Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “computational efficiency”

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 307 records · Page 17

Redesign of the DFT/MRCI Hamiltonian

The combined density functional theory and multireference configuration interaction (DFT/MRCI) method of Grimme and Waletzke [J. Chem. Phys. 111, 5645 (1999)] is a well-established semi-empirical quantum chemical method for efficiently computing excited-state properties of organic molecules. As it turns out, the method fails to treat bi-chromophores owing to the strong dependence of the parameters on the excitation class. In this work, we present an alternative form of correcting the matrix elements of a MRCI Hamiltonian which is built from a Kohn-Sham set of orbitals. It is based on the idea of constructing individual energy shifts for each of the state functions of a configuration. The new parameterization is spin-invariant and incorporates less empirism compared to the original formulation. By utilizing damping techniques together with an algorithm of selecting important configurations for treating static electron correlation, the high computational efficiency has been preserved. The robustness of the original and redesigned Hamiltonians has been tested on experimentally known vertical excitation energies of organic molecules yielding similar statistics for the two parameterizations. Besides that, our new formulation is free from artificially low-lying doubly excited states, producing qualitatively correct and consistent results for excimers. The way of modifying matrix elements of the MRCI Hamiltonian presented here shall be considered as default choice when investigating photophysical processes of bi-chromophoric systems such as singlet fission or triplet-triplet upconversion.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

Dynamical modeling of tidal streams

I present a new framework for modeling the dynamics of tidal streams. The framework consists of simple models for the initial action-angle distribution of tidal debris, which can be straightforwardly evolved forward in time. Taking advantage of the essentially one-dimensional nature of tidal streams, the transformation to position-velocity coordinates can be linearized and interpolated near a small number of points along the stream, thus allowing for efficient computations of a stream's properties in observable quantities. I illustrate how to calculate the stream's average location (its 'track') in different coordinate systems, how to quickly estimate the dispersion around its track, and how to draw mock stream data. As a generative model, this framework allows one to compute the full probability distribution function and marginalize over or condition it on certain phase-space dimensions as well as convolve it with observational uncertainties. This will be instrumental in proper data analysis of stream data. In addition to providing a computationally efficient practical tool for modeling the dynamics of tidal streams, the action-angle nature of the framework helps elucidate how the observed width of the stream relates to the velocity dispersion or mass of the progenitor, and how the progenitors of 'orphan' streams could be located. The practical usefulness of the proposed framework crucially depends on the ability to calculate action-angle variables for any orbit in any gravitational potential. A novel method for calculating actions, frequencies, and angles in any static potential using a single orbit integration is described in the Appendix.

79 ASTRONOMY AND ASTROPHYSICS↗

Likelihood-informed dimension reduction for nonlinear inverse problems

The intrinsic dimensionality of an inverse problem is affected by prior information, the accuracy and number of observations, and the smoothing properties of the forward operator. From a Bayesian perspective, changes from the prior to the posterior may, in many problems, be confined to a relatively low-dimensional subspace of the parameter space. We present a dimension reduction approach that defines and identifies such a subspace, called the 'likelihood-informed subspace' (LIS), by characterizing the relative influences of the prior and the likelihood over the support of the posterior distribution. This identification enables new and more efficient computational methods for Bayesian inference with nonlinear forward models and Gaussian priors. In particular, we approximate the posterior distribution as the product of a lower-dimensional posterior defined on the LIS and the prior distribution marginalized onto the complementary subspace. Markov chain Monte Carlo sampling can then proceed in lower dimensions, with significant gains in computational efficiency. We also introduce a Rao–Blackwellization strategy that de-randomizes Monte Carlo estimates of posterior expectations for additional variance reduction. We demonstrate the efficiency of our methods using two numerical examples: inference of permeability in a groundwater system governed by an elliptic PDE, and an atmospheric remote sensing problem based on Global Ozone Monitoring System (GOMOS) observations.

42 ENGINEERING↗

SPARTAN (Scalable Probabilistic Application Reconfigurable Tensor Autonomous Network)

The technical founder of Ludwig Computing Inc has been competitively selected for support by Cyclotron Road, a U.S. Department of Energy (DOE) Advanced Manufacturing Office (AMO) Lab-Embedded Entrepreneurship Program (LEEP) through an approved merit review process. Ludwig Computing Inc, supported by the U.S. Department of Energy's Advanced Manufacturing Office through the Cyclotron Road program, has investigated the advantages of probabilistic computing for real-world compute-intensive applications. This research adds to the understanding of alternative computing paradigms by exploring a unique hardware-software co-design that integrates quantum computing methods with nature-inspired problem-solving techniques. The project's focus on areas such as combinatorial optimization, graph analytics, and machine learning demonstrates the potential for significant advancements in computational efficiency and performance. By harnessing natural randomness to streamline large circuits into fewer devices, Ludwig's approach enables massive parallelism, potentially offering higher throughput, speed, and energy efficiency compared to conventional hardware solutions. This work benefits the public by paving the way for more efficient computing solutions that could address complex real-world problems while potentially reducing energy consumption in data-intensive industries.

