SEARCH · Search NASA

Results for “cluster methods”

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.

SourceResearch connectionFrom yearThrough year

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 1,639 records · Page 91

All sources NASA NTRS DOE OSTI

DOE OSTI2015

Optimizing deep learning hyper-parameters through an evolutionary algorithm. In: MLHPC '15 Proceedings of the Workshop on Machine Learning in High-Performance Computing Environments, Article No. 4

There has been a recent surge of success in utilizing Deep Learning (DL) in imaging and speech applications for its relatively automatic feature generation and, in particular for convolutional neural networks (CNNs), high accuracy classification abilities. While these models learn their parameters through data-driven methods, model selection (as architecture construction) through hyper-parameter choices remains a tedious and highly intuition driven task. To address this, Multi-node Evolutionary Neural Networks for Deep Learning (MENNDL) is proposed as a method for automating network selection on computational clusters through hyper-parameter optimization performed via genetic algorithms.

Young, Steven R.↗

DOE OSTI2023

The Generalized Green’s function Cluster Expansion: A Python package for simulating polarons

We present an efficient implementation of the Generalized Green’s function Cluster Expansion (GGCE), which is a new method for computing the ground-state properties and dynamics of polarons (single electrons coupled to lattice vibrations) in model electron-phonon systems. The GGCE works at arbitrary temperature and is well suited for a variety of electron-phonon couplings, including, but not limited to, site and bond Holstein and Peierls (Su-SchriefferHeeger) couplings, and couplings to multiple phonon modes with different energy scales and coupling strengths. Quick calculations can be performed efficiently on a laptop using solvers from NumPy and SciPy, or in parallel at scale using the PETSc sparse linear solver engine.

36 MATERIALS SCIENCE↗

DOE OSTI2022

Electronic and Geometric effects on Photochemistry of Molecules in Well-Defined Environments

Assemblies of molecular switches at surfaces are of great interest for applications in solar energy conversion, optoelectronic and optomechanical devices. Experimental studies have demonstrated that it is possible to optically probe photoactive molecules in well-defined nanoscale environments. To understand these systems, we have developed a polarizable frozen-density embedding (FDEpol) to model the optical properties of molecules adsorbed on metal clusters. We have demonstrated that the FDEpol method can describe the ground state and linear response properties of strongly coupled systems with results that are in good agreement with exact projection based embedding method.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

DOE OSTI2022

Integrating Concentrating Solar Power Technologies into the Hybrid Optimization and Performance Platform (HOPP)

As the world increases renewable energy deployment, there is an increasing interest in hybridizing various generation and storage technologies to maximize net benefit to the developer and/or off-taker. A particularly interesting combination of renewable technologies is concentrating solar power (CSP) with thermal energy storage (TES), photovoltaics (PV), and electrochemical battery energy storage (BESS). Due to the system complexity of CSP technology, it is difficult to evaluate the technological and financial performance of a CSP-PV hybrid system without detailed modeling of annual operations. To address this challenge, we have developed a modeling framework for evaluating the performance and financial viability of CSP systems hybridized with PV and battery technologies. This modeling effort incorporates CSP tower and trough systems into an existing modeling tool recently developed by NREL referred to as the Hybrid Optimization and Performance Platform (HOPP). This report outlines the modeling methodology as well as preliminary results from example case studies conducted using the model. The methodology describes: (i) the integration of CSP tower and troughs into HOPP using python interfaces to access System Advisor Model (SAM) underlining technology models, (ii) the mathematical formulation of the mixed integer linear program dispatch optimization model which optimizes operations of storage asset to either maximize system revenue or minimize operating cost while load following, (iii) the design analysis methods implemented within HOPP, and (iv) simulation clustering for the purposes of reducing computational expense. We exercise the model using a case study of a future scenario where we assume (i) CSP and PV technologies achieve the 2030 cost targets provided by the Solar Energy Technologies Office (SETO), (ii) battery costs reduce to the 2030 mid cost projection presented by NREL. Lastly, (iii) electricity prices for southern California in 2030 are provided by NREL's Cambium database, and (iv) a capacity payment of $150/kW-yr based on the system capacity factor during the to 100 net-load hours.

