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 1,081 records · Page 60

Comparative study of the convergence rates of two numerical techniques

The paper examines the applicability of the three-step Stetter (1968) method to the problem of hypersonic viscous flow over a blunt axisymmetric body used for planetary entry probes at zero angle of attack. The flow-field results using the two-step finite-difference MacCormack (1969) method are reported by Kumar and Graves (1977). Only the computational efficiency of Stetter's method is compared with that of MacCormack's in terms of the iterative time steps and computing time required for the steady-state solution. Advantages of Stetter's method over MacCormack's are established.

Kumar, A.↗

Computation of wing-vortex interaction in transonic flow using implicit finite difference algorithm

An implicit delta form finite difference algorithm for Euler equations in conservation law form was used in preliminary calculations of three dimensional wing vortex interaction. Both steady and unsteady transonic flow wing vortex interactions are computed. The computations themselves are meant to guide upcoming wind tunnel experiments of the same flow field. Various modifications to the numerical method that are intended to improve computational efficiency are also described and tested in both two and three dimensions. Combination of these methods can reduce the overall computational time by a factor of 4.

Srinivasan, G.↗

Comparison of methods for developing the dynamics of rigid-body systems

Several approaches for developing the equations of motion for a three-degree-of-freedom PUMA robot were compared on the basis of computational efficiency (i.e., the number of additions, subtractions, multiplications, and divisions). Of particular interest was the investigation of the use of computer algebra as a tool for developing the equations of motion. Three approaches were implemented algebraically: Lagrange's method, Kane's method, and Wittenburg's method. Each formulation was developed in absolute and relative coordinates. These six cases were compared to each other and to a recursive numerical formulation. The results showed that all of the formulations implemented algebraically required fewer calculations than the recursive numerical algorithm. The algebraic formulations required fewer calculations in absolute coordinates than in relative coordinates. Each of the algebraic formulations could be simplified, using patterns from Kane's method, to yield the same number of calculations in a given coordinate system.

Ju, M. S.↗

A viscous-inviscid interaction method for 2-D unsteady, compressible flows

A Navier-Stokes/potentia/flow interactive solution method suitable for the solution of steady-state and unsteady flowfields around airfoils has been developed. The Navier-Stokes equations are solved in the close proximity of an airfoil and in its wake. The inviscid flowfield surrounding the viscous flow regions is assumed to be irrotational and isentropic. In the inviscid flow region, the potential flow equations, which are based on distributed sources and vortices, are solved. The two solutions are strongly coupled through the application of their boundary conditions. The strongly coupled Navier-Stokes/potential flow interactive solution method confines computations to a small domain in the proximity of an airfoil. For steady flow solutions, the computational domain may be confined to a region which extends less than one-fifth of a chord length distance away from the airfoil surface. Computed solutions have the same accuracy as the full domain NavierStokes solutions. Yet, as a result of the significantly reduced computational domain and increased convergence rates, the Navier-Stokes/potential flow interactive solution method is about 40 percent more efficient computationally.

Tuncer, Ismail H.↗

Comparison of Coupled Radiative Flow Solutions with Project Fire 2 Flight Data

A nonequilibrium, axisymmetric, Navier-Stokes flow solver with coupled radiation has been developed for use in the design or thermal protection systems for vehicles where radiation effects are important. The present method has been compared with an existing now and radiation solver and with the Project Fire 2 experimental data. Good agreement has been obtained over the entire Fire 2 trajectory with the experimentally determined values of the stagnation radiation intensity in the 0.2-6.2 eV range and with the total stagnation heating. The effects of a number of flow models are examined to determine which combination of physical models produces the best agreement with the experimental data. These models include radiation coupling, multitemperature thermal models, and finite rate chemistry. Finally, the computational efficiency of the present model is evaluated. The radiation properties model developed for this study is shown to offer significant computational savings compared to existing codes.

