Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “Smooth approximation”

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.

At least 199 records · Page 11
All sourcesNASA NTRSDOE OSTI
NASA NTRS1973

A computer program for calculating the perfect gas inviscid flow field about blunt axisymmetric bodies at an angle of attack of 0 deg

A method for the rapid calculation of the inviscid shock layer about blunt axisymmetric bodies at an angle of attack of 0 deg has been developed. The procedure is of an inverse nature, that is, a shock wave is assumed and calculations proceed along rays normal to the shock. The solution is iterated until the given body is computed. The flow field solution procedure is programed at the Langley Research Center for the Control Data 6600 computer. The geometries specified in the program are sphores, ellipsoids, paraboloids, and hyperboloids which may conical afterbodies. The normal momentum equation is replaced with an approximate algebraic expression. This simplification significantly reduces machine computation time. Comparisons of the present results with shock shapes and surface pressure distributions obtained by the more exact methods indicate that the program provides reasonably accurate results for smooth bodies in axisymmetric flow. However, further research is required to establish the proper approximate form of the normal momentum equation for the two-dimensional case.

Zoby, E. V.↗
NASA NTRS1999

Radiative Smoothing in Clouds at Transparent and Absorbing Wavelengths

For absorbing and transparent wavelengths, we discuss the effect of horizontal solar radiative fluxes in clouds on the accuracy of a conventional plane-parallel radiative transfer calculations for a single pixel, known as the Independent Pixel Approximation (IPA). We address the question of correlations between horizontal fluxes, IPA accuracies and radiative smoothing. By smoothing we understand a radiative transfer process whereby radiation does not follow the small-scale fluctuations of cloud structure, producing much smoother radiation fields. The scale eta that characterizes this process is called "radiative smoothing scale." We relate radiative smoothing to the photon's horizontal displacement that characterizes a "spot" of reflected light associated with a point source. We generalize the "spot-size" estimate derived for conservative scattering using the diffusion theory to the case of non-conservative scattering. For reflected light, theoretical results are confirmed with numerical simulations. The radiative smoothing scale eta is a critical value where IPA effectively breaks down; for scales smaller than TI, real radiation field are much smoother than their IPA counterparts for the same cloud structure. In addition to the estimate of il for absorbing wavelengths, we show that: (1) with more absorption, the scale break determined by eta in a log-log plot of wavenumber spectra moves towards smaller scales and (2) the smaller eta the flatter the small-scale slope which means less radiative smoothing, thus more accuracy in the IPA reflection.

Marshak, A.↗
DOE OSTI2023

Cross-correlating radial peculiar velocities and CMB lensing convergence

We study, for the first time, the cross correlation between the angular distribution of radial peculiar velocities (PV) and the lensing convergence of cosmic microwave background (CMB) photons. We derive theoretical expectations for the signal and its covariance and assess its detectability with existing and forthcoming surveys. We find that such cross-correlations are expected to improve constraints on different gravitational models by partially breaking degeneracies with the matter density. We identify in the distance-scaling dispersion of the peculiar velocities the most relevant source of noise in the cross correlation. For this reason, we also study how the above picture changes assuming a redshift-independent scatter for the PV, obtained for example using a reconstruction technique. Our results show that the cross correlation might be detected in the near future combining PV measurements from DESI and the convergence map from CMB-S4. Using realistic direct PV measurements we predict a cumulative signal-to-noise ratio of approximately 3.8σ using data on angular scales 3 ≤ ℓ ≤ 200. For an idealized reconstructed peculiar velocity map extending up to redshift z = 0.15 and a smoothing scale of 4 Mpc h -1 we predict a cumulative signal-to-noise ratio of approximately 27σ from angular scales 3 ≤ ℓ ≤ 200. We conclude that currently reconstructed peculiar velocities have more constraining power than directly observed ones, even though they are more cosmological-model dependent.

