SEARCH · Engineering Papers
Results for “numerical calculations: variational”
Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.
Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.
Nonlinear saturation of ballooning modes in stellarators
Ballooning mode saturation is investigated in realistic stellarator configurations using the flux tube approach of Ham et al (2018 Plasma Phys. Control. Fusion 60 075017), Ham et al (2016 Phys. Rev. Lett. 116 235001). The method is adapted to account for the lack of exact force balance in stellarator equilibrium solvers that assume existence of nested flux surfaces. A variational approach for calculating flux tube energy is developed to overcome this force error problem in stellarator numerical equilibria. Saturated (equilibrium) flux tube states that cross 10%–20% of the plasma minor radius are shown to exist for linearly ballooning unstable profiles. It is shown that several features of the displaced flux tube structure in a full nonlinear MHD simulation of Wendelstein 7X are reproduced by our model. Saturated states are found in a compact stellarator equilibrium close but below the marginal ballooning linear instability, i.e. the unperturbed equilibrium is metastable. This suggests that edge-localized-mode-like explosive MHD behavior may be possible in stellarators.
On the Laplace relation among the satellites of Uranus
Three of Uranus' satellites are involved in an interaction similar to the Laplace relation among the Galilean satellites of Jupiter. However, there are no pair-wise commensurabilities in the Uranus system as there are among the Galilean satellites. Thus analysis by Fourier expansion of disturbing functions becomes too unwieldy for practical calculation. Instead, the unexpanded perturbations are numerically averaged over a time scale short compared to the important variations in the Laplace relation. Perturbations of Miranda's orbit are estimated to be of observable magnitude for reasonable masses of Ariel and Umbriel.
A nonlinear theory for spinning anisotropic beams using restrained warping functions
A geometrically nonlinear theory is developed for spinning anisotropic beams having arbitrary cross sections. An assumed displacement field is developed using the standard 3D kinematics relations to describe the global beam behavior supplemented with an additional field that represents the local deformation within the cross section and warping out of the cross section plane. It is assumed that the magnitude of this additional field is directly proportional to the local stress resultants. In order to take into account the effects of boundary conditions, a restraining function is introduced. This function plays the role of reducing the amount of free warping deformation throughout the field due to the restraint of the cross section(s) at the end(s) of the beam, e.g., in the case of a cantilever beam. Using a developed ordering scheme, the nonlinear strains are calculated to the third order. The FEM is developed using the weak form variational formulation. Preliminary interesting numerical results have been obtained that indicate the role of the restraining function in the case of a cantilever beam with circular cross section. These results are for the cases of a tip displacement (static) and free vibration studies for both isotropic and anisotropic materials with varied fiber orientations.
Hydrostatic temperature calculations
Comparisons are made between hydrostatically computed temperatures and ambient temperatures associated with nine different data sources, including analyses, forecasts and conventional observations. Five-day averages and the day-to-day variations in the root-mean-square temperature differences are presented. Several different numerical and interpolation procedures are examined. Error correction and a constrained optimum procedure that minimizes ambient minus calculated hydrostatic temperature differences are introduced. Systematic differences between ambient and hydrostatic temperatures are found to be associated with the sinoptic situation. When compared with ambient temperatures, hydrostatic temperatures at 500 mb tend to be too warm at or in front of a trough and too cold behind the trough. In the vertical direction, for the eight-level configuration tested, the average hydrostatic temperatures are too cold at low levels (850, 700 mb) and too warm at upper levels, (300, 250 mb).
Elastic Buckling of Laminated Plates Under Varying Axial Stresses
The elastic buckling load of simply supported rectangular composite plates subjected to a second degree parabolic variation of axial stresses in the longitudinal direction is calculated using analytical methods. The variation of axial stresses is equilibrated by nonuniform shear stresses along the plate edges and transverse normal stresses. Numerical results are reported for three different cases: (1) orthotropic plates, (2) symmetrically laminated plates with multiple generally orthotropic layers exhibiting coupling between normal moments and twist, and twisting moment and normal curvatures, and (3) unsymmetrically laminated plates. Rayleigh-Ritz method is used to calculate the buckling load. An approximate solution using "reduced bending stiffness" is adopted for unsymmetrically laminated plates. The influence of the aspect ratio is examined, and the results are compared with plates subjected to uniform axial stresses.