Olynick, David R.↗

Comparison of Coupled Radiative Navier-Stokes Flow Solutions with the Project Fire II Flight Data

A nonequilibrium, axisymmetric, Navier-Stokes flow solver with coupled radiation has been developed to use in the design of thermal protection systems for vehicles where radiation effects are important. The present method has been compared with an existing flow and radiation solver and with the Project Fire II experimental data. Very good agreement has been obtained over the entire Fire II trajectory with the experimentally determined values of the stagnation radiation intensity in the .2 to 6.2 eV range and with the total stagnation heating. The agreement was significantly better than previous numerical predictions. The effects of a number of flow models are examined to determine which combination of physical models produces the best agreement with the experimental data. These models include radiation coupling, multi-temperature thermal models, finite-rate chemistry, and a quasi-steady-state or Boltzmann assumption for the calculation of the excited electronic states. Finally, the computational efficiency of the present model is evaluated. The radiation properties model developed for this study is shown to offer significant computational savings compared to existing codes.

Olynick, David R.↗

An Exact Dual Adjoint Solution Method for Turbulent Flows on Unstructured Grids

An algorithm for solving the discrete adjoint system based on an unstructured-grid discretization of the Navier-Stokes equations is presented. The method is constructed such that an adjoint solution exactly dual to a direct differentiation approach is recovered at each time step, yielding a convergence rate which is asymptotically equivalent to that of the primal system. The new approach is implemented within a three-dimensional unstructured-grid framework and results are presented for inviscid, laminar, and turbulent flows. Improvements to the baseline solution algorithm, such as line-implicit relaxation and a tight coupling of the turbulence model, are also presented. By storing nearest-neighbor terms in the residual computation, the dual scheme is computationally efficient, while requiring twice the memory of the flow solution. The scheme is expected to have a broad impact on computational problems related to design optimization as well as error estimation and grid adaptation efforts.

Nielsen, Eric J.↗

Dynamic Programming for Structured Continuous Markov Decision Problems

We describe an approach for exploiting structure in Markov Decision Processes with continuous state variables. At each step of the dynamic programming, the state space is dynamically partitioned into regions where the value function is the same throughout the region. We first describe the algorithm for piecewise constant representations. We then extend it to piecewise linear representations, using techniques from POMDPs to represent and reason about linear surfaces efficiently. We show that for complex, structured problems, our approach exploits the natural structure so that optimal solutions can be computed efficiently.

Dearden, Richard↗

hydro-hifem v.1.0

SAND2021-3180 O The hydro-hifem 1.0 software is used to simulate transient physical processes in large geologic environments consisting of complex fractures and man-made infrastructures with affordable computational cost and mesh refinement. The Earth models with such complex model features can be designed by considering the reduction of insignificant dimensions of fractures/infrastructures and embracing their corresponding hierarchical material properties, and their computational meshes are generated by using Cubit. The finite element solutions are automatically distributed across nodes of parallel cluster for efficient computation of transient fluid flow or heat conduction in geologic media. Sandia National Laboratories is a multimission laboratory managed and operated by National Technology & Engineering Solutions of Sandia, LLC, a wholly owned subsidiary of Honeywell International Inc., for the U.S. Department of Energy’s National Nuclear Security Administration under contract DE-NA0003525.

Beskardes, GungorDidem↗

Strong Lensing Cosmology with Population-level Calibrated Neural Ratio Estimation

Strong gravitational lensing contains key information about cosmic acceleration. Modern and next-generation galaxy imaging surveys are expected to provide high-quality data on $\mathcal{O}(10^5)$ galaxy-galaxy lensing systems. The plethora and complexity of the data are likely to present computational challenges for parameter inference methods for fitting high-dimensional likelihoods, which are often analytically intractable. Neural Ratio Estimation (NRE) efficiently computes individual likelihood ratios that can be combined into population-level posteriors. We use simulations to study the capacity of NRE to jointly predict the dark energy equation-of-state parameter $w$ and the total matter density $Ω_{m}$ from lensing images and companion spectroscopic information. We also introduce a post hoc posterior coverage calibration procedure that mitigates the model overconfidence that is typically found in neural density estimation applications. Our experiments show that the errors on both parameters decrease with increasing inference population sizes. In particular, for 100 lenses in a standard $Λ$CDM Universe, our calibrated NRE model achieves median fractional uncertainty of $22.8\%$ in $w$ and $2.9\%$ in $Ω_{m}$. This proof of concept demonstrates a potentially scalable approach for efficient cosmological parameter inference with large populations of galaxy-scale lenses observed in future surveys.