97 MATHEMATICS AND COMPUTING↗

Analysis of axisymmetric body effects on rotor aerodynamics using modified slender body theory

A modified slender body theory was used in modelling the flowfield around an axisymmetric body at zero angle of attack. The model was defined in an attempt to characterize body-induced effects on helicopter rotor performance and loading using efficient computation methods. The streamfunction for an axial source distribution was applied in the body model, which was then adapted to the helicopter analysis code (CAMRAD). The body-induced velocities were calculated for separate locations on the rotor disk. Three different rotor models in nonuniform inflow with a free wake geometry were considered. Comparison of predictions with data on ellipsoidal and axisymmetric bodies showed good agreement for moderately thick ellipsoidal bodies. Oscillatory blade bending moments were increased by the presence of the body, but overall performance impact was negligible. Computational efficiency was 10-20 percent better than with a potential flow panel code.

Yamauchi, G. K.↗

A high-accuracy optical linear algebra processor for finite element applications

Optical linear processors are computationally efficient computers for solving matrix-matrix and matrix-vector oriented problems. Optical system errors limit their dynamic range to 30-40 dB, which limits their accuray to 9-12 bits. Large problems, such as the finite element problem in structural mechanics (with tens or hundreds of thousands of variables) which can exploit the speed of optical processors, require the 32 bit accuracy obtainable from digital machines. To obtain this required 32 bit accuracy with an optical processor, the data can be digitally encoded, thereby reducing the dynamic range requirements of the optical system (i.e., decreasing the effect of optical errors on the data) while providing increased accuracy. This report describes a new digitally encoded optical linear algebra processor architecture for solving finite element and banded matrix-vector problems. A linear static plate bending case study is described which quantities the processor requirements. Multiplication by digital convolution is explained, and the digitally encoded optical processor architecture is advanced.

Casasent, D.↗

Compression behavior of delaminated composite plates

The response of delaminated composite plates to compressive in-plane loads was investigated. The delaminated region may be either circular or elliptical, and may be located between any two plies of the laminate. For elliptical delaminations, the axes of the ellipse may be arbitrarily oriented with respect to the applied loads. A model was developed that describes the stresses, strains, and deformation of the sublaminate created by the delamination. The mathematical model is based on a two dimensional nonlinear plate theory that includes the effects of transverse shear deformation. The model takes into account thermal and moisture induced strains, transverse pressures acting on the sublaminate, and contact between the sublaminate and plate. The solution technique used is the Ritz method. A computationally efficient computer implementation of the model was developed. The code can be used to predict the nonlinear-load-strain behavior of the sublaminate including the buckling load, postbuckling behavior, and the onset of delamination growth. The accuracy of the code was evaluated by comparing the model results to benchmark analytical solutions. A series of experiments was conducted on Fiberite T300/976 graphite/epoxy laminates bonded to an aluminum honeycomb core forming a sandwich panel. Either circles or ellipses made from Teflon film were embedded in the laminates, simulating the presence of a delamination. Each specimen was loaded in compression and the strain history of the sublaminate was recorded far into the postbuckling regime. The extent of delamination growth was evaluated by C-scan examination of each specimen. The experimental data were compared to code predictions. The code was found to describe the data with reasonable accuracy. A sensitivity study examined the relative importance of various material properties, the delamination dimensions, the contact model, the transverse pressure differential, the critical strain energy release rate, and the relative growth direction on the buckling load, the postbuckling behavior, and the growth load of the sublaminate.

