT was specifically developed to overcome these limitations. On a vector type graphics device the two best ways to show depth are by hidden line plotting or haloed line plotting. A hidden line algorithm generates views of models with all hidden lines removed, and a haloed line algorithm displays views with aft lines broken in order to show depth while keeping the entire model visible. A haloed line algorithm is especially useful for plotting models composed of many line elements and few surface elements. The most important feature of NPLOT is its ability to create both hidden line and haloed line views accurately and much more quickly than with any other existing hidden or haloed line algorithms. NPLOT is also capable of plotting a normal wire frame view to display all lines of a model. NPLOT is able to aid in viewing all elements, but it has special features not generally available for plotting BAR elements. These features include plotting of TRUE LENGTH and NORMALIZED offset vectors and orientation vectors. Standard display operations such as rotation and perspective are possible, but different view planes such as X-Y, Y-Z, and X-Z may also be selected. Another display option is the Z-axis cut which allows a portion of the fore part of the model to be cut away to reveal details of the inside of the model. A zoom function is available to terminals with a locator (graphics cursor, joystick, etc.). The user interface of NPLOT is designed to make the program quick and easy to use. A combination of menus and commands with help menus for detailed information about each command allows experienced users greater speed and efficiency. Once a plot is on the screen the interface becomes command driven, enabling the user to manipulate the display or execute a command without having to return to the menu. NPLOT is also able to plot deformed shapes allowing it to perform post-processing. The program can read displacements, either static displacements or eigenvectors, from a MSC/NASTRAN F06 file or a UAI/NASTRAN PRT file. The displacements are written into a unformatted scratch file where they are available for rapid access when the user wishes to display a deformed shape. All subcases or mode shapes can be read in at once. Then it is easy to enable the deformed shape, to change subcases or mode shapes and to change the scale factor for subsequent plots. NPLOT is written in VAX FORTRAN for DEC VAX series computers running VMS. As distributed, the NPLOT source code makes calls to the DI3000 graphics package from Precision Visuals; however, a set of interface routines is provided to translate the DI3000 calls into Tektronix PLOT10/TCS graphics library calls so that NPLOT can use the standard Tektronix 4010 which many PC terminal emulation software programs support. NPLOT is available in VAX BACKUP format on a 9-track 1600 BPI DEC VAX BACKUP format magnetic tape (standard media) or a TK50 tape cartridge. This program was developed in 1991. DEC, VAX, VMS, and TK50 are trademarks of Digital Equipment Corporation. Tektronix, PLOT10, and TCS are trademarks of Tektronix, Inc. DI3000 is a registered trademark of Precision Visuals, Inc. NASTRAN is a registered trademark of the National Aeronautics and Space Administration. MSC/ is a trademark of MacNeal-Schwendler Corporation. UAI is a trademark of Universal Analytics, Inc.

Description: New research on hypersonic vehicles, such as the National Aero-Space Plane (NASP), has raised concerns about the effects of shock-wave interference on various structural components of the craft. State-of-the-art aerothermal analysis software is inadequate to predict local flow and heat flux in areas of extremely high heat transfer, such as the surface impingement of an Edney-type supersonic jet. EASI revives and updates older computational methods for calculating inviscid flow field and maximum heating from shock wave interference. The program expands these methods to solve problems involving the six shock-wave interference patterns on a two-dimensional cylindrical leading edge with an equilibrium chemically reacting gas mixture (representing, for example, the scramjet cowl of the NASP). The inclusion of gas chemistry allows for a more accurate prediction of the maximum pressure and heating loads by accounting for the effects of high temperature on the air mixture. Caloric imperfections and specie dissociation of high-temperature air cause shock-wave angles, flow deflection angles, and thermodynamic properties to differ from those calculated by a calorically perfect gas model. EASI contains pressure- and temperature-dependent thermodynamic and transport properties to determine heating rates, and uses either a calorically perfect air model or an 11-specie, 7-reaction reacting air model at equilibrium with temperatures up to 15,000 K for the inviscid flowfield calculations. EASI solves the flow field and the associated maximum surface pressure and heat flux for the six common types of shock wave interference. Depending on the type of interference, the program solves for shock-wave/boundary-layer interaction, expansion-fan/boundary-layer interaction, attaching shear layer or supersonic jet impingement. Heat flux predictions require a knowledge (from experimental data or relevant calculations) of a pertinent length scale of the interaction. Output files contain flow-field information for the various shock-wave interference patterns and their associated maximum surface pressure and heat flux predictions. EASI is written in FORTRAN 77 for a DEC VAX 8500 series computer using the VAX/VMS operating system, and requires 75K of memory. The program is available on a 9-track 1600 BPI magnetic tape in DEC VAX BACKUP format. EASI was developed in 1989. DEC, VAX, and VMS are registered trademarks of the Digital Equipment Corporation.