Jarugula, Sreevani [Fermilab] (ORCID:0000000253867↗

ADROC: An Emulation Experimentation Platform for Advancing Resilience of Control Systems

Cyberattacks against industrial control systems have increased over the last decade, making it more critical than ever for system owners to have the tools necessary to understand the cyber resilience of their systems. However, existing tools are often qualitative, subject matter expertise-driven, or highly generic, making thorough, data-driven cyber resilience analysis challenging. The ADROC project proposed to develop a platform to enable efficient, repeatable, data-driven cyber resilience analysis for cyber-physical systems. The approach consists of two phases of modeling: computationally efficient math modeling and high-fidelity emulations. The first phase allows for scenarios of low concern to be quickly filtered out, conserving resources available for analysis. The second phase supports more detailed scenario analysis, which is more predictive of real-world systems. Data extracted from experiments is used to calculate cyber resilience metrics. ADROC then ranks scenarios based on these metrics, enabling prioritization of system resources to improve cyber resilience.

97 MATHEMATICS AND COMPUTING↗

Implicit approximate-factorization schemes for the efficient solution of steady transonic flow problems

Implicit approximate-factorization algorithms (AF) are developed for the solution of steady-state transonic flow problems. The performance of the AF solution method is evaluated relative to that of the standard solution method for transonic flow problems, successive line over-relaxation (SLOR). Both methods are applied to the solution of the nonlinear, two-dimensional transonic small-disturbance equation. Results indicate that the AF method requires substantially less computer time than SLOR to solve the nonlinear finite-difference matrix equation for a transonic flow field. This increase in computational efficiency is achieved with no appreciable increase in computer storage or coding complexity.

Ballhaus, W. F.↗

Numerical studies of dilute and dense spray characteristics

Several issues involving the improvement of physical submodels and the computational efficiency in modeling dilute and dense spray combustion are discussed. First, the implementations of a dispersion width approach accounting for turbulent dispersion within each computational parcel is discussed. This is essentially a statistical transport model and the testings of this model confirm the capability of accurately representing dispersion in nearly-homogeneous and inhomogeneous turbulent flows with improved efficiency over the delta function stochastic separated flow model. To account for the dense spray effects, an existing drop collision and coalescence model and a Taylor analogy breakup (TAB) model were employed. These models were incorporated into a state-of-the-art multiphase all-speed transient flow solution procedure. Several examples including nonevaporating, evaporating, and burning dense spray cases were studied. The numerical results show reasonably good comparisons with available experimental data in terms of spray penetration, drop sizes, and overall configuration of a spray flame.

Kim, Y. M.↗

A high-order multiscale finite-element method for time-domain elastic wave modeling in strongly heterogeneous media

Efficient and accurate numerical methods for elastic wave modeling in complex media have many important applications. However, it is fairly challenging to model elastic wave propagation in strongly heterogeneous media with high computational efficiency and high-order accuracy simultaneously. We develop a novel high-order multiscale finite-element method to model elastic wave propagation in strongly heterogeneous media in the time domain. The most important feature of our method is a generalization of standard multiscale finite element method by using high-order multiscale finite-element basis functions to capture the fine-scale heterogeneities on the coarse mesh, in contrast to conventional finite-element basis functions that are merely determined by the order of polynomials. These multiscale basis functions leads to a system matrix with significantly reduced dimension, thus enable us to solve the elastic wave equation on the coarse mesh with high-order accuracy and very low computational time cost. In conclusion, we use 2D and 3D numerical examples to demonstrate the superior efficiency and accuracy of our new modeling method compared with the conventional spectral-element method.

58 GEOSCIENCES↗

Approximate boundary condition procedure for the two-dimensional numerical solution of vortex wakes

Research on efficient computational methods for general vorticity fields has been conducted in connection with a need for basic research on vortex-dominated flows. The present investigation is concerned with the evolution of vortex wakes behind aircraft wings. An efficient procedure is presented for the calculation of the boundary values used in the numerical solution of the unsteady, incompressible, two-dimensional Navier-Stokes equations for an unbounded flow field. The extent of the computational grid can be reduced compared to methods utilizing standard boundary conditions, without loss of accuracy. The efficiencies realized make it feasible to calculate the vortex wake development for realistic wing configurations, including the merging of multiple vortices, for Reynolds numbers of about 10,000 based on wing chord.

