Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “Computing methodologies”

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 793 records · Page 44

A Nonlocal-Gradient Descent Method for Inverse Design in Nanophotonics

Local-gradient-based optimization approaches lack nonlocal exploration abilityrequired for escaping from local minima when searching non-convex landscapes.A directional Gaussian smoothing (DGS) approach was recently proposed in [29]and used to define a truly nonlocal gradient, referred to as the DGS gradient, inorder to enable nonlocal exploration in high-dimensional black-box optimization.Promising results show that replacing the traditional local gradient with the nonlocalDGS gradient can significantly improve the performance of gradient-based methodsin optimizing highly multi-modal loss functions. However, the current DGS methodis designed for unbounded and uncontrained optimization problems, making itinapplicable to real-world engineering optimization problems where the tuningparameters are often bounded and the loss function is usually constrained byphysical processes. In this work, we propose to extend to the DGS approachto the constrained inverse design framework in order to find better optima ofmulti-modal loss functions. A series of adaptive strategies for smoothing radiusand learning rate updating are developed to improve the computational efficiencyand robustness. Our methodology is demonstrated by an example of designing ananoscale wavelength demultiplexer, and shows superior performance compared tothe state-of-the-art approaches. By incorporating volume constraints, the optimizeddesign achieves an equivalently high performance but significantly reduces theamount of material usage.

Bi, Sirui↗

Optimal Transport for Super Resolution Applied to Astronomy Imaging

Super resolution is an essential tool in optics, especially on interstellar scales, due to physical laws restricting possible imaging resolution. We propose using optimal transport and entropy for super resolution applications. We prove that the reconstruction is accurate when sparsity is known and noise or distortion is small enough. We prove that the optimizer is stable and robust to noise and perturbations. We compare this method to a state of the art convolutional neural network and get similar results for much less computational cost and greater methodological flexibility.

Rawson, Michael G.↗

PARTICLE RESOLVED DNS STUDY OF TURBULENCE EFFECTS ON HYPORHEIC MIXING IN RANDOMLY PACKED SEDIMENT BEDS

Pore-resolved direct numerical simulations (DNS) are used to investigate the interactions between stream-water flow turbulence and groundwater flow through a porous sediment bed in the hyporheic zone. Two permeability Reynolds numbers (2.56 and 5.17), representative of aquatic systems and representing ratio of permeability to viscous length scales, were simulated to understand its influence on the momentum exchange at the sediment-water interface (SWI). A doubleaveraging methodology is used to compute the Reynolds stresses, form-induced stresses, and pressure fluctuations. It is observed that both shear layer and turbulent shear stress penetration increases with ReK. Reynolds and form-induced bed-normal stresses increase with ReK. The peak values of the form-induced stresses for the lower (2.56) and higher (5.17) ReK happen within the top layer of the sediment bed. The sum of turbulent and form-induced pressure fluctuations, analyzed at their respective zero-displacement planes, are statistically similar and can be well approximated by a t location-scale distribution fit providing with a model that could potentially be used to impose boundary conditions at the SWI in reach scale simulations.

Karra, Shashank↗

Force Matching Approaches to Extend Density Functional Theory to Large Time and Length Scales

We present methods for the creation of semi-empirical quantum approaches and reactive force fields through force matching to quantum simulation data for materials under reactive conditions. Our methodologies overcome the extreme computational cost of standard Kohn-Sham Density Functional Theory (DFT) by mapping DFT computed simulation data onto functional forms with linear dependence on their parameters. This allows for quick parameterization of our models by avoiding the non-linear fitting bottlenecks associated with most molecular dynamics model development. We illustrate our approach with two different systems: (i) determination of density functional tight binding models for aqueous glycine dimerization, and (ii) determination of the Chebyshev Interactional Model for Efficient Simulation (ChIMES) reactive force field for metallic liquid carbon. In each case, we observe that our approach is easy to parametrize and yields a model that is orders of magnitude faster than DFT while largely retaining its accuracy. Overall, our methods have potential use for studying complex long time and length scale chemical reactivity at extreme conditions, where there is a significant need for computationally efficient atomistic simulations methods to aid in the interpretation and design of experiments.