79 ASTRONOMY AND ASTROPHYSICS↗
NASA NTRS1977

Minimal-resource computer program for automatic generation of ocean wave ray or crest diagrams in shoaling waters

A computer program for studying linear ocean wave refraction is described. The program features random-access modular bathymetry data storage. Three bottom topography approximation techniques are available in the program which provide varying degrees of bathymetry data smoothing. Refraction diagrams are generated automatically and can be displayed graphically in three forms: Ray patterns with specified uniform deepwater ray density, ray patterns with controlled nearshore ray density, or crest patterns constructed by using a cubic polynomial to approximate crest segments between adjacent rays.

Poole, L. R.↗
NASA NTRS1991

Recent developments in shock-capturing schemes

The development of the shock capturing methodology is reviewed, paying special attention to the increasing nonlinearity in its design and its relation to interpolation. It is well-known that higher-order approximations to a discontinuous function generate spurious oscillations near the discontinuity (Gibbs phenomenon). Unlike standard finite-difference methods which use a fixed stencil, modern shock capturing schemes use an adaptive stencil which is selected according to the local smoothness of the solution. Near discontinuities this technique automatically switches to one-sided approximations, thus avoiding the use of discontinuous data which brings about spurious oscillations.

Harten, Ami↗
NASA NTRS1993

On the prediction of multigrid efficiency through local mode analysis

A single grid local mode analysis is used to predict the smoothing properties of numerical schemes for solving the Navier-Stokes equations with factorization based on Stone's Strongly Implicit Method. Four difference approximations for the convection terms are considered, namely, hybrid, central, second-order upwind, and third-order upwind. Smoothing factors from the analysis are compared with practical convergence factors in a multigrid method for flow over a backward facing step and it is found that the local mode analysis correctly predicts the effects of Reynolds number and higher-order schemes.

Wilson, R. V.↗
NASA NTRS1986

Intelligent information extraction from reflectance spectra Absorption band positions

A multiple high-order derivative analysis algorithm has been developed which can automatically extract absorption band positions from low-quality reflectance spectra with little degredation of accuracy. Overlapping bands with comparable widths and intensities can be resolved whose centers are as close as 0.3-0.5 W, with safer resolution limits of 0.6-1.0 W band center separations suggested for overlapping bands that are dissimilar. The segment length for smoothing is continually adjusted to about 0.5 W to minimize signal distortion, and a spectral pattern recognition algorithm predicts the signal spectrum and calculates approximate W across the spectrum using its second derivative. A single-pass cubic spline is applied to the smoothed data, and a sliding segment sixth-order polynomial is fit to the spectrum, with the length of the segment being continuously locally adjusted to 1.0 W across the spectrum. Good reliability and consistency of the algorithm is demonstrated with application to laboratory and earth-based telescope spectra.

Huguenin, R. L.↗
NASA NTRS1939

Determination of the Profile Drag of an Airplane Wing in Flight at High Reynolds Numbers

Flight tests were made to determine the profile-drag coefficients of a portion of the original wing surface of an all-metal airplane and of a portion of the wing made aerodynamically smooth and more nearly fair than the original section. The wing section was approximately the NACA 2414.5. The tests were carried out over a range of airplane speeds giving a maximum Reynolds number of 15,000,000. Tests were also carried out to locate the point of transition from laminar to turbulent boundary layer and to determine the velocity distribution along the upper surface of the wing. The profile-drag coefficients of the original and of the smooth wing portions at a Reynolds number of 15,000,000 were 0.0102 and 0.0068, respectively; i.e., the surface irregularities on the original wing increased the profile-drag coefficient 50 percent above that of the smooth wing.