14 SOLAR ENERGY↗

NASA NTRS1983

An interactive Barnes objective map analysis scheme for use with satellite and conventional data

The Barnes (1973) objective map analysis scheme is employed to develop an interactive analysis package for assessing the impact of satellite-derived data on analyses of conventional meteorological data sets. The method permits modification of the values of input parameters in the objective analysis within objectively determined, internally set limits. The effects of the manipulations are rapidly displayed, and methods are included for assimilating the spatially clustered characteristics of satellite data and the various horizontal resolutions of the data types. Data sets from the SESAME rawinsonde wind data with uniform spatial distribution, with the same data set plus satellite cloud motion data, and a data set from the atmospheric sounder radiometer on the GOES satellite were analyzed as examples. The scheme is demonstrated to recover details after two iterations through the data.

Koch, S. E.↗

NASA NTRS1996

Self-Consistent Simulation of the Brownian Stage of Dust Growth

It is a widely accepted view that in proto-planetary accretion disks the collision and following sticking of dust particles embedded in the gas eventually leads to the formation of planetesimals (coagulation). For the smallest dust grains, Brownian motion is assumed to be the dominant source of their relative velocities leading to collisions between these dust grains. As the dust grains grow they eventually couple to the turbulent motion of the gas which then drives the coagulation much more efficiently. Many numerical coagulation simulations have been carried out to calculate the fractal dimension of the aggregates, which determines the duration of the ineffective Brownian stage of growth. Predominantly on-lattice and off-lattice methods were used. However, both methods require simplification of the astrophysical conditions. The aggregates found by those methods had a fractal dimension of approximately 2 which is equivalent to a constant, mass-independent friction time. If this value were valid for the conditions in an accretion disk, this would mean that the coagulation process would finally 'freeze out' and the growth of a planetesimal would be impossible within the lifetime of an accretion disk. In order to investigate whether this fractal dimension is model independent, we simulate self-consistently the Brownian stage of the coagulation by an N-particle code. This method has the advantage that no further assumptions about homogeneity of the dust have to be made. In our model, the dust grains are considered as aggregates built up of spheres. The equation of motion of the dust grains is based on the probability density for the diffusive transport within the gas atmosphere. Because of the very low number density of the dust grains, only 2-body-collisions have to be considered. As the Brownian stage of growth is very inefficient, the system is to be simulated over long periods of time. In order to find close particle pairs of the system which are most likely to undergo a collision, we use a particle-in-cell (PIC) method for the early stages of the simulation where the system is still very homogeneous and a tree method later when the particles are more clustered.

Kempf, S.↗

NASA NTRS1993

Statistics of Experiments on Cluster Formation and Transport in a Gravitational Field

Metastable state relaxation in a gravitational field is investigated in the case of non-critical binary solutions. A relaxation description is presented in terms of the time-dependent Ginzburg-Landau formalism for a non-conserved order parameter. A new ansatz for solution of the corresponding partial nonlinear stochastic differential equation is discussed. It is proved that, for the supersaturated solution under consideration, the metastable state relaxation in a gravitational field leads to formation of solute concentration gradients due to the sedimentation of subcritical solute clusters. The pure discussion of the possible methods to compare theoretical results and experimental data related to solute sedimentation in a gravitational field is presented. It is shown that in order to describe these experiments it is necessary to deal both with the value of the solute concentration gradient and with its formation rate. The stochastic nature of the sedimentation process is shown.

