Search NASA⌕ Search

SEARCH · Search NASA

Results for “hybrid algorithm”

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 991 records · Page 55

Electrical Power Grid Delivery Dynamic Analysis: Using Prime Mover Engines to Balance Dynamic Wind Turbine Output

This paper presents an investigation into integrated wind + combustion engine high penetration electrical generation systems. Renewable generation systems are now a reality of electrical transmission. Unfortunately, many of these renewable energy supplies are stochastic and highly dynamic. Conversely, the existing national grid has been designed for steady state operation. The research team has developed an algorithm to investigate the feasibility and relative capability of a reciprocating internal combustion engine to directly integrate with wind generation in a tightly coupled Hybrid Energy System. Utilizing the Idaho National Laboratory developed Phoenix Model Integration Platform, the research team has coupled demand data with wind turbine generation data and the Aspen Custom Modeler reciprocating engine electrical generator model to investigate the capability of reciprocating engine electrical generation to balance stochastic renewable energy.

17 WIND ENERGY↗

MHD SIMULATIONS OF ACCRETION ONTO Sgr A*: QUIESCENT FLUCTUATIONS, OUTBURSTS, AND QUASIPERIODICITY

High-resolution observations of Sgr A* have revealed a wide variety of phenomena, ranging from intense rapid flares to quasi-periodic oscillations (QPOs), making this object an ideal system to study the properties of low luminosity accreting black holes. In this paper, we use a pseudospectral algorithm to construct and evolve a three-dimensional magnetohydrodynamic (MHD) model of the accretion disk in Sgr A*. Assuming a hybrid thermal-nonthermal emission scheme and calibrating the parameters by observations, we show that the MHD turbulence in the environment of Sgr A* can by itself only produce factor two fluctuations in luminosity. These fluctuations cannot explain the magnitude of flares observed in this system. However, we also demonstrate that external forcing of the accretion disk, which may be generated by the 'clumpy material' raining down onto the disk from the large-scale flow, do produce outbursts qualitatively similar to those observed by XMM-Newton in X-rays and by ground-based facilities in the near infrared. Strong, but short-term QPOs emerge naturally in the simulated light curves. We attribute these to nonaxisymmetric density perturbations that emerge as the disk evolves back toward its quiescent state.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Calibration of linear contact stiffnesses in discrete element models using a hybrid analytical-computational framework

Efficient selections of particle-scale contact parameters in discrete element modelling remain an open question. The purpose of this study is to provide a hybrid calibration framework to estimate linear contact stiffnesses (normal and tangential) for both two-dimensional and three-dimensional simulations. Analytical formulas linking macroscopic parameters (Young's modulus, Poisson's ratio) to mesoscopic particle parameters for granular systems are derived based on statistically isotropic packings under small-strain isotropic stress conditions. By taking the derived analytical solutions as initial approximations, the gradient descent algorithm automatically obtains a reliable numerical estimation. The introduced framework is validated with several numerical cases including randomly distributed monodisperse and polydisperse packings. The results show that this hybrid method practically reduces the time for artificial trials and errors to obtain reasonable stiffness parameters. The proposed framework can also be extended to other parameter calibration problems in DEM.

Calibration method↗

A communication-avoiding, hybrid-parallel, rank-revealing orthogonalization method.

Orthogonalization consumes much of the run time of many iterative methods for solving sparse linear systems and eigenvalue problems. Commonly used algorithms, such as variants of Gram-Schmidt or Householder QR, have performance dominated by communication. Here, 'communication' includes both data movement between the CPU and memory, and messages between processors in parallel. Our Tall Skinny QR (TSQR) family of algorithms requires asymptotically fewer messages between processors and data movement between CPU and memory than typical orthogonalization methods, yet achieves the same accuracy as Householder QR factorization. Furthermore, in block orthogonalizations, TSQR is faster and more accurate than existing approaches for orthogonalizing the vectors within each block ('normalization'). TSQR's rank-revealing capability also makes it useful for detecting deflation in block iterative methods, for which existing approaches sacrifice performance, accuracy, or both. We have implemented a version of TSQR that exploits both distributed-memory and shared-memory parallelism, and supports real and complex arithmetic. Our implementation is optimized for the case of orthogonalizing a small number (5-20) of very long vectors. The shared-memory parallel component uses Intel's Threading Building Blocks, though its modular design supports other shared-memory programming models as well, including computation on the GPU. Our implementation achieves speedups of 2 times or more over competing orthogonalizations. It is available now in the development branch of the Trilinos software package, and will be included in the 10.8 release.