Bicknell, Joseph↗
DOE OSTI2021

An adaptive Hessian approximated stochastic gradient MCMC method

Bayesian approaches have been successfully integrated into training deep neural networks. One popular family is stochastic gradient Markov chain Monte Carlo methods (SG-MCMC), which have gained increasing interest due to their ability to handle large datasets and the potential to avoid overfitting. Although standard SG-MCMC methods have shown great performance in a variety of problems, they may be inefficient when the random variables in the target posterior densities have scale differences or are highly correlated. Here, we present an adaptive Hessian approximated stochastic gradient MCMC method to incorporate local geometric information while sampling from the posterior. The idea is to apply stochastic approximation (SA) to sequentially update a preconditioning matrix at each iteration. The preconditioner possesses second-order information and can guide the random walk of a sampler efficiently. Instead of computing and saving the full Hessian of the log posterior, we use limited memory of the samples and their stochastic gradients to approximate the inverse Hessian-vector multiplication in the updating formula. Moreover, by smoothly optimizing the preconditioning matrix via SA, our proposed algorithm can asymptotically converge to the target distribution with a controllable bias under mild conditions. To reduce the training and testing computational burden, we adopt a magnitude-based weight pruning method to enforce the sparsity of the network. Our method is user-friendly and demonstrates better learning results compared to standard SG-MCMC updating rules. The approximation of inverse Hessian alleviates storage and computational complexities for large dimensional models. Numerical experiments are performed on several problems, including sampling from 2D correlated distribution, synthetic regression problems, and learning the numerical solutions of heterogeneous elliptic PDE. The numerical results demonstrate great improvement in both the convergence rate and accuracy.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2021

Numerical and Analytical Modeling of Flow Partitioning in Partially Saturated Fracture Networks

Infiltration processes in fractured-porous media remain a crucial, yet not very well understood component of recharge and vulnerability assessment. Under partially-saturated conditions flows in fractures, percolating fracture networks, and fault zones contribute to the fastest spectrum of infiltration velocities via preferential pathways. Specifically, the partitioning dynamics at fracture intersections determine the magnitude of flow fragmentation into vertical and horizontal components and hence the bulk flow velocity and dispersion of fracture networks. In this work, we derive an approximate analytical solution for the partitioning process and validate it using smoothed particle hydrodynamics simulations. The developed transfer function is conceptually based on simulation results and laboratory experiments carried out in previous works. It allows to efficiently simulate flow through fracture networks with simple cubic structure and arbitrary number of fractures and aperture sizes via linear response theory and convolution of a given input signal. We derive a non-dimensional bulk flow velocity ($\widetilde{v}$) and dispersion coefficient ($\widetilde{D}$) to characterize fracture networks in terms of dimensionless horizontal and vertical time scales $\tau_m$ and $\tau_0$. The dispersion coefficient is shown to strongly depend on the horizontal time scale and converges towards a constant value of $0.08$ within reasonable fluid and geometrical parameter ranges, while the non-dimensional velocity exhibits a characteristic $\widetilde{v} \sim \tau_m^{-1/2}$ scaling. Given that hydraulic information is often only available at limited places within (fractured-porous) aquifer system, such as boreholes or springs, our study intends to provide a rudimentary analytical concept to potentially reconstruct internal fracture network geometries from external boundary information, e.g., the dispersive properties of discharge (groundwater level fluctuations).

42 ENGINEERING↗
NASA NTRS1972

Locally nonsimilar solutions for radiating shock layer about smooth axisymmetric bodies

Locally nonsimilar solutions for nongray radiating shock layers about smooth axisymmetric bodies have been obtained based on a newly developed approximate method. Good agreement is found with numerical solutions for inviscid cases (both radiating and nonradiating) and with series solutions for the radiating viscous case. For the inviscid case, the effect of radiative cooling is to destroy the entropy layer; at a distance far from the stagnation point, the shock layer is nearly isothermal. For the viscous case, the radiative wall flux approaches that of the inviscid case at a distance far downstream of the stagnation point. The method can also treat surface mass injection.