Description: LATDYN is a computer code for modeling the Large Angle Transient DYNamics of flexible articulating structures and mechanisms involving joints about which members rotate through large angles. LATDYN extends and brings together some of the aspects of Finite Element Structural Analysis, Multi-Body Dynamics, and Control System Analysis; three disciplines that have been historically separate. It combines significant portions of their distinct capabilities into one single analysis tool. The finite element formulation for flexible bodies in LATDYN extends the conventional finite element formulation by using a convected coordinate system for constructing the equation of motion. LATDYN's formulation allows for large displacements and rotations of finite elements subject to the restriction that deformations within each are small. Also, the finite element approach implemented in LATDYN provides a convergent path for checking solutions simply by increasing mesh density. For rigid bodies and joints LATDYN borrows extensively from methodology used in multi-body dynamics where rigid bodies may be defined and connected together through joints (hinges, ball, universal, sliders, etc.). Joints may be modeled either by constraints or by adding joint degrees of freedom. To eliminate error brought about by the separation of structural analysis and control analysis, LATDYN provides symbolic capabilities for modeling control systems which are integrated with the structural dynamic analysis itself. Its command language contains syntactical structures which perform symbolic operations which are also interfaced directly with the finite element structural model, bypassing the modal approximation. Thus, when the dynamic equations representing the structural model are integrated, the equations representing the control system are integrated along with them as a coupled system. This procedure also has the side benefit of enabling a dramatic simplification of the user interface for modeling control systems. Three FORTRAN computer programs, the LATDYN Program, the Preprocessor, and the Postprocessor, make up the collective LATDYN System. The Preprocessor translates user commands into a form which can be used while the LATDYN program provides the computational core. The Postprocessor allows the user to interactively plot and manage a database of LATDYN transient analysis results. It also includes special facilities for modeling control systems and for programming changes to the model which take place during analysis sequence. The documentation includes a Demonstration Problem Manual for the evaluation and verification of results and a Postprocessor guide. Because the program should be viewed as a byproduct of research on technology development, LATDYN's scope is limited. It does not have a wide library of finite elements, and 3-D Graphics are not available. Nevertheless, it does have a measure of "user friendliness". The LATDYN program was developed over a period of several years and was implemented on a CDC NOS/VE & Convex Unix computer. It is written in FORTRAN 77 and has a virtual memory requirement of 1.46 MB. The program was validated on a DEC MICROVAX operating under VMS 5.2.