Peck, Scott O.↗

Modeling of Unsteady Three-dimensional Flows in Multistage Machines

Despite many years of development, the accurate and reliable prediction of unsteady aerodynamic forces acting on turbomachinery blades remains less than satisfactory, especially when viewed next to the great success investigators have had in predicting steady flows. Hall and Silkowski (1997) have proposed that one of the main reasons for the discrepancy between theory and experiment and/or industrial experience is that many of the current unsteady aerodynamic theories model a single blade row in an infinitely long duct, ignoring potentially important multistage effects. However, unsteady flows are made up of acoustic, vortical, and entropic waves. These waves provide a mechanism for the rotors and stators of multistage machines to communicate with one another. In other words, wave behavior makes unsteady flows fundamentally a multistage (and three-dimensional) phenomenon. In this research program, we have has as goals (1) the development of computationally efficient computer models of the unsteady aerodynamic response of blade rows embedded in a multistage machine (these models will ultimately be capable of analyzing three-dimensional viscous transonic flows), and (2) the use of these computer codes to study a number of important multistage phenomena.

Hall, Kenneth C.↗

Alloy Design Workbench-Surface Modeling Package Developed

NASA Glenn Research Center's Computational Materials Group has integrated a graphical user interface with in-house-developed surface modeling capabilities, with the goal of using computationally efficient atomistic simulations to aid the development of advanced aerospace materials, through the modeling of alloy surfaces, surface alloys, and segregation. The software is also ideal for modeling nanomaterials, since surface and interfacial effects can dominate material behavior and properties at this level. Through the combination of an accurate atomistic surface modeling methodology and an efficient computational engine, it is now possible to directly model these types of surface phenomenon and metallic nanostructures without a supercomputer. Fulfilling a High Operating Temperature Propulsion Components (HOTPC) project level-I milestone, a graphical user interface was created for a suite of quantum approximate atomistic materials modeling Fortran programs developed at Glenn. The resulting "Alloy Design Workbench-Surface Modeling Package" (ADW-SMP) is the combination of proven quantum approximate Bozzolo-Ferrante-Smith (BFS) algorithms (refs. 1 and 2) with a productivity-enhancing graphical front end. Written in the portable, platform independent Java programming language, the graphical user interface calls on extensively tested Fortran programs running in the background for the detailed computational tasks. Designed to run on desktop computers, the package has been deployed on PC, Mac, and SGI computer systems. The graphical user interface integrates two modes of computational materials exploration. One mode uses Monte Carlo simulations to determine lowest energy equilibrium configurations. The second approach is an interactive "what if" comparison of atomic configuration energies, designed to provide real-time insight into the underlying drivers of alloying processes.

Abel, Phillip B.↗

Issues and Methods Concerning the Evaluation of Hypersingular and Near-Hypersingular Integrals in BEM Formulations

It is known that higher order modeling of the sources and the geometry in Boundary Element Modeling (BEM) formulations is essential to highly efficient computational electromagnetics. However, in order to achieve the benefits of hIgher order basis and geometry modeling, the singular and near-singular terms arising in BEM formulations must be integrated accurately. In particular, the accurate integration of near-singular terms, which occur when observation points are near but not on source regions of the scattering object, has been considered one of the remaining limitations on the computational efficiency of integral equation methods. The method of singularity subtraction has been used extensively for the evaluation of singular and near-singular terms. Piecewise integration of the source terms in this manner, while manageable for bases of constant and linear orders, becomes unwieldy and prone to error for bases of higher order. Furthermore, we find that the singularity subtraction method is not conducive to object-oriented programming practices, particularly in the context of multiple operators. To extend the capabilities, accuracy, and maintainability of general-purpose codes, the subtraction method is being replaced in favor of the purely numerical quadrature schemes. These schemes employ singularity cancellation methods in which a change of variables is chosen such that the Jacobian of the transformation cancels the singularity. An example of the sin,oularity cancellation approach is the Duffy method, which has two major drawbacks: 1) In the resulting integrand, it produces an angular variation about the singular point that becomes nearly-singular for observation points close to an edge of the parent element, and 2) it appears not to work well when applied to nearly-singular integrals. Recently, the authors have introduced the transformation u(x(prime))= sinh (exp -1) x(prime)/Square root of ((y prime (exp 2))+ z(exp 2) for integrating functions of the form I = Integral of (lambda(r(prime))((e(exp -jkR))/(4 pi R) d D where A (r (prime)) is a vector or scalar basis function and R = Square root of( (x(prime)(exp2) + (y(prime)(exp2) + z(exp 2)) is the distance between source and observation points. This scheme has all of the advantages of the Duffy method while avoiding the disadvantages listed above. In this presentation we will survey similar approaches for handling singular and near-singular terms for kernels with 1/R(exp 2) type behavior, addressing potential pitfalls and offering techniques to efficiently handle special cases.