99 GENERAL AND MISCELLANEOUS↗

Singular value decomposition utilizing parallel algorithms on graphical processors

One of the current challenges in underwater acoustic array signal processing is the detection of quiet targets in the presence of noise. In order to enable robust detection, one of the key processing steps requires data and replica whitening. This, in turn, involves the eigen-decomposition of the sample spectral matrix, Cx = 1/K xKX(k)XH(k) where X(k) denotes a single frequency snapshot with an element for each element of the array. By employing the singular value decomposition (SVD) method, the eigenvectors and eigenvalues can be determined directly from the data without computing the sample covariance matrix, reducing the computational requirements for a given level of accuracy (van Trees, Optimum Array Processing). (Recall that the SVD of a complex matrix A involves determining V, , and U such that A = U VH where U and V are orthonormal and is a positive, real, diagonal matrix containing the singular values of A. U and V are the eigenvectors of AAH and AHA, respectively, while the singular values are the square roots of the eigenvalues of AAH.) Because it is desirable to be able to compute these quantities in real time, an efficient technique for computing the SVD is vital. In addition, emerging multicore processors like graphical processing units (GPUs) are bringing parallel processing capabilities to an ever increasing number of users. Since the computational tasks involved in array signal processing are well suited for parallelization, it is expected that these computations will be implemented using GPUs as soon as users have the necessary computational tools available to them. Thus, it is important to have an SVD algorithm that is suitable for these processors. This work explores the effectiveness of two different parallel SVD implementations on an NVIDIA Tesla C2050 GPU (14 multiprocessors, 32 cores per multiprocessor, 1.15 GHz clock - peed). The first algorithm is based on a two-step algorithm which bidiagonalizes the matrix using Householder transformations, and then diagonalizes the intermediate bidiagonal matrix through implicit QR shifts. This is similar to that implemented for real matrices by Lahabar and Narayanan ("Singular Value Decomposition on GPU using CUDA", IEEE International Parallel Distributed Processing Symposium 2009). The implementation is done in a hybrid manner, with the bidiagonalization stage done using the GPU while the diagonalization stage is done using the CPU, with the GPU used to update the U and V matrices. The second algorithm is based on a one-sided Jacobi scheme utilizing a sequence of pair-wise column orthogonalizations such that A is replaced by AV until the resulting matrix is sufficiently orthogonal (that is, equal to U ). V is obtained from the sequence of orthogonalizations, while can be found from the square root of the diagonal elements of AH A and, once is known, U can be found from column scaling the resulting matrix. These implementations utilize CUDA Fortran and NVIDIA's CUB LAS library. The primary goal of this study is to quantify the comparative performance of these two techniques against themselves and other standard implementations (for example, MATLAB). Considering that there is significant overhead associated with transferring data to the GPU and with synchronization between the GPU and the host CPU, it is also important to understand when it is worthwhile to use the GPU in terms of the matrix size and number of concurrent SVDs to be calculated.

99 GENERAL AND MISCELLANEOUS↗

Formulation of boundary conditions for the multigrid acceleration of the Euler and Navier Stokes equations

An explicit, Multigrid algorithm was written to solve the Euler and Navier-Stokes equations with special consideration given to the coarse mesh boundary conditions. These are formulated in a manner consistent with the interior solution, utilizing forcing terms to prevent coarse-mesh truncation error from affecting the fine-mesh solution. A 4-Stage Hybrid Runge-Kutta Scheme is used to advance the solution in time, and Multigrid convergence is further enhanced by using local time-stepping and implicit residual smoothing. Details of the algorithm are presented along with a description of Jameson's standard Multigrid method and a new approach to formulating the Multigrid equations.