Numerical model of solar dynamic radiator for parametric analysis
Growth power requirements for Space Station Freedom will be met through addition of 25 kW solar dynamic (SD) power modules. The SD module rejects waste heat from the power conversion cycle to space through a pumped-loop, multi-panel, deployable radiator. The baseline radiator configuration was defined during the Space Station conceptual design phase and is a function of the state point and heat rejection requirements of the power conversion unit. Requirements determined by the overall station design such as mass, system redundancy, micrometeoroid and space debris impact survivability, launch packaging, costs, and thermal and structural interaction with other station components have also been design drivers for the radiator configuration. Extensive thermal and power cycle modeling capabilities have been developed which are powerful tools in Station design and analysis, but which prove cumbersome and costly for simple component preliminary design studies. In order to aid in refining the SD radiator to the mature design stage, a simple and flexible numerical model was developed. The model simulates heat transfer and fluid flow performance of the radiator and calculates area mass and impact survivability for many combinations of flow tube and panel configurations, fluid and material properties, and environmental and cycle variations. A brief description and discussion of the numerical model, it's capabilities and limitations, and results of the parametric studies performed is presented.
Interaction between an elastic circular inclusion and two symmetrically placed collinear cracks
The paper investigates the fracture toughness of a flat large elastic cracked plate containing an elastic circular inclusion, using the plane stress crack-tip stress intensity factor as a criterion for fracture. In this plane stress geometry, each of the two collinear finite cracks is located on either side of the elastic circular inclusion and the geometry is subjected to uniform stresses at infinity. The analysis is based on the two-dimensional theory of elasticity by using the Muskhelishvili complex variable approach. Numerical calculations are reported for the case of simple tension normal to the crack direction and show the variation of the stress intensity factor with the configuration and elastic properties of the plate and the inclusion.
Broadband jet noise amplification by a pure tone excitation
A model is proposed for the change in turbulent structure of a round jet in the presence of an acoustic excitation. The excitation is initial amplitude at the jet exit. As these waves propagate downstream they extract energy from the mean flow and transfer it to the random turbulence. This results in an increase in the levels of the turbulence and a resulting increase in the radiated broadband noise. An examination is made of the effect of excitation level and frequency on the jet flow. The numerical procedure allows for radial as well as axial variations in the averaged properties of jet to be calculated. The results indicate that the presence of a finite amplitude instability wave increases the spreading of the jet.
Numerical study of the transient flow in the driven tube and the nozzle section of a shock tunnel
The initial flow in a shock tunnel was examined numerically using computational fluid dynamics (CFD). A finite-volume total variation diminishing (TVD) scheme was used to calculate the transient flow in a shock tunnel. Both viscous and inviscid, chemically nonreacting flows were studied. The study consisted of two parts, the first dealt with the transient flow in the driven-tube/nozzle interface region (inviscid calculations). The effects of varying the geometry in this region was examined. The second part of the study examined the transient flow in the nozzle (viscous calculations). The results were compared to experimental data.
The magnetic field of solar prominences
A model is presented which accounts for the formation of coronal magnetic field lines with the appropriate 'dipped' structure to support prominences. The critical ingredients of the model are that the prominence magnetic field is a truly three-dimensional structure with significant variation along the prominence length, and the magnetic field is strongly sheared near the photospheric neutral line. Numerical calculations are presented which demonstrate that these two features lead to dip formation. In addition our model is able to account for the long-puzzling observation of inverse polarity in quiescent prominences.
Turbulence and deterministic chaos
Several turbulent and nonturbulent solutions of the Navier-Stokes equations are obtained. The unaveraged equations are used numerically in conjunction with tools and concepts from nonlinear dynamics, including time series, phase portraits, Poincare sections, largest Liapunov exponents, power spectra, and strange attractors. Initially neighboring solutions for a low Reynolds number fully developed turbulence are compared. Several flows are noted: fully chaotic, complex periodic, weakly chaotic, simple periodic, and fixed-point. Of these, only fully chaotic is classified as turbulent. Besides the sustained flows, a flow which decays as it becomes turbulent is examined. For the finest grid, 128(exp 3) points, the spatial resolution appears to be quite good. As a final note, the variation of the velocity derivatives skewness of a Navier-Stokes flow as the Reynolds number goes to zero is calculated numerically. The value of the skewness is shown to become small at low Reynolds numbers, in agreement with intuitive arguments that nonlinear terms should be negligible.
Numerical and experimental simulations of the postbuckling response of laminated anisotropic panels
Numerical simulations are used in conjuction with experiments to study the buckling and postbuckling responses and failure initiation of flat unstiffened composite panels. The numerical simulations are conducted using two-dimensional shear-flexible finite elements. The effect of the laminate stacking sequence on the buckling and postbuckling responses is studied. Correlation between numerical and experimental results is good through buckling, but the numerical models overestimate the postbuckling stiffness of the panels. To explain the discrepancies in the postbuckling stiffnesses, analytic sensitivity derivatives are calculated and are used to study the sensitivity of the buckling and postbuckling responses to variations in different material and lamination parameters of the panel. Experimental results indicate that failure occurs along a nodal line. Numerical results show that the location of failure initiation corresponds to that of the maximum transverse shear strain energy density in the panel, which occurs at the edge of the panel at a nodal line.