Lindsey, R.↗

Accelerating Science Driven System Design With RAMP

Researchers from UC Berkeley, in collaboration with the Lawrence Berkeley National Lab, are engaged in developing an Infrastructure for Synthesis with Integrated Simulation (ISIS). The ISIS Project was a cooperative effort for “application-driven hardware design” that engages application scientists in the early parts of the hardware design process for future generation supercomputing systems. This project served to foster development of computing systems that are better tuned to the application requirements of demanding scientific applications and result in more cost-effective and efficient HPC system designs. In order to overcome long conventional design-cycle times, we leveraged reconfigurable devices to aid in the design of high-efficiency systems, including conventional multi- and many-core systems. The resulting system emulation/prototyping environment, in conjunction with the appropriate intermediate abstractions, provided both a convenient user programming experience and retained flexibility, and thus efficiency, of a reconfigurable platform. We initially targeted the Berkeley RAMP system (Research Accelerator for Multiple Processors) as that hardware emulation environment to facilitate and ultimately accelerate the iterative process of science-driven system design. Our goal was to develop and demonstrate a design methodology for domain-optimized computer system architectures. The tangible outcome is a methodology and tools for rapid prototyping and design-space exploration, leading to highly optimized and efficient HPC systems.

97 MATHEMATICS AND COMPUTING↗

Hierarchical ensemble Kalman methods with sparsity-promoting generalized gamma hyperpriors

This paper introduces a computational framework to incorporate flexible regularization techniques in ensemble Kalman methods, generalizing the iterative alternating scheme to nonlinear inverse problems. The proposed methodology approximates the maximum a posteriori (MAP) estimate of a hierarchical Bayesian model characterized by a conditionally Gaussian prior and generalized gamma hyperpriors. Suitable choices of hyperparameters yield sparsity-promoting regularization. We propose an iterative algorithm for MAP estimation, which alternates between updating the unknown with an ensemble Kalman method and updating the hyperparameters in the regularization to promote sparsity. Here, the effectiveness of our methodology is demonstrated in several computed examples, including compressed sensing and subsurface flow inverse problems.

Ensemble Kalman methods↗

Computational simulation of coupled material degradation processes for probabilistic lifetime strength of aerospace materials

The research included ongoing development of methodology that provides probabilistic lifetime strength of aerospace materials via computational simulation. A probabilistic material strength degradation model, in the form of a randomized multifactor interaction equation, is postulated for strength degradation of structural components of aerospace propulsion systems subjected to a number of effects or primative variables. These primative variable may include high temperature, fatigue or creep. In most cases, strength is reduced as a result of the action of a variable. This multifactor interaction strength degradation equation has been randomized and is included in the computer program, PROMISS. Also included in the research is the development of methodology to calibrate the above described constitutive equation using actual experimental materials data together with linear regression of that data, thereby predicting values for the empirical material constraints for each effect or primative variable. This regression methodology is included in the computer program, PROMISC. Actual experimental materials data were obtained from the open literature for materials typically of interest to those studying aerospace propulsion system components. Material data for Inconel 718 was analyzed using the developed methodology.

Boyce, Lola↗

Bayesian probability of predictive agreement for comparing the outcome of two separate regressions

The concept of a Bayesian probability of agreement was recently introduced to give the posterior probabilities that the response surfaces for two different groups are within δ of one another. For example, a difference of less than δ in the mean response at fixed levels of the predictor variables might be thought to be practically unimportant. In such a case, we would say that the mean responses are in agreement. The posterior probability of this is called the Bayesian probability of agreement. In this article, we quantify the probability that new response observations from two groups will be within δ for a continuous response, and the probability that the two responses agree completely for categorical cases such as logistic regression and Poisson regression. We call these Bayesian comparative predictive probabilities, with the former being the predictive probability of agreement. We use Markov chain Monte Carlo simulation to estimate the posterior distribution of the model parameters and then the predictive probability of agreement. Here, we illustrate the use of this methodology with three examples and provide a freely available R Shiny app that automates the computation and estimation associated with the methodology.