Fink, P. W.↗

Experimental and Analytical Characterization of the Macromechanical Response for Triaxial Braided Composite Materials

Increasingly, carbon composite structures are being used in aerospace applications. Their highstrength, high-stiffness, and low-weight properties make them good candidates for replacing many aerospace structures currently made of aluminum or steel. Recently, many of the aircraft engine manufacturers have developed new commercial jet engines that will use composite fan cases. Instead of using traditional composite layup techniques, these new fan cases will use a triaxially braided pattern, which improves case performance. The impact characteristics of composite materials for jet engine fan case applications have been an important research topic because Federal regulations require that an engine case be able to contain a blade and blade fragments during an engine blade-out event. Once the impact characteristics of these triaxial braided materials become known, computer models can be developed to simulate a jet engine blade-out event, thus reducing cost and time in the development of these composite jet engine cases. The two main problems that have arisen in this area of research are that the properties for these materials have not been fully determined and computationally efficient computer models, which incorporate much of the microscale deformation and failure mechanisms, are not available. The research reported herein addresses some of the deficiencies present in previous research regarding these triaxial braided composite materials. The current research develops new techniques to accurately quantify the material properties of the triaxial braided composite materials. New test methods are developed for the polymer resin composite constituent and representative composite coupons. These methods expand previous research by using novel specimen designs along with using a noncontact measuring system that is also capable of identifying and quantifying many of the microscale failure mechanisms present in the materials. Finally, using the data gathered, a new hybrid micromacromechanical computer model is created to simulate the behavior of these composite material systems under static and ballistic impact loading using the test data acquired. The model also quantifies the way in which the fiber/matrix interface affects material response under static and impact loading. The results show that the test methods are capable of accurately quantifying the polymer resin under a variety of strain rates and temperature for three loading conditions. The resin strength and stiffness data show a clear rate and temperature dependence. The data also show the hydrostatic stress effects and hysteresis, all of which can be used by researchers developing composite constitutive models for the resins. The results for the composite data reveal noticeable differences in strength, failure strain, and stiffness in the different material systems presented. The investigations into the microscale failure mechanisms provide information about the nature of the different material system behaviors. Finally, the developed computer model predicts composite static strength and stiffness to within 10 percent of the gathered test data and also agrees with composite impact data, where available.

Littell, Justin D.↗

Space Radiation Transport Code Development: 3DHZETRN

The space radiation transport code, HZETRN, has been used extensively for research, vehicle design optimization, risk analysis, and related applications. One of the simplifying features of the HZETRN transport formalism is the straight-ahead approximation, wherein all particles are assumed to travel along a common axis. This reduces the governing equation to one spatial dimension allowing enormous simplification and highly efficient computational procedures to be implemented. Despite the physical simplifications, the HZETRN code is widely used for space applications and has been found to agree well with fully 3D Monte Carlo simulations in many circumstances. Recent work has focused on the development of 3D transport corrections for neutrons and light ions (Z < 2) for which the straight-ahead approximation is known to be less accurate. Within the development of 3D corrections, well-defined convergence criteria have been considered, allowing approximation errors at each stage in model development to be quantified. The present level of development assumes the neutron cross sections have an isotropic component treated within N explicit angular directions and a forward component represented by the straight-ahead approximation. The N = 1 solution refers to the straight-ahead treatment, while N = 2 represents the bi-directional model in current use for engineering design. The figure below shows neutrons, protons, and alphas for various values of N at locations in an aluminum sphere exposed to a solar particle event (SPE) spectrum. The neutron fluence converges quickly in simple geometry with N > 14 directions. The improved code, 3DHZETRN, transports neutrons, light ions, and heavy ions under space-like boundary conditions through general geometry while maintaining a high degree of computational efficiency. A brief overview of the 3D transport formalism for neutrons and light ions is given, and extensive benchmarking results with the Monte Carlo codes Geant4, FLUKA, and PHITS are provided for a variety of boundary conditions and geometries. Improvements provided by the 3D corrections are made clear in the comparisons. Developments needed to connect 3DHZETRN to vehicle design and optimization studies will be discussed. Future theoretical development will relax the forward plus isotropic interaction assumption to more general angular dependence.

Wilson, John W.↗

The Minos Computing Library: Efficient Parallel Programming for Extremely Heterogeneous Systems

Hardware specialization has become the silver bullet to achieve efficient high performance, from Systems-on-Chip systems, where hardware specialization can be ``extreme'', to large-scale HPC systems. As the complexity of the systems increases, so does the complexity of programming such architectures in a portable way. This work introduces the Minos Computing Library (MCL), as system software, programming model, and programming model runtime that facilitate programming extremely heterogeneous systems. MCL supports the execution of several multi-threaded applications within the same compute node, performs asynchronous execution of application tasks, efficiently balances computation across hardware resources, and provides performance portability. We show that code developed on a personal desktop automatically scales up to fully utilize powerful workstations with 8 GPUs and down to power-efficient embedded systems. MCL provides up to 17.5x speedup over OpenCL on NVIDIA DGX-1 systems and up to 1.88x speedup on single-GPU systems. In multi-application workloads, MCL dynamically resource allocation provides up to 2.43x performance improvement over manual, static allocation of computing resources.