Description: The increasing number of applications of fiber-reinforced composites in industry demands a detailed understanding of their material properties and behavior. A three-dimensional finite-element computer program called PAFAC (Plastic and Failure Analysis of Composites) has been developed for the elastic-plastic analysis of fiber-reinforced composite materials and structures. The evaluation of stresses and deformations at edges, cut-outs, and joints is essential in understanding the strength and failure for metal-matrix composites since the onset of plastic yielding starts very early in the loading process as compared to the composite's ultimate strength. Such comprehensive analysis can only be achieved by a finite-element program like PAFAC. PAFAC is particularly suited for the analysis of laminated metal-matrix composites. It can model the elastic-plastic behavior of the matrix phase while the fibers remain elastic. Since the PAFAC program uses a three-dimensional element, the program can also model the individual layers of the laminate to account for thickness effects. In PAFAC, the composite is modeled as a continuum reinforced by cylindrical fibers of vanishingly small diameter which occupy a finite volume fraction of the composite. In this way, the essential axial constraint of the phases is retained. Furthermore, the local stress and strain fields are uniform. The PAFAC finite-element solution is obtained using the displacement method. Solution of the nonlinear equilibrium equations is obtained with a Newton-Raphson iteration technique. The elastic-plastic behavior of composites consisting of aligned, continuous elastic filaments and an elastic-plastic matrix is described in terms of the constituent properties, their volume fractions, and mutual constraints between phases indicated by the geometry of the microstructure. The program uses an iterative procedure to determine the overall response of the laminate, then from the overall response determines the stress state in each phase of the composite material. Failure of the fibers or matrix within an element can also be modeled by PAFAC. PAFAC is written in FORTRAN IV for batch execution and has been implemented on a CDC CYBER 170 series computer with a segmented memory requirement of approximately 66K (octal) of 60 bit words. PAFAC was developed in 1982.

Description: The Panel Analysis and Sizing Code (PASCO) was developed for the buckling and vibration analysis and sizing of prismatic structures having an arbitrary cross section. PASCO is primarily intended for analyzing and sizing stiffened panels made of laminated orthotropic materials and is of particular value in analyzing and sizing filamentary composite structures. When used in the analysis mode, PASCO calculates laminate stiffnesses, lamina stress and strains (including the effects of temperature and panel bending), buckling loads, vibration frequencies, and overall panel stiffness. When used in the sizing mode, PASCO adjusts sizing variables to provide a low-mass panel design that carries a set of specified loadings without exceeding buckling or material strength allowables and that meets other design requirements such as upper and lower bounds on sizing variables, upper and lower bounds on overall bending, extensional and shear stiffnesses, and lower bounds on vibration frequencies. Although emphasis in PASCO is placed on flat panels having several identical bays, the only restriction on configuration modeling is that the structure is assumed to be prismatic. In addition, it is assumed that loads and temperatures do not vary along the length of a panel. Because of their wide application in aerospace structures, stiffened panels are readily handled by PASCO. The panel cross section may be composed of an arbitrary assemblage of thin, flat, rectangular plate elements that are connected together along their longitudinal edges. Each plate element consists of a balanced symmetric laminate of any number of layers of orthotropic material. Any group of element widths, layer thicknesses, and layer orientation angles may be selected as sizing variables. Substructuring is available to increase the efficiency of the analysis and to simplify the modeling of complicated structures. The Macintosh version of PASCO includes an interactive, graphic preprocessor called MacPASCO. The main objective of MacPASCO is to make the use of PASCO faster, simpler, and less error-prone. By using a graphical user interface (GUI), MacPASCO simplifies the specification of panel geometry and reduces user input errors, thus making the modeling and analysis of panel designs more efficient. The user draws the initial structural geometry on the computer screen, then uses a combination of graphic and text inputs to: refine the structural geometry, specify information required for analysis such as panel load conditions, and define design variables and constraints for minimum-mass optimization. Composite panel design is an ideal application because the graphical user interface can: serve as a visual aid, eliminate the tedious aspects of text-based input, and eliminate many sources of input errors. The current version of MacPASCO does not implement all the modeling features of PASCO, but has been found to be sufficient for many users. Many difficulties common to text-based inputs are avoided because MacPASCO uses a GUI. First, the graphic displays eliminate syntax errors, like misplaced commas and incorrect command names, because there is no text-based syntax. Second, graphic displays allow the user to see the geometry as it is created and immediately detect and correct any errors. Third, MacPASCO's drawing tools have been designed to avoid modeling errors. Fourth, the graphic displays make revisions to existing structural designs much easier and less error-prone by eliminating the need for the user to conceptualize the text input as geometry. The user can work directly with the geometry displayed on the screen. Finally, MacPASCO automatically generates the correct PASCO input from the geometry displayed on the screen. This input file can be used with any machine version of PASCO to actually perform the analysis and sizing and to output results. The DEC VAX version of PASCO is written in FORTRAN IV for batch execution and has been implemented on a DEC VAX series computer. The Macintosh version of PASCO was developed for Macintosh II series computers with at least 2Mb of RAM running MPW Pascal 3.0 and Language Systems FORTRAN 2.0 under the MPW programming environment. It includes MPW compatible makefiles for compiling the source code. The Macintosh version uses input files compatible with versions of PASCO running on different platforms. MacPASCO is written in Macintosh Programmers Workbench 3.0, MPW Pascal 3.0, and MacAPP 2.0. The Pascal source code is included on the distribution diskette. MacAPP is a development library which is not included. MacPASCO requires a Mac Plus, SE/30, or MacII, IIx, IIcx, IIci, or IIfx running System 6.0 or greater. MacPASCO is System 7.0 compatible. A minimum of 2Mb of RAM is required for execution. The Macintosh version of PASCO is distributed on four 3.5 inch 800K Macintosh format diskettes. The DEC VAX version is distributed on a 9-track 1600 BPI magnetic tape. The PASCO program was developed in 1981, adapted to the DEC VAX in 1983 and to the Macintosh in 1991. MacPASCO was released in 1992.