Chou, Y. S.↗
NASA NTRS1981

New approach to the solution of large, full matrix equations

A new approach to the solution of matrix equations resulting from integral equations is presented and applied to the solution of two-dimensional Neumann problems describing the inviscid, incompressible flow past an airfoil. The problem is reformulated in terms of a preselected set of mode functions giving an equivalent matrix equation to be solved for the mode-function expansion coefficients. Because of the inherent smoothness of the original problem, the coefficient problem can be solved approximately without significantly affecting the accuracy of the final solution. Very promising two-dimensional results are obtained and the extension of the method to three-dimensional problems is investigated. On the basis of these results it is shown that the computing time for the matrix solution for a large three-dimensional panel method calculation could be reduced by an order of magnitude compared with that required for a direct solution.

Clark, R. W.↗
NASA NTRS1985

Comparison of predicted and measured drag for a single-engine airplane

Renewed interest in natural laminar flow (NLF) has rekindled designers' concerns that manufacturing deviations, (loss of surface contours or other surface imperfections) may destroy the effectiveness of NLF for an operational airplane. This paper reports on experimental research that compares predicted and measured boundary layer transition, total drag, and two-dimensional drag coefficients for three different wing surface conditions on an airplane typical of general aviation manufacturing technology. The three flight test phases included: (1) assessment of an unpainted airframe, (2) flight tests of the same airplane after painstakingly filling and sanding the wings to design contours, and (3) similar measurements after this airplane was painted. In each flight phase, transition locations were monitored using either sublimating chemicals or pigmented oil. As expected, total drag changes were difficult to measure. Two-dimensional drag coefficients were estimated using the Eppler-Somers code and measured with a wake rake in a method very similar to Jones' pitot traverse method. The net change in two-dimensional drag was approximately 20 counts between the unpainted airplane and the 'hand-smoothed' airplane for typical cruise flight conditions.

Ward, D. T.↗
NASA NTRS1987

Measurement of the effect of manufacturing deviations on natural laminar flow for a single engine general aviation airplane

Renewed interest in natural laminar flow (NLF) had rekindled designer concern that manufacuring deviations may destroy the effectiveness of NLF for an operational aircraft. Experiments are summarized that attemtped to measure total drag changes associated with three different wing surface conditions on an aircraft typical of current general aviation high performance singles. The speed power technique was first used in an attempt to quantify the changes in total drag. Predicted and measured boundary layer transition locations for three different wing surface conditions were also compared, using two different forms of flow visualization. The three flight test phases included: assessment of an unpainted airframe, flight tests of the same aircraft after painstakingly filling and sanding the wings to design contours, and similar measurement after this aricraft was painted. In each flight phase, transition locations were monitored using with sublimating chemicals or pigmented oil. Two-dimensional drag coefficients were estimated using the Eppler-Somers code and measured with a wake rake in a method very similar to Jones' pitot traverse method. The net change in two-dimensional drag coefficient was approximately 20 counts between the unpainted aircraft and the hand-smoothed aircraft for typical cruise flight conditions.

Source record↗
NASA NTRS1986

Geomagnetic temporal change: 1903-1982 - A spline representation

The secular variation of the earth's magnetic field is itself subject to temporal variations. These are investigated with the aid of the coefficients of a series of spherical harmonic models of secular variation deduced from data for the interval 1903-1982 from the worldwide network of magnetic observatories. For some studies it is convenient to approximate the time variation of the spherical harmonic coefficients with a smooth, continuous, function; for this a spline fitting is used. The phenomena that are investigated include periodicities, discontinuities, and correlation with the length of day. The numerical data presented will be of use for further investigations and for the synthesis of secular variation at any place and at any time within the interval of the data - they are not appropriate for temporal extrapolations.

Langel, R. A.↗
NASA NTRS1992

A regularization method for extrapolation of solar potential magnetic fields

The mathematical basis of a Tikhonov regularization method for extrapolating the chromospheric-coronal magnetic field using photospheric vector magnetograms is discussed. The basic techniques show that the Cauchy initial value problem can be formulated for potential magnetic fields. The potential field analysis considers a set of linear, elliptic partial differential equations. It is found that, by introducing an appropriate smoothing of the initial data of the Cauchy potential problem, an approximate Fourier integral solution is found, and an upper bound to the error in the solution is derived. This specific regularization technique, which is a function of magnetograph measurement sensitivities, provides a method to extrapolate the potential magnetic field above an active region into the chromosphere and low corona.