Gioiosa, Roberto↗

Computing the energy of a water molecule using multideterminants: A simple, efficient algorithm

Quantum Monte Carlo (QMC) methods such as variational Monte Carlo and fixed node diffusion Monte Carlo depend heavily on the quality of the trial wave function. Although Slater-Jastrow wave functions are the most commonly used variational ansatz in electronic structure, more sophisticated wave functions are critical to ascertaining new physics. One such wave function is the multi-Slater-Jastrow wave function which consists of a Jastrow function multiplied by the sum of Slater determinants. In this paper we describe a method for working with these wave functions in QMC codes that is easy to implement, efficient both in computational speed as well as memory, and easily parallelized. The computational cost scales quadratically with particle number making this scaling no worse than the single determinant case and linear with the total number of excitations. Additionally, we implement this method and use it to compute the ground state energy of a water molecule.

97 MATHEMATICS AND COMPUTING↗

Three-dimensional transonic nacelle/inlet flowfield computations using an efficient approximate factorization algorithm

A highly efficient computer analysis has been developed for predicting transonic nacelle/inlet flowfields. This algorithm can compute the three-dimensional transonic flowfield about axisymmetric or asymmetric nacelle/inlet configurations at zero or nonzero incidence. The flowfield is determined by solving the full-potential equation in conservative form on a body-fitted curvilinear computational mesh. The difference equations are solved using the AF2 approximate factorization scheme. The effects of boundary layer viscous entrainment are approximated in the inviscid algorithm by applying a surface transpiration velocity which is determined from the calculated boundary layer growth. Computed results and correlations with existing methods and experiment are presented to illustrate application of the analysis.

Vadyak, J.↗

An Efficient and Accurate Algorithm for Computing Grid-Averaged Solar Fluxes for Horizontally Inhomogeneous Clouds

A computationally efficient method is presented to account for the horizontal cloud inhomogeneity by using a radiatively equivalent plane parallel homogeneous (PPH) cloud. The algorithm can accurately match the calculations of the reference (rPPH) independent column approximation (ICA) results, but use only the same computational time required for a single plane parallel computation. The effective optical depth of this synthetic sPPH cloud is derived by exactly matching the direct transmission to that of the inhomogeneous ICA cloud. The ffective9 scattering asymmetry factor is found from a pre-calculated albedo inverse look-up-table that is allowed to vary over the range from -1.0 to 1.0. In the special cases of conservative scattering and total absorption, the synthetic method is exactly equivalent to the ICA, with only a small bias (about 0.2% in flux) relative to ICA due to imperfect interpolation in using the look-up tables. In principle, the ICA albedo can be approximated accurately regardless of cloud inhomogeneity. For a more complete comparison, the broadband shortwave albedo and transmission calculated from the synthetic sPPH cloud and averaged over all incident directions, have the RMS biases of 0.26% and 0.76%, respectively, for inhomogeneous clouds over a wide variation of particle size. The advantages of the synthetic PPH method are that (1) it is not required that all the cloud subcolumns have uniform microphysical characteristic, (2) it is applicable to any 1D radiative transfer scheme, and (3) it can handle arbitrary cloud optical depth distributions and an arbitrary number of cloud subcolumns with uniform computational efficiency.

cloud inhomogeneity↗

Energy Efficiency Scaling for 2 Decades (EES2) Roadmap for Computing

In response to the looming crisis in global energy consumption required for advanced computing applications, the United States Department of Energy (DOE) Advanced Materials and Manufacturing Technology Office (AMMTO) is leading a multi-organizational effort to define a roadmap for energy efficiency scaling for two decades (EES2) with the aim to reduce energy use in all aspects of computation by more than a factor of 1000 in two decades. By July of 2024, over 60 organizations representing industry, academia, and the national laboratories have pledged to work in various aspects of research and development to enable energy efficiency in computing including in the development of the EES2 roadmap, with an initial public release in 2024 as the first phase of an ongoing commitment to energy-efficient and sustainable computation.

Kaarsberg, Tina [U.S. Department of Energy (DOE)]↗