Description: Current investigations of the hydrogen-fueled supersonic combustion ramjet engine have delineated several technological problem areas. One area, the analysis of the injection, turbulent mixing, and combusiton of hydrogen, requires the accurate calculation of the supersonic combustion flow fields. This calculation has proven difficult because of an interesting phenomena which makes possible the transition from supersonic to subsonic flow in the combustion field, due to the temperature transitions which occur in the flow field. This computer program was developed to use viscous characteristics theory to analyze supersonic combustion flow fields with imbedded subsonic regions. Intended to be used as a practical design tool for two-dimensional and axisymmetric supersonic combustor development, this program has proven useful in the analysis of such problems as determining the flow field of a single underexpanded hydrogen jet, the internal flow of a gas sampling probe, the effects of fuel-injector strut shape, and the effects of changes in combustor configuration. Both combustion and diffusive effects can significantly alter the wave pattern in a supersonic field and generate significant pressure gradients in both the axial and radial directions. The induced pressure, in turn, substantially influences the ignition delay and reaction times as well as the velocity distribution. To accurately analyze the flow fields, the effects of finite rate chemistry, mixing, and wave propagation must be properly linked to one another. The viscous characteristics theory has been used in the past to describe flows that are purely supersonic; however, the interacting pressure effects in the combustor often allow for the development of shock waves and imbedded subsonic regions. Numerical investigation of these transonic situations has required the development of a new viscous characteristics procedure which is valid within the subsonic region and can be coupled with the standard viscous characteristics procedure in the supersonic region. The basic governing equations used are the 'viscous-inviscid' equations, similar to those employed in higher-order boundary layer analyses, with finite rate chemistry terms included. In addition, the Rankine-Hugoniot and Prandtl-Meyer relations are used to compute shock and expansion conditions. The program can handle up to 20 simultaneous shock waves. Chemistry terms are computed for a 7-species 8-mechanism hydrogen-air reaction scheme. The user input consists of a physical description of the combustor and flow determination parameters. Output includes detail flow parameter values at selected points within the flow field. This computer program is written in FORTRAN IV for batch execution and has been implemented on a CDC CYBER 175 with a central memory requirement of approximately 114K (octal) of 60 bit words. The program was developed in 1978.