Jentink, Thomas Neil↗

A new configurational bias scheme for sampling supramolecular structures

We present a new simulation scheme which allows an efficient sampling of reconfigurable supramolecular structures made of polymeric constructs functionalized by reactive binding sites. The algorithm is based on the configurational bias scheme of Siepmann and Frenkel and is powered by the possibility of changing the topology of the supramolecular network by a non-local Monte Carlo algorithm. Such a plan is accomplished by a multi-scale modelling that merges coarse-grained simulations, describing the typical polymer conformations, with experimental results accounting for free energy terms involved in the reactions of the active sites. We test the new algorithm for a system of DNA coated colloids for which we compute the hybridisation free energy cost associated to the binding of tethered single stranded DNAs terminated by short sequences of complementary nucleotides. In order to demonstrate the versatility of our method, we also consider polymers functionalized by receptors that bind a surface decorated by ligands. In particular, we compute the density of states of adsorbed polymers as a function of the number of ligand–receptor complexes formed. Such a quantity can be used to study the conformational properties of adsorbed polymers useful when engineering adsorption with tailored properties. We successfully compare the results with the predictions of a mean field theory. We believe that the proposed method will be a useful tool to investigate supramolecular structures resulting from direct interactions between functionalized polymers for which efficient numerical methodologies of investigation are still lacking.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

The XFaster Power Spectrum and Likelihood Estimator for the Analysis of Cosmic Microwave Background Maps

We present the XFaster analysis package, XFaster is a fast, iterative angular power spectrum estimator based on a diagonal approximation to the quadratic Fisher matrix estimator. XFaster uses Monte Carlo simulations to compute noise biases and filter transfer functions and is thus a hybrid of both Monte Carlo and quadratic estimator methods. In contrast to conventional pseudo-C ℓ based methods, the algorithm described here requires a minimal number of simulations, and does not require them to be precisely representative of the data to estimate accurate covariance matrices for the bandpowers. The formalism works with polarization-sensitive observations and also data sets with identical, partially overlapping, or independent survey regions. The method was first implemented for the analysis of BOOMERanG data (Netterfield et al. 2002; Jones et al. 2006), and also used as part of the Planck analysis (Rocha et al. 2011). Here, we describe the full, publicly available analysis package, written in Python, as developed for the analysis of data from the 2015 flight of the SPIDER instrument (SPIDER Collaboration 2021). The package includes extensions for self-consistently estimating null spectra and for estimating fits for Galactic foreground contributions. We show results from the extensive validation of XFaster using simulations, and its application to the SPIDER data set.

79 ASTRONOMY AND ASTROPHYSICS↗

Antenna pattern control using impedance surfaces

During this research period, we have effectively transferred existing computer codes from CRAY supercomputer to work station based systems. The work station based version of our code preserved the accuracy of the numerical computations while giving a much better turn-around time than the CRAY supercomputer. Such a task relieved us of the heavy dependence of the supercomputer account budget and made codes developed in this research project more feasible for applications. The analysis of pyramidal horns with impedance surfaces was our major focus during this research period. Three different modeling algorithms in analyzing lossy impedance surfaces were investigated and compared with measured data. Through this investigation, we discovered that a hybrid Fourier transform technique, which uses the eigen mode in the stepped waveguide section and the Fourier transformed field distributions across the stepped discontinuities for lossy impedances coating, gives a better accuracy in analyzing lossy coatings. After a further refinement of the present technique, we will perform an accurate radiation pattern synthesis in the coming reporting period.

Balanis, Constantine A.↗

Viscous simulation method for unsteady flows past multicomponent configurations

The present algorithm for the numerical simulation of flows about complex configurations (whose multiple and nonsimilar components have arbitraty geometries) employs a hybridization of the domain decomposition techniques for grid generation as well as to reduce computer-memory requirements. A fully vectorized, finite-volume, upwind-biased, approximately factored multigrid method is used to solve 3D Reynolds-averaged unsteady and compressible Navier-Stokes equations simulating supersonic flows past an ogive-nose-cylinder near or within a cavity. The time-averaged surface pressures obtained compare favorably with the wind tunnel data.