97 MATHEMATICS AND COMPUTING↗

Current Grid Generation Strategies and Future Requirements in Hypersonic Vehicle Design, Analysis and Testing

Recent advances in computational power enable computational fluid dynamic modeling of increasingly complex configurations. A review of grid generation methodologies implemented in support of the computational work performed for the X-38 and X-33 are presented. In strategizing topological constructs and blocking structures factors considered are the geometric configuration, optimal grid size, numerical algorithms, accuracy requirements, physics of the problem at hand, computational expense, and the available computer hardware. Also addressed are grid refinement strategies, the effects of wall spacing, and convergence. The significance of grid is demonstrated through a comparison of computational and experimental results of the aeroheating environment experienced by the X-38 vehicle. Special topics on grid generation strategies are also addressed to model control surface deflections, and material mapping.

Papadopoulos, Periklis↗

A New Finite Element Approach for Prediction of Aerothermal Loads: Progress in Inviscid Flow Computations

Recent progress in the development of finite element methodology for the prediction of aerothermal loads is described. Two dimensional, inviscid computations are presented, but emphasis is placed on development of an approach extendable to three dimensional viscous flows. Research progress is described for: (1) utilization of a commerically available program to construct flow solution domains and display computational results, (2) development of an explicit Taylor-Galerkin solution algorithm, (3) closed form evaluation of finite element matrices, (4) vector computer programming strategies, and (5) validation of solutions. Two test problems of interest to NASA Langley aerothermal research are studied. Comparisons of finite element solutions for Mach 6 flow with other solution methods and experimental data validate fundamental capabilities of the approach for analyzing high speed inviscid compressible flows.

Bey, K. S.↗

A new finite element approach for prediction of aerothermal loads - Progress in inviscid flow computations

Recent progress in the development of finite element methodology for the prediction of aerothermal loads is described. Two dimensional, inviscid computations are presented, but emphasis is placed on development of an approach extendable to three dimensional viscous flows. Research progress is described for: (1) utilization of a commercially available program to construct flow solution domains and display computational results, (2) development of an explicit Taylor-Galerkin solution algorithm, (3) closed form evaluation of finite element matrices, (4) vector computer programming strategies, and (5) validation of solutions. Two test problems of interest to NASA Langley aerothermal research are studied. Comparisons of finite element solutions for Mach 6 flow with other solution methods and experimental data validate fundamental capabilities of the approach for analyzing high speed inviscid compressible flows.

Bey, K. S.↗

Transfinite element methodology towards a unified thermal/structural analysis

The paper describes computational developments towards thermal/structural modeling and analysis via a generalized common numerical methodology for effectively and efficiently interfacing interdisciplinary areas. The proposed formulations use transform methods in conjunction with finite element developments for each of the heat transfer and structural disciplines, respectively, providing avenues for obtaining the structural response due to thermal effects. An alternative methodology for unified thermal/structural analysis is presented. The potential of the approach is outlined in comparison with conventional schemes and existing practices. Highlights and characteristic features of the approach are described via general formulations and applications to several problems. Results obtained demonstrate excellent agreement in comparison with analytic and/or conventional finite element schemes accurately and efficiently.

Tamma, K. K.↗

Approaches to the verification of rule-based expert systems

Expert systems are a highly useful spinoff of artificial intelligence research. One major stumbling block to extended use of expert systems is the lack of well-defined verification and validation (V and V) methodologies. Since expert systems are computer programs, the definitions of verification and validation from conventional software are applicable. The primary difficulty with expert systems is the use of development methodologies which do not support effective V and V. If proper techniques are used to document requirements, V and V of rule-based expert systems is possible, and may be easier than with conventional code. For NASA applications, the flight technique panels used in previous programs should provide an excellent way to verify the rules used in expert systems. There are, however, some inherent differences in expert systems that will affect V and V considerations.