Description: SINDA, the Systems Improved Numerical Differencing Analyzer, is a software system for solving lumped parameter representations of physical problems governed by diffusion-type equations. SINDA was originally designed for analyzing thermal systems represented in electrical analog, lumped parameter form, although its use may be extended to include other classes of physical systems which can be modeled in this form. As a thermal analyzer, SINDA can handle such interrelated phenomena as sublimation, diffuse radiation within enclosures, transport delay effects, and sensitivity analysis. FLUINT, the FLUid INTegrator, is an advanced one-dimensional fluid analysis program that solves arbitrary fluid flow networks. The working fluids can be single phase vapor, single phase liquid, or two phase. The SINDA'85/FLUINT system permits the mutual influences of thermal and fluid problems to be analyzed. The SINDA system consists of a programming language, a preprocessor, and a subroutine library. The SINDA language is designed for working with lumped parameter representations and finite difference solution techniques. The preprocessor accepts programs written in the SINDA language and converts them into standard FORTRAN. The SINDA library consists of a large number of FORTRAN subroutines that perform a variety of commonly needed actions. The use of these subroutines can greatly reduce the programming effort required to solve many problems. A complete run of a SINDA'85/FLUINT model is a four step process. First, the user's desired model is run through the preprocessor which writes out data files for the processor to read and translates the user's program code. Second, the translated code is compiled. The third step requires linking the user's code with the processor library. Finally, the processor is executed. SINDA'85/FLUINT program features include 20,000 nodes, 100,000 conductors, 100 thermal submodels, and 10 fluid submodels. SINDA'85/FLUINT can also model two phase flow, capillary devices, user defined fluids, gravity and acceleration body forces on a fluid, and variable volumes. SINDA'85/FLUINT offers the following numerical solution techniques. The Finite difference formulation of the explicit method is the Forward-difference explicit approximation. The formulation of the implicit method is the Crank-Nicolson approximation. The program allows simulation of non-uniform heating and facilitates modeling thin-walled heat exchangers. The ability to model non-equilibrium behavior within two-phase volumes is included. Recent improvements to the program were made in modeling real evaporator-pumps and other capillary-assist evaporators. SINDA'85/FLUINT is available by license for a period of ten (10) years to approved licensees. The licensed program product includes the source code and one copy of the supporting documentation. Additional copies of the documentation may be purchased separately at any time. SINDA'85/FLUINT is written in FORTRAN 77. Version 2.3 has been implemented on Cray series computers running UNICOS, CONVEX computers running CONVEX OS, and DEC RISC computers running ULTRIX. Binaries are included with the Cray version only. The Cray version of SINDA'85/FLUINT also contains SINGE, an additional graphics program developed at Johnson Space Flight Center. Both source and executable code are provided for SINGE. Users wishing to create their own SINGE executable will also need the NASA Device Independent Graphics Library (NASADIG, previously known as SMDDIG; UNIX version, MSC-22001). The Cray and CONVEX versions of SINDA'85/FLUINT are available on 9-track 1600 BPI UNIX tar format magnetic tapes. The CONVEX version is also available on a .25 inch streaming magnetic tape cartridge in UNIX tar format. The DEC RISC ULTRIX version is available on a TK50 magnetic tape cartridge in UNIX tar format. SINDA was developed in 1971, and first had fluid capability added in 1975. SINDA'85/FLUINT version 2.3 was released in 1990.