Gary, G. A.↗
NASA NTRS1994

Electron injection and acceleration at nonlinear shocks: Results of numerical simulations

We present results of numerical simulations of electron injection and acceleration at nonlinear high Mach number shocks. The electrons are assumed to be heated at the thermal subshock to an energy E(sub inj), which is treated as a free parameter, above which they are injected by self-generated whistlers to momentum m(sub p) x V(sub A). This injection mechanism requires Mach numbers greater than (43/(beta(sub -)))(((k T(sub e))/(E(sub inj)))(exp 1/2)), where T(sub e) and beta (sub -) are the upstream electron temperature and plasma beta parameter. Above m(sub p) x V(sub A) electrons are trapped in the shock by Alfven waves. In the proton precursor region the Alfven waves are assumed to be generated by protons accelerated at the shock, and have nonlinear intensities. Below GeV, however, electrons of a given rigidity propagate faster than protons with a similar rigidity and therefore diffuse to regions ahead of the proton precursor. In those regions the Alfven waves are generated by the electrons themselves. The diffusion coefficient appears to increase with decreasing acceleration efficiency. As a result, the number of electrons accelerated to energies GeV and above and, hence, the electron to proton ratio, depend only weakly on the extent of electron heating at the subshock. The negative feedback also renders the electron spectra insensitive to shock compression ratio and smoothing length scale. The estimated e/p ratio at GeV is between approximately 1%-10%.

Levinson, Amir↗
DOE OSTI2025

Bias-Variance Trade-Off in Physics-Informed Neural Networks with Randomized Smoothing for High-Dimensional PDEs

Physics-Informed Neural Networks (PINNs) have triggered a paradigm shift in scientific computing, leveraging mesh-free properties and robust approximation capabilities. While proving effective for low-dimensional partial differential equations (PDEs), the computational cost of PINNs remains a hurdle in high-dimensional scenarios. This is particularly pronounced when computing high-order and high-dimensional derivatives in the physics-informed loss. Randomized Smoothing PINN (RS-PINN) introduces Gaussian noise for stochastic smoothing of the original neural net model, enabling the use of Monte Carlo methods for derivative approximation, which eliminates the need for costly automatic differentiation. Despite its computational efficiency, especially in the approximation of high-dimensional derivatives, RS-PINN introduces biases in both loss and gradients, negatively impacting convergence, especially when coupled with stochastic gradient descent (SGD) algorithms. We present a comprehensive analysis of biases in RS-PINN, attributing them to the nonlinearity of the Mean Squared Error (MSE) loss as well as the intrinsic nonlinearity of the PDE itself. We propose tailored bias correction techniques, delineating their application based on the order of PDE nonlinearity. The derivation of an unbiased RS-PINN allows for a detailed examination of its advantages and disadvantages compared to the biased version. Specifically, the biased version has a lower variance and runs faster than the unbiased version, but it is less accurate due to the bias. To optimize the bias-variance trade-off, we combine the two approaches in a hybrid method that balances the rapid convergence of the biased version with the high accuracy of the unbiased version. In addition to methodological contributions, we present an enhanced implementation of RS-PINN. Extensive experiments on diverse high-dimensional PDEs, including Fokker-Planck, Hamilton-Jacobi-Bellman (HJB), viscous Burgers’, Allen-Cahn, and Sine-Gordon equations, illustrate the bias-variance trade-off and highlight the effectiveness of the hybrid RS-PINN. Empirical guidelines are provided for selecting biased, unbiased, or hybrid versions, depending on the dimensionality and nonlinearity of the specific PDE problem.

97 MATHEMATICS AND COMPUTING↗