Culbert, Chris↗

PROTEUS-SN User Manual

PROTEUS-SN is a three-dimensional, highly scalable, high-fidelity neutron transport code developed at Argonne National Laboratory. The code is applicable to all spectrum reactor transport calculations, particularly those in which a high degree of fidelity is needed either to represent spatial detail or to resolve solution gradients. PROTEUS-SN solves the second order formulation of the transport equation using the continuous Galerkin finite element method in space, the discrete ordinates approximation in angle, and the multigroup approximation in energy. PROTEUS-SN’s parallel methodology permits the efficient decomposition of the problem by both space and angle, permitting large problems to run efficiently on hundreds of thousands of cores. PROTEUS-SN can also be used in serial or on smaller compute clusters (10’s to 100’s of cores) for smaller homogenized problems, although it is generally more computationally expensive than traditional homogenized methodology codes. PROTEUS-SN has been used to model partially homogenized systems, where regions of interest are represented explicitly and other regions are homogenized to reduce the problem size and required computational resources. PROTEUS-SN solves forward and adjoint eigenvalue problems and permits both neutron upscattering and downscattering. An adiabatic kinetics option has recently been included for performing simple time-dependent calculations in addition to standard steady state calculations. PROTEUS-SN handles void and reflective boundary conditions. Multigroup cross sections can be generated externally using the MC2-3 fast reactor multigroup cross section generation code or internally using the cross section application programming interface (API) which can treat the subgroup or resonance table libraries. PROTEUS-SN is written in Fortran 90 and also includes C preprocessor definitions. The code links against the PETSc, METIS, HDF5, and MPICH libraries. It optionally links against the MOAB library and is a part of the SHARP multi-physics suite for coupled multi-physics analysis of nuclear reactors. This user manual describes how to set up a neutron transport simulation with the PROTEUS-SN code. A companion methodology manual describes the theory and algorithms within PROTEUS-SN.

97 MATHEMATICS AND COMPUTING↗

Top-down methodology for human factors research

User computer interaction as a conversation is discussed. The design of user interfaces which depends on viewing communications between a user and the computer as a conversion is presented. This conversation includes inputs to the computer (outputs from the user), outputs from the computer (inputs to the user), and the sequencing in both time and space of those outputs and inputs. The conversation is viewed from the user's side of the conversation. Two languages are modeled: the one with which the user communicates with the computer and the language where communication flows from the computer to the user. Both languages exist on three levels; the semantic, syntactic and lexical. It is suggested that natural languages can also be considered in these terms.

Sibert, J.↗

Report of the LSPI/NASA Workshop on Lunar Base Methodology Development

Groundwork was laid for computer models which will assist in the design of a manned lunar base. The models, herein described, will provide the following functions for the successful conclusion of that task: strategic planning; sensitivity analyses; impact analyses; and documentation. Topics addressed include: upper level model description; interrelationship matrix; user community; model features; model descriptions; system implementation; model management; and plans for future action.

Nozette, Stewart↗

A comparison of the various helicopter mathematical models used in the methodology assessment

Various features of the computer codes used in the helicopter industry and by government agencies for rotorcraft aeroelastic stability analysis are compared. Mathematical rigor in modeling rotorcraft is given primarily to the rotor-system dynamic behavior; the aerodynamic modeling is still limited to strip theory and to uneven application of corrections for stall, reversed flow, yawed flow, radial flow, and unsteady aerodynamic effects. The forward-flight regime analysis is included in five of the 11 codes surveyed. However, only two of these codes are capable of a Floquet analysis for aeroelastic stability. For the hover regime, nine of the 11 codes use eigen-analysis approach. The remaining codes perform a harmonic analysis of the transient response of system.

Stephens, Wendell B.↗

Computational Analysis and Optimized Modeling of Geomagnetically Induced Currents in Power Transformers

In this project we aim to better understand the effect of geomagnetically induced currents (GIC) on power transformers. Expanding upon our previous work focused on producing a methodology for accuracy-enhanced computation of GIC signatures (i.e., time-domain current magnitude variation for the event duration) from a combination of physics-based and data-driven computational tools, we propose the use of these GIC signatures as inputs for a physically-detailed and optimized model of the power transformer to investigate how GIC determination and transformer modeling influence the evaluation of GIC effects on the transformer operation, as well as in its interaction with the power grid.

24 POWER TRANSMISSION AND DISTRIBUTION↗