Description: The transient response of a thermal protection material to heat applied to the surface can be calculated using the CHAP III computer program. CHAP III can be used to analyze pyrolysis gas chemical kinetics in detail and examine pyrolysis reactions-indepth. The analysis includes the deposition of solid products produced by chemical reactions in the gas phase. CHAP III uses a modelling technique which can approximate a wide range of ablation problems. The energy equation used in CHAP III incorporates pyrolysis (both solid and gas reactions), convection, conduction, storage, work, kinetic energy, and viscous dissipation. The chemically reacting components of the solid are allowed to vary as a function of position and time. CHAP III employs a finite difference method to approximate the energy equations. Input values include specific heat, thermal conductivity, thermocouple locations, enthalpy, heating rates, and a description of the chemical reactions expected. The output tabulates the temperature at locations throughout the ablator, gas flow within the solid, density of the solid, weight of pyrolysis gases, and rate of carbon deposition. A sample case is included, which analyzes an ablator material containing several pyrolysis reactions subjected to an environment typical of entry at lunar return velocity. CHAP III is written in FORTRAN IV for batch execution and has been implemented on a CDC CYBER 170 series computer operating under NOS with a central memory requirement of approximately 102K (octal) of 60 bit words. This program was developed in 1985.

Description: BUCKO is a computer program developed to predict the buckling load of a rectangular compression-loaded orthotropic plate with a centrally located cutout. The plate is assumed to be a balanced, symmetric laminate of uniform thickness. The cutout shape can be elliptical, circular, rectangular, or square. The BUCKO package includes sample data that demonstrates the essence of the program and its ease of usage. BUCKO uses an approximate one-dimensional formulation of the classical two-dimensional buckling problem following the Kantorovich method. The boundary conditions are considered to be simply supported unloaded edges and either clamped or simply supported loaded edges. The plate is loaded in uniaxial compression by either uniformly displacing or uniformly stressing two opposite edges of the plate. The BUCKO analysis consists of two parts: calculation of the inplane stress distribution prior to buckling, and calculation of the plate axial load and displacement at buckling. User input includes plate planform and cutout geometry, plate membrane and bending stiffnesses, finite difference parameters, boundary condition data, and loading data. Results generated by BUCKO are the prebuckling strain energy, inplane stress resultants, buckling mode shape, critical end shortening, and average axial and transverse strains at buckling. BUCKO is written in FORTRAN V for batch execution and has been implemented on a CDC CYBER 170 series computer operating under NOS with a central memory requirement of approximately 343K of 60 bit words. This program was developed in 1984 and was last updated in 1990.

Description: This program performs a one-dimensional numerical analysis of the transient thermal response of multi-layer insulative systems. The analysis can determine the temperature distribution through a system consisting of from one to four layers, one of which can be an air gap. Concentrated heat sinks at any interface can be included. The computer program based on the analysis will determine the thickness of a specified layer that will satisfy a temperature limit criterion at any point in the insulative system. The program will also automatically calculate the thickness at several points on a system and determine the total system mass. This program was developed as a tool for designing thermal protection systems for high-speed aerospace vehicles but could be adapted to many areas of industry involved in thermal insulation systems. In this package, the equations describing the transient thermal response of a system are developed. The governing differential equation for each layer and boundary condition are put in finite-difference form using a Taylor's series expansion. These equations yield an essentially tridiagonal matrix of unknown temperatures. A procedure based on Gauss' elimination method is used to solve the matrix. This program is written in FORTRAN IV for the CDC RUN compiler and has been implemented on a CDC 6000 series machine operating under SCOPE 3.0. This program requires a minimum of 44K (octal) of 60 bit words of memory.

Description: A computer program has been developed to analyze the transient response of an ablating axisymmetric body, including the effect of shape change. The governing differential equation, the boundary conditions for the analysis on which the computer program is based, and the method of solution of the resulting finite-difference equations are discussed in the documentation. Some of the features of the analysis and the associated program are (1) the ablation material is considered to be orthotropic with temperature-dependent thermal properties; (2) the thermal response of the entire body is considered simultaneously; (3) the heat transfer and pressure distribution over the body are adjusted to the new geometry as ablation occurs; (4) the governing equations and several boundary-condition options are formulated in terms of generalized orthogonal coordinates for fixed points in a moving coordinate system; (5) the finite-difference equations are solved implicitly; and (6) other instantaneous body shapes can be displayed with a user-supplied plotting routine. The physical problem to be modeled with the analysis is described by FORTRAN input variables. For example, the external body geometry is described in the W, Z coordinates; material density is given; and the stagnation cold-wall heating rate is given in a time-dependent array. Other input variables are required which control the solution, specify boundary conditions, and determine output from the program. The equations have been programmed so that either the International System of Units or the U. S. Customary Units may be used. This program is written in FORTRAN IV for batch execution and has been implemented on a CDC 6000 Series computer. This program was developed in 1972.