Fouladi, Kamran↗

An Advanced Approach to Simultaneous Monitoring of Multiple Bacteria in Space

The utility of a novel microarray-based microbial analyzer was demonstrated by the rapid detection, imaging, and identification of a mixture of microorganisms found in a waste water sample from the Lunar-Mars Life Support Test Project through the synergistic combination of: (1) judicious RNA probe selection via algorithms developed by University of Houston scientists; (2) tuned surface chemistries developed by Baylor College of Medicine scientists to facilitate hybridization of rRNA targets to DNA probes under very low salt conditions, thereby minimizing secondary structure; and (3) integration of the microarray printing and detection/imaging instrumentation by Genometrix to complete the quantitative analysis of microorganism mixtures.

Eggers, M.↗

Semi-analytic preliminary design of low-thrust missions

Using generalized logarithmic spirals to approximate low-thrust trajectories, a new strategy for the design of low-thrust gravity-assist transfers has been developed. Each transfer leg is defined by a semi-analytic model, and its solution is equivalent to a hybrid Lambert’s problem. The method is suitable for approximating both flyby and rendezvous transfer legs. A branch and prune algorithm is used to generate a collection of initial guesses for further optimization. The analytic nature of the low-thrust model simplifies the pruning step, since dynamical and operational constraints (like maximum thrust or total v) can be imposed easily. The solutions obtained with the global search algorithm can be post-processed, filtered, and ranked according to various criteria. This is where the versatility of the method resides, because changing the selection criteria does not require a new search. Selected candidates are then optimized further, in order to generate actual low-thrust orbits. Two mission design examples are presented: an asteroid deflection mission using a kinetic impactor, and a rendezvous mission to Jupiter. These examples are used to analyze the convergence of the optimization stage, in particular how far from the optimal solution the initial guesses are.

Park, Ryan S.↗

Hybrid Flush and Synthetic Air Data Filter for Entry Vehicle Atmospheric State Estimation

A hybrid flush/synthetic air data sensing filter utilizing Kalman-Schmidt and Rach-Tung-Striebel smoothers is developed to obtain entry vehicle atmosphere estimates. The filter/smoother blends information from pressure sensors distributed on the heatshield with measurements of the vehicle aerodynamic forces and moments computed from mass properties and inertial measurement unit data, and prior estimates of the atmosphere. The filter produces estimates of the atmospheric conditions along the entry trajectory, and systematic error estimates to reconcile differences between the pressure and aerodynamic data sources. The filter is applied to data acquired during the Mars Science Laboratory and Mars 2020 entry, descent, and landing at Gale crater and at Jezero crater, respectively. The results show that the hybrid filter produces estimates of the freestream flight condition with lower uncertainty than either the flush or synthetic air data algorithms. The filter accomplishes this result by incorporating additional data and computing estimates of systematic error parameters in the pressure data and the aerodynamic model to further reduce the uncertainties.

Christopher D. Karlgaard↗

A hybrid-stress finite element approach for stress and vibration analysis in linear anisotropic elasticity

A hybrid stress finite element method is developed for accurate stress and vibration analysis of problems in linear anisotropic elasticity. A modified form of the Hellinger-Reissner principle is formulated for dynamic analysis and an algorithm for the determination of the anisotropic elastic and compliance constants from experimental data is developed. These schemes were implemented in a finite element program for static and dynamic analysis of linear anisotropic two dimensional elasticity problems. Specific numerical examples are considered to verify the accuracy of the hybrid stress approach and compare it with that of the standard displacement method, especially for highly anisotropic materials. It is that the hybrid stress approach gives much better results than the displacement method. Preliminary work on extensions of this method to three dimensional elasticity is discussed, and the stress shape functions necessary for this extension are included.

Oden, J. Tinsley↗

ExaWind: Open‐source CFD for hybrid‐RANS/LES geometry‐resolved wind turbine simulations in atmospheric flows

Abstract Predictive high‐fidelity modeling of wind turbines with computational fluid dynamics, wherein turbine geometry is resolved in an atmospheric boundary layer, is important to understanding complex flow accounting for design strategies and operational phenomena such as blade erosion, pitch‐control, stall/vortex‐induced vibrations, and aftermarket add‐ons. The biggest challenge with high‐fidelity modeling is the realization of numerical algorithms that can capture the relevant physics in detail through effective use of high‐performance computing. For modern supercomputers, that means relying on GPUs for acceleration. In this paper, we present ExaWind, a GPU‐enabled open‐source incompressible‐flow hybrid‐computational fluid dynamics framework, comprising the near‐body unstructured grid solver Nalu‐Wind, and the off‐body block‐structured‐grid solver AMR‐Wind, which are coupled using the Topology Independent Overset Grid Assembler. Turbine simulations employ either a pure Reynolds‐averaged Navier–Stokes turbulence model or hybrid turbulence modeling wherein Reynolds‐averaged Navier–Stokes is used for near‐body flow and large eddy simulation is used for off‐body flow. Being two‐way coupled through overset grids, the two solvers enable simulation of flows across a huge range of length scales, for example, 10 orders of magnitude going from O(μm) boundary layers along the blades to O(10 km) across a wind farm. In this paper, we describe the numerical algorithms for geometry‐resolved turbine simulations in atmospheric boundary layers using ExaWind. We present verification studies using canonical flow problems. Validation studies are presented using megawatt‐scale turbines established in literature. Additionally presented are demonstration simulations of a small wind farm under atmospheric inflow with different stability states.

17 WIND ENERGY↗

SU-F-T-606: Monte Carlo Evaluation of Tissue Inhomogeneity Corrections in the Treatment of Liver Cancer Patients Using Stereotactic Body Radiotherapy

Purpose: To evaluate the dosimetric performance of X-ray Voxel Monte Carlo(XVMC) algorithm in effort to clinically validate Monte Carlo-approach in heterogeneous liver phantom and liver SBRT plans. Methods: An anthropomorphic RPC liver phantom incorporating a liver structure with two cylindrical targets and organs-at-risk (OARs) was used in phantom validation. Subsequently, five patients with metastatic liver cancer were treated using heterogeneity-corrected pencil-beam(PB-hete)algorithm were analyzed following RTOG-1112 criteria .ITV was delineated on MinIP and OARs were contoured on MeanIP images of 4D-CT. PTV was generated from ITV with 5-10mm uniform margin. Mean PTV was 81.3±46.4cc. Prescription was 30–45Gy in 5 fractions, with at least PTV(D95%)=100%. Hybrid SBRT plans were generated with noncoplanar/3D-conformal arcs plus static-beams at Novalis-TX consisting of HD-MLC and 6MV-SRS. SBRT plans were re-computed using XVMC algorithm utilizing identical beam-geometry, MLC-positions, and monitor units and subsequently compared to clinical PB-hete plans. Results: Our results using RPC liver motion phantom validation were all compliance with MD Anderson standards. However, compared to PB-hete, average target volume encompassed by the prescribed percent isodose (Vp) was 9.1% and 8.5% less for PTV1 and PTV2 with XVMC. For the clinical liver SBRT plans, PB-hete systematically overestimated PTV dose (D95, Dmean and D10) within ±2.0% (p<0.05) compared to XVMC. Mean value of Vp was about 3.8% less with XVMC compared to PB-hete (ranged 2.9–5.7% (p<0.003)). However, mean liver dose (MLD) was 3.2% higher (p<0.003), on average, with XVMC compared to clinical PB-hete (ranged −1.0to−3.9%). OARs doses were statistically insignificant. Conclusion: Results from our XVMC dose calculations and validation study for liver SBRT indicate small-to-moderate under-dosing of the tumor volume when compared to PB-hete. Results were consistent with phantom validation and patients plans. However, Vp was less by up to 5.7% for some liver SBRT patients with XVMC–suggesting under-dosing of the target volume and overdosing of MLD by up to 3.9% occurred with PB-hete plan. These differences between PB-hete and XVMC dose calculations may be of clinical interest.

60 APPLIED LIFE SCIENCES↗

FY11 annual Report: PHEV Engine Control and Energy Management Strategy

Objectives are to: (1) Investigate novel engine control strategies targeted at rapid engine/catalyst warming for the purpose of mitigating tailpipe emissions from plug-in hybrid electric vehicles (PHEV) exposed to multiple engine cold start events; and (2) Validate and optimize hybrid supervisory control techniques developed during previous and on-going research projects by integrating them into the vehicle level control system and complementing them with the modified engine control strategies in order to further reduce emissions during both cold start and engine re-starts. Approach used are: (1) Perform a literature search of engine control strategies used in conventional powertrains to reduce cold start emissions; (2) Develop an open source engine controller providing full access to engine control strategies in order to implement new engine/catalyst warm-up behaviors; (3) Modify engine cold start control algorithms and characterize impact on cold start behavior; and (4) Develop an experimental Engine-In-the-Loop test stand in order to validate control methodologies and verify transient thermal behavior and emissions of the real engine when combined with a virtual hybrid powertrain. Some major accomplishments are: (1) Commissioned a prototype engine controller on a GM Ecotec 2.4l direct injected gasoline engine on an engine test cell at the University of Tennessee. (2) Obtained from Bosch (with GM's approval) an open calibration engine controller for a GM Ecotec LNF 2.0l Gasoline Turbocharged Direct Injection engine. Bosch will support the bypass of cold start strategies if calibration access proves insufficient. The LNF engine and its open controller were commissioned on an engine test cell at ORNL. (3) Completed a literature search to identify key engine cold start control parameters and characterized their impact on the real engine using the Bosch engine controller to calibrate them. (4) Ported virtual hybrid vehicle model from offline simulation environment to real-time Hardware-In-the-Loop platform.

33 ADVANCED PROPULSION SYSTEMS↗

Molecular dynamics simulation of effects of solutes on dislocation propagation in Ni-based superalloys

Ni-based superalloys are used in the hot sections of jet turbine engines because of their high strength, stability and resistance to oxidation. The properties of these alloys can be further optimized by adding alloying elements. Therefore, a fundamental knowledge on the effect of different elements on properties of Ni-based superalloys is required. Molecular dynamics simulation could shed light here but its application is hindered by the absence of reliable and computationally cheap semi-empirical potential of the interatomic interaction for 4 and more element alloys. We will present a new Ni-Al-Cr-Nb Finnis-Sinclair (FS) potential specially designed to simulate the dislocation propagation from to  phase. In order to construct this potential, we designed a special algorithm to incorporate the data on element partitioning in the potential development procedure. For example, it is known from experiment, that Cr is mostly present in the gamma phase. Figure 1 shows a snapshot obtained after equilibration of the model of the Ni68Al17Cr15 alloy at T=1000 K using the hybrid Monte-Carlo (MC)/molecular dynamics (MD) simulation with the developed semi-empirical potential. One can clearly see that the Cr partitioning is in agreement with the experimental data. We will discuss the developed algorithm to incorporate the solute partition data in details. Using the developed semi-empirical potential, we first investigated the effect of anti-site defects in the  phase on the single dislocation propagation. It was found that the dislocation velocity increases with the increasing of the anti-site defect concentration. This effect was attributed to smaller number of Al-Al pairs forming during the dislocation migration in the presence of the anti-site defects. Next, we investigated the effect of Nb on the dislocation pair propagation in the Kolbe mechanism. It was found that the addition of Nb leads to considerable decrease in the dislocation propagation rate. This is in agreement with the experimental data on the effect of Nb on the creep resistance of the Ni-based superalloys. We will discuss the origin of this effect.

Mikhail I. Mendelev↗