Influence of fiber packing structure on permeability
The study on the permeability of an aligned fiber bundle is the key building block in modeling the permeability of advanced woven and braided preforms. Available results on the permeability of fiber bundles in the literature show that a substantial difference exists between numerical and analytical calculations on idealized fiber packing structures, such as square and hexagonal packing, and experimental measurements on practical fiber bundles. The present study focuses on the variation of the permeability of a fiber bundle under practical process conditions. Fiber bundles are considered as containing openings and fiber clusters within the bundle. Numerical simulations on the influence of various openings on the permeability were conducted. Idealized packing structures are used, but with introduced openings distributed in different patterns. Both longitudinal and transverse flow are considered. The results show that openings within the fiber bundle have substantial effect on the permeability. In the longitudinal flow case, the openings become the dominant flow path. In the transverse flow case, the fiber clusters reduce the gap sizes among fibers. Therefore the permeability is greatly influenced by these openings and clusters, respectively. In addition to the porosity or fiber volume fraction, which is commonly used in the permeability expression, another fiber bundle status parameter, the ultimate fiber volume fraction, is introduced to capture the disturbance within a fiber bundle.
Thermodynamics and kinetics of reactions in protective coating systems
Investigations of fluoride activated packs with Al:Ni ratios greater than 50 a/o prove that the specimen surface is not in equilibrium with the pack at high Al:Ni ratios but that an activity gradient exists between pack and specimen. Therefore, gaseous diffusion and possibly surface reactions play a role in determining the overall rate of Al deposition in such packs. Noticeable differences in coating behavior have been obtained in packs activated with chloride and iodide, and it appears that poorest results are obtained with iodides, better with chlorides, and best with fluorides. A numerical method has been perfected for calculating rates of solid-state diffusion controlled coating formation, allowing for the variation of diffusivity with composition in the NiAl phase. Layer growth rates can now be accurately predicted from a knowledge of the surface and substrate compositions. Furthermore, the correct diffusion profiles are obtained by this method. These differ substantially from the profile obtained when the diffusivity is assumed constant.
Nonlinear filters for efficient shock computation
A new type of methods for the numerical approximation of hyperbolic conservation laws with discontinuous solution is introduced. The methods are based on standard finite difference schemes. The difference solution is processed with a nonlinear conservation form filter at every time level to eliminate spurious oscillations near shocks. It is proved that the filter can control the total variation of the solution and also produce sharp discrete shocks. The method is simpler and faster than many other high resolution schemes for shock calculations. Numerical examples in one and two space dimensions are presented.
Analytic fits to atom-in-jellium shear modulus predictions
Fits to AJ shear modulus calculations from ρ0 to the limit of the corresponding AJ EOS table (usually 1000 ρ0), adjusted to reproduce STP shear modulus where possible. Fitting minimized the fractional difference, so low-pressure points would be reproduced accurately despite the wide-ranging tabulation. AJ calculations were at T=0. Isochoric variation G(T) has been small for all cases spot-tested so far, and may have either sign. The fitted equation usually matches the AJ data to within a few percent. Between numerical noise in the AJ calculation and probably-physical structure not captured by the equation, the deviation could be up to 20% in some places in most models, and 30% in a few. AJ does not capture structural phase transitions or phases stabilized by directional bonds, where the shear modulus may vary by a greater amount. Where possible, STP ρ,G were used as parameters, and low pressure AJ points were de-weighted or removed if necessary. Otherwise, such as where AJ fails to capture solid phases with a significantly different shear modulus, the AJ data were fitted as far down in pressure as possible, G0 was also fitted if necessary, and ρ0 was also adjusted if needed to keep G0>0. The resulting models are not intended for use at low pressure, though some are probably adequate for practical purposes
Dynamical effects on the global distribution of thermospheric atomic oxygen
Results are presented of calculations pertaining to the effects of global-scale dynamical processes on the coupled seasonal and diurnal variations of atomic oxygen in the altitude range 90-500 km. The calculations are based on a three-dimensional numerical model of thermospheric dynamics which treats the three gases N2, O2, and O simultaneously, coupled together by collisional processes. The model is described, and the results of the calculations under both equinox and solstice conditions and a comparison with observational data are discussed. It is shown that none of these gases can be considered to be in hydrostatic equilibrium below about 200 km, and that global-scale winds cause the observed winter enhancement of atomic oxygen.