Description: The GTRAN program was developed to solve transient, as well as steady state, problems for gas piping systems. GTRAN capabilities allow for the analysis of a variety of system configurations and components. These include: multiple pipe junctions; valves that change position with time; fixed restrictions (orifices, manual valves, filters, etc.); relief valves; constant pressure sources; and heat transfer for insulated piping and piping subjected to free or forced convection. In addition, boundary conditions can be incorporated to simulate specific components. The governing equations of GTRAN are the one-dimensional transient gas dynamic equations. The three equations for pressure, velocity, and density are reduced to numerical equations using an implicit Crank-Nicholson finite difference technique. Input to GTRAN includes a description of the piping network, the initial conditions, and any events (e.g. valve closings) occuring during the period of analysis. Output includes pressure, velocity, and density versus time. GTRAN is written in FORTRAN 77 for batch execution and has been implemented on a DEC VAX series computer. GTRAN was developed in 1983.

Description: Continuous optical observations of cusp/cleft auroral activities within approximately equal to 09-15 MLT and 70-76 deg magnetic latitude are studied in relation to changes in solar wind dynamic pressure and interplanetary magnetic field (IMF) variability. The observed latitudinal movements of the cusp/cleft aurora in response to IMF B(sub z) changes may be explained as an effect of a variable magnetic field intensity in the outer dayside magnetosphere associated with the changing intensity of region 1 field-aligned currents and associated closure currents. Ground magnetic signatures related to such currents were observed in the present case (January 10, 1993). Strong, isolated enhancements in solar wind dynamic pressure (Delta p/p is greater than or equal to 0.5) gave rise to equatorward shifts of the cusp/cleft aurora, characteristic auroral transients, and distinct ground magnetic signatures of enhanced convection at cleft latitudes. A sequence of auroral events of approximately equal to 5-10 min recurrence time, moving eastward along the poleward boundary of the persistent cusp/cleft aurora in the approximately equal to 10-14 MLT sector, during negative IMF B(sub z) and B(sub y) conditions, were found to be correlated with brief pulses in solar wind dynamic pressure (0.1 is less than Delta p/p is less than 0.5). Simultaneous photometer observations from Ny Alesund, Svalbard, and Danmarkshavn, Greenland, show that the events often appeared on the prenoon side (approximately equal to 10-12 MLT), before moving into the postnoon sector in the case we study here, when IMF B(sub y) is less than 0. In other cases, similar auroral event sequences have been observed to move westward in the prenoon sector, during intervals of positive B(sub y). Thus a strong prenoon/postnoon asymmetry of event occurence and motion pattern related to the IMF B(sub y) polarity is observed. We find that this category of auroral event sequence is stimulated bursts of electron precipitation that originate from magnetosheath plasma that has accessed that dayside magnetosphere in the noon or near-noon sector, possibly at high latitudes, partly governed by the IMF orientation as well as by solar wind dynamic pressure pulses.

Description: Results are presented for the numerical simulation of unsteady viscous incompressible flow past thick airfoils. Specifically, flow past a NACA 4424 at an angle of attack of 2.5 deg and Reynolds numbers in the range of 1700-4000 has been simulated using the spectral element method. At these conditions the flow is separatedd and an unsteady wake is formed. Application of the method of empirical eigenfunction reveals the structure of the most energetic components of the flow. These are found to occur in pairs that, through phase exchange, are responsible for the vortex shedding. A set of ordinary differential equations is obtained for the amplitudes of these eigenfunctions by a Galerkin projection of the Navier-Stokes equations. The solutions of the model system are compared with the full simulation. The work is of relevance to the transition process and observed routes to chaos in airfoil wakes.