Weston, R. P.↗

Enhanced accuracy through ensembling of randomly initialized auto-regressive models for dynamical systems

Computational mechanics simulations using traditional finite element methods (FEM) require prohibitively expensive computational resources for real-time engineering applications, design optimization, and digital twin implementations. While machine learning (ML) surrogate models offer significant computational speedups, autoregressive ML models for time-dependent mechanical systems suffer from error accumulation that compromises long-term prediction reliability - a critical concern for engineering applications where accuracy over extended time horizons is essential for safety and performance assessments. Here, we propose a deep ensemble framework specifically designed to address this challenge in computational mechanics applications, where multiple ML surrogate models with random weight initializations are trained in parallel and their predictions aggregated during inference. This approach leverages statistical diversity to maximize information gain from a fixed set of training data and to mitigate error propagation, while maintaining the computational efficiency that makes ML surrogates attractive for engineering practice. We validate the framework on three representative problems spanning critical areas of computational mechanics: stress field evolution in heterogeneous microstructures under complex loading (relevant to advanced materials design and composite analysis), planetary-scale shallow water dynamics (applicable to environmental and geotechnical engineering), and Gray-Scott reaction-diffusion systems (relevant to mass transport and chemical process engineering). Across all test cases, the ensemble approach demonstrates consistent error reduction of 15-33% compared to individual models. The codes for this work are available on GitHub (https://github.com/Graham-Brady-Research-Group/AutoregressiveEnsemble_SpatioTemporal_Evolution).

autoregressive prediction↗

Groundwater flow and heat transport for systems undergoing freeze-thaw: Intercomparison of numerical simulators for 2D test cases

In high-elevation, boreal and arctic regions, hydrological processes and associated water bodies can be strongly influenced by the distribution of permafrost. Recent field and modeling studies indicate that a fully-coupled multidimensional thermo-hydraulic approach is required to accurately model the evolution of these permafrost-impacted landscapes and groundwater systems. However, the relatively new and complex numerical codes being developed for coupled non-linear freeze-thaw systems require verification. Here in this paper, this issue is addressed by means of an intercomparison of thirteen numerical codes for two-dimensional test cases with several performance metrics (PMs). These codes comprise a wide range of numerical approaches, spatial and temporal discretization strategies, and computational efficiencies. Results suggest that the codes provide robust results for the test cases considered and that minor discrepancies are explained by computational precision. However, larger discrepancies are observed for some PMs resulting from differences in the governing equations, discretization issues, or in the freezing curve used by some codes.

54 ENVIRONMENTAL SCIENCES↗

Pairwise-interaction extended point-particle model for particle-laden flows

In this work we consider the pairwise interaction extended point-particle (PIEP) model for Euler–Lagrange simulations of particle-laden flows. By accounting for the precise location of neighbors the PIEP model goes beyond local particle volume fraction, and distinguishes the influence of upstream, downstream and laterally located neighbors. The two main ingredients of the PIEP model are (i) the undisturbed flow at any particle is evaluated as a superposition of the macroscale flow and a microscale flow that is approximated as a pairwise superposition of perturbation fields induced by each of the neighboring particles, and (ii) the forces and torque on the particle are then calculated from the undisturbed flow using the Faxén form of the force relation. The computational efficiency of the standard Euler–Lagrange approach is retained, since the microscale perturbation fields induced by a neighbor are pre-computed and stored as PIEP maps. Here we extend the PIEP force model of Akiki et al. with a corresponding torque model to systematically include the effect of perturbation fields induced by the neighbors in evaluating the net torque. Also, we use DNS results from a uniform flow over two stationary spheres to further improve the PIEP force and torque models. We then test the PIEP model in three different sedimentation problems and compare the results against corresponding DNS to assess the accuracy of the PIEP model and improvement over the standard point-particle approach. In the case of two sedimenting spheres in a quiescent ambient the PIEP model is shown to capture the drafting-kissing-tumbling process. In cases of 5 and 80 sedimenting spheres a good agreement is obtained between the PIEP simulation and the DNS. For all three simulations, the DEM-PIEP was able to recreate, to a good extent, the results from the DNS, while requiring only a negligible fraction of the numerical resources required by the fully-resolved DNS.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