Izmailov, Alexander F.↗

NASA NTRS2012

3D Drop Size Distribution Extrapolation Algorithm Using a Single Disdrometer

Determining the Z-R relationship (where Z is the radar reflectivity factor and R is rainfall rate) from disdrometer data has been and is a common goal of cloud physicists and radar meteorology researchers. The usefulness of this quantity has traditionally been limited since radar represents a volume measurement, while a disdrometer corresponds to a point measurement. To solve that problem, a 3D-DSD (drop-size distribution) method of determining an equivalent 3D Z-R was developed at the University of Central Florida and tested at the Kennedy Space Center, FL. Unfortunately, that method required a minimum of three disdrometers clustered together within a microscale network (.1-km separation). Since most commercial disdrometers used by the radar meteorology/cloud physics community are high-cost instruments, three disdrometers located within a microscale area is generally not a practical strategy due to the limitations of these kinds of research budgets. A relatively simple modification to the 3D-DSD algorithm provides an estimate of the 3D-DSD and therefore, a 3D Z-R measurement using a single disdrometer. The basis of the horizontal extrapolation is mass conservation of a drop size increment, employing the mass conservation equation. For vertical extrapolation, convolution of a drop size increment using raindrop terminal velocity is used. Together, these two independent extrapolation techniques provide a complete 3DDSD estimate in a volume around and above a single disdrometer. The estimation error is lowest along a vertical plane intersecting the disdrometer position in the direction of wind advection. This work demonstrates that multiple sensors are not required for successful implementation of the 3D interpolation/extrapolation algorithm. This is a great benefit since it is seldom that multiple sensors in the required spatial arrangement are available for this type of analysis. The original software (developed at the University of Central Florida, 1998.- 2000) has also been modified to read standardized disdrometer data format (Joss-Waldvogel format). Other modifications to the software involve accounting for vertical ambient wind motion, as well as evaporation of the raindrop during its flight time.

Lane, John↗

NASA NTRS2020

The Atacama Cosmology Telescope: arcminute-resolution maps of 18,000 square degrees of the microwave sky from ACT 2008-2018 data combined with Planck

This paper presents a maximum-likelihood algorithm for combining sky maps with disparate sky coverage, angular resolution and spatially varying anisotropic noise into a single map of the sky. We use this to merge hundreds of individual maps covering the 2008–2018 ACT observing seasons, resulting in by far the deepest ACT maps released so far. We also combine the maps with the full Planck maps, resulting in maps that have the best features of both Planck and ACT: Planck’s nearly white noise on intermediate and large angular scales and ACT’s high-resolution and sensitivity on small angular scales. The maps cover over 18 000 square degrees, nearly half the full sky, at 100, 150 and 220 GHz. They reveal 4 000 optically-confirmed clusters through the Sunyaev Zel’dovich effect (SZ) and 18 500 point source candidates at > 5σ, the largest single collection of SZ clusters and millimeter wave sources to date. The multi-frequency maps provide millimeter images of nearby galaxies and individual Milky Way nebulae, and even clear detections of several nearby stars. Other anticipated uses of these maps include, for example, thermal SZ and kinematic SZ cluster stacking, CMB cluster lensing and galactic dust science. The method itself has negligible bias. However, due to the preliminary nature of some of the component data sets, we caution that these maps should not be used for precision cosmological analysis. The maps are part of ACT DR5, and will be made available on LAMBDA no later than three months after the journal publication of this article, along with an interactive sky atlas.

Sigurd Naess↗

NASA NTRS1995

Enhancements to the SSME transfer function modeling code

This report details the results of a one year effort by Ohio University to apply the transfer function modeling and analysis tools developed under NASA Grant NAG8-167 (Irwin, 1992), (Bartholomew, 1992) to attempt the generation of Space Shuttle Main Engine High Pressure Turbopump transfer functions from time domain data. In addition, new enhancements to the transfer function modeling codes which enhance the code functionality are presented, along with some ideas for improved modeling methods and future work. Section 2 contains a review of the analytical background used to generate transfer functions with the SSME transfer function modeling software. Section 2.1 presents the 'ratio method' developed for obtaining models of systems that are subject to single unmeasured excitation sources and have two or more measured output signals. Since most of the models developed during the investigation use the Eigensystem Realization Algorithm (ERA) for model generation, Section 2.2 presents an introduction of ERA, and Section 2.3 describes how it can be used to model spectral quantities. Section 2.4 details the Residue Identification Algorithm (RID) including the use of Constrained Least Squares (CLS) and Total Least Squares (TLS). Most of this information can be found in the report (and is repeated for convenience). Section 3 chronicles the effort of applying the SSME transfer function modeling codes to the a51p394.dat and a51p1294.dat time data files to generate transfer functions from the unmeasured input to the 129.4 degree sensor output. Included are transfer function modeling attempts using five methods. The first method is a direct application of the SSME codes to the data files and the second method uses the underlying trends in the spectral density estimates to form transfer function models with less clustering of poles and zeros than the models obtained by the direct method. In the third approach, the time data is low pass filtered prior to the modeling process in an effort to filter out high frequency characteristics. The fourth method removes the presumed system excitation and its harmonics in order to investigate the effects of the excitation on the modeling process. The fifth method is an attempt to apply constrained RID to obtain better transfer functions through more accurate modeling over certain frequency ranges. Section 4 presents some new C main files which were created to round out the functionality of the existing SSME transfer function modeling code. It is now possible to go from time data to transfer function models using only the C codes; it is not necessary to rely on external software. The new C main files and instructions for their use are included. Section 5 presents current and future enhancements to the XPLOT graphics program which was delivered with the initial software. Several new features which have been added to the program are detailed in the first part of this section. The remainder of Section 5 then lists some possible features which may be added in the future. Section 6 contains the conclusion section of this report. Section 6.1 is an overview of the work including a summary and observations relating to finding transfer functions with the SSME code. Section 6.2 contains information relating to future work on the project.

Irwin, R. Dennis↗

NASA NTRS1991

A parallel vectorized implementation of triple excitations in CCSD(T) - Application to the binding energies of the AlH3, AlH2F, AlHF2 and AlF3 dimers

An efficient method for various noniterative estimates of connected triple excitations in coupled-cluster theory is outlined and related to a similar expression occurring in Moller-Plesset perturbation theory. The method is highly vectorized and capable of utilizing multiple processors on a shared-memory machine, leading to computational rates in excess of one billion floating-point operations per second on four processors of a CRAY Y-MP. Using the new procedure, the binding energies of the D(2h) diborane-type dimers of AlH3, AlH2F, AlHF2, and AlF3 have been determined to be 32, 40, 20, and 47 kcal/mol, respectively. For Al2F6, the correlation procedure includes 232 molecular orbitals and over 1.5 x 10 to the 6th single and double coupled-cluster amplitudes, effectively accounting for over 2 x 10 to the 9th connected triple excitations.

Rendell, Alistair P.↗

DOE OSTI2019

Controlling internal degrees: general discussion

Temperature is a very important concept in ion spectroscopy and its control has seen tremendous advances in recent years. Cryogenic ion traps are now found in many experiments and serve as thermostats, reaction vessels, clustering sources and spectroscopic sample chambers. The control over internal degrees of freedom provides routes to forming ions that are relevant at cryogenic temperatures such as in the interstellar medium and to control and lock-in conformations. It also provides a condensing environment and allows weakly bound clusters to be formed, which in turn provides new routes to action spectroscopy. This session will discuss the most recent developments in cryogenic spectroscopy and the research it enables and how new theoretical methods allow the prediction of experimental spectra as a function of temperature. The main discussion topics will cover: the best cooling methods; how temperature can be used to form exotic clusters and weakly interacting systems; and the possibilities for probing conformational locking and metastable structures.

Ahmed, Musahid↗

DOE OSTI2024

40 Ca transverse response function from coupled-cluster theory

Here, we present calculations of the 40 Ca transverse response function obtained from coupled-cluster theory used in conjunction with the Lorentz integral transform method. We employ nuclear forces derived at next-to-next-to leading order in chiral effective field theory with and without Δ degrees of freedom. We first benchmark this approach on the 4 He nucleus and compare both the transverse sum rule and the response function to earlier calculations based on different methods. As expected from the power counting of the chiral expansion of electromagnetic currents and from previous studies, our results retaining only one-body term underestimate the experimental data for 4 He by about 20%. However, when the method is applied to 40 Ca at the same order of the expansion, response functions do not lack strength and agree well with the world electron scattering data. We discuss various sources of theoretical uncertainties and comment on the comparison of our results with the available experiments.

39 ≤ A ≤ 58↗

DOE OSTI2014

Generalized constructive tree weights

The Loop Vertex Expansion (LVE) is a quantum field theory (QFT) method which explicitly computes the Borel sum of Feynman perturbation series. This LVE relies in a crucial way on symmetric tree weights which define a measure on the set of spanning trees of any connected graph. In this paper we generalize this method by defining new tree weights. They depend on the choice of a partition of a set of vertices of the graph, and when the partition is non-trivial, they are no longer symmetric under permutation of vertices. Nevertheless we prove they have the required positivity property to lead to a convergent LVE; in fact we formulate this positivity property precisely for the first time. Our generalized tree weights are inspired by the Brydges-Battle-Federbush work on cluster expansions and could be particularly suited to the computation of connected functions in QFT. Several concrete examples are explicitly given.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗

NASA NTRS1991

Euler solutions to nonlinear acoustics of non-lifting hovering rotor blades

For the first time a computational fluid dynamics (CFD) method is used to calculate directly the high-speed impulsive (HSI) noise of a non-lifting hovering rotor blade out to a distance of over three rotor radii. In order to accurately propagate the acoustic wave in a stable and efficient manner, an implicit upwind-biased Euler method is solved on a grid with points clustered along the line of propagation. A detailed validation of the code is performed for a rectangular rotor blade at tip Mach numbers ranging from 0.88 to 0.92. The agreement with experiment is excellent at both the sonic cylinder and at 2.18 rotor radii. The agreement at 3.09 rotor radii is still very good, showing improvements over the results from the best previous method. Grid sensitivity studies indicate that with special attention to the location of the boundaries a grid with approximately 60,000 points is adequate. This results in a computational time of approximately 40 minutes on a Cray-XMP. The practicality of the method to calculate HSI noise is demonstrated by expanding the scope of the investigation to examine the rectangular blade as well as a highly swept and tapered blade over a tip Mach number range of 0.80 to 0.95. Comparisons with experimental data are excellent and the advantages of planform modifications are clearly evident. New insight is gained into the mechanisms of nonlinear propagation and the minimum distance at which a valid comparison of different rotors can be made: approximately two rotor radii from the center of rotation.

Baeder, J. D.↗

NASA NTRS1991

Euler solutions to nonlinear acoustics of non-lifting rotor blades

Baeder, J. D.↗

DOE OSTI2020

The Generality of the GUGA MRCI Approach in COLUMBUS for Treating Complex Quantum Chemistry

The core part of the program system COLUMBUS allows highly efficient calculations using variational multireference (MR) methods in the framework of configuration interaction with single and double excitations (MR-CISD) and averaged quadratic coupled-cluster calculations (MR-AQCC), based on uncontracted sets of configurations and the graphical unitary group approach (GUGA). The availability of analytic MR-CISD and MR-AQCC energy gradients and analytic nonadiabatic couplings for MR-CISD enables exciting applications including, e.g., investigations of π-conjugated biradicaloid compounds, calculations of multitudes of excited states, development of diabatization procedures, and furnishing the electronic structure information for on-the-fly surface nonadiabatic dynamics. With fully variational uncontracted spin-orbit MRCI, COLUMBUS provides a unique possibility of performing high-level calculations on compounds containing heavy atoms up to lanthanides and actinides. Crucial for carrying out all of these calculations effectively is the availability of an efficient parallel code for the CI step. Configuration spaces of several billion in size now can be treated quite routinely on standard parallel computer clusters. Emerging developments in COLUMBUS, including the all configuration mean energy multiconfiguration self-consistent field method and the graphically contracted function method, promise to allow practically unlimited configuration space dimensions. Spin density based on the GUGA approach, analytic spin-orbit energy gradients, possibilities for local electron correlation MR calculations, development of general interfaces for nonadiabatic dynamics, and MRCI linear vibronic coupling models conclude this overview.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

DOE OSTI2013

A measurement of the kinetic Sunyaev-Zel'dovich signal toward