Description: The objective of this study is to develop a reduced mechanism for ethylene oxidation. The authors are interested in a model with a minimum number of species and reactions that still models the chemistry with reasonable accuracy for the expected combustor conditions. The model will be validated by comparing the results to those calculated with a detailed kinetic model that has been validated against the experimental data.

Description: Artificial viscosity is added either implicity or explicitly in practically every numerical scheme for suppressing spurious oscillations in the solution of fluid-dynamics equations. In the present central-difference scheme, artificial viscosity is added explicitly for suppressing high-frequency oscillations and achieving good convergence properties. The amount of artificial viscosity added is controlled through the use of preselected coefficients. In the standard scheme, scalar coefficients based on the spectral radii of the Jacobian of the convective fluxes are used. However, this can add too much viscosity to the slower waves. Hence, the use of matrix-valued coefficients, which give appropriate viscosity for each wave component, is suggested. With the matrix-valued coefficients, the central-difference scheme produces more accurate solutions on a given grid, particularly in the vicinity of shocks and boundary layers, while still maintaining good convergence properties.

Description: A potential flow based three-dimensional panel method was modified to treat time-dependent conditions in which several submerged bodies can move within the fluid along different trajectories. This modification was accomplished by formulating the momentary solution in an inertial frame of reference, attached to the undisturbed stationary fluid. Consequently, the numerical interpretation of the multiple-body, solid-surface boundary condition and the viscous wake rollup was considerably simplified. The usteady capability of this code was calibrated and validated by comparing computed results with closed-form analytical results available for an airfoil, which was impulsively set into a constant speed forward motion. To demonstrate the multicomponent capability, computations were made for two wings following closely intersecting paths (i.e., simulations aimed at avoiding mid-air collisions) and for a flowfield with relative rotation (i.e., the case of a helicopter rotor rotating relative to the fuselage). Computed results for the cases were compared to experimental data, when such data was available.

Description: Stratospheric aerosols, collected near 19 km altitude on wire impactors over western North America from August 20, 1991 to May 11, 1993, show strong influence of the June 1991 Mt. Pinatubo eruption. Lognormal size distributions are bimodal; each of the mode radii increases and reaches maximum value at about 15 months after eruption. The second (large particle) mode becomes well developed then, and about 40% of the droplets are larger than 0.4 micron radius. The eruption of Mt. Spurr (Alaska) may also have contributed to this. Sulfate mass loading decays exponentially (e-folding 216 days), similar to El Chichon. Silicates are present in samples only immediately after eruption. Two years after eruption, sulfate mass loading is about 0.4 micrograms/cu m, about an order of magnitude higher than background pre-volcanic values. Aerosol size distributions are still bimodal with a very well-defined large droplet mode.

Description: Comparisons are presented between Nimbus 7 LIMS (Limb Infrared Monitor of the Stratosphere) mapped temperatures and both Datasonde and sphere in situ rocketsonde temperature measurements. With this approach up to 666 LIMS/Datasonde pairs were obtained for various pressure levels to look for small biases in LIMS temperatures as a function of altitude, latitude and season. Between 10-1 hPa LIMS and Datasonde agree everywhere to better than +/- 2 K with the exception of a warm bias of about 3 K at 2 hPa at high latitudes. However, LIMS is colder than the Datasonde by about 4 K at 0.4 hPa and by about 8-10 K at 0.1 hPa. When compared with the more accurate sphere temperatures the bias at 0.1 hPa is reduced by nearly one-half. These results indicate that the LIMS zonal mean constituent profiles are nearly free of temperature bias, except perhaps at 0.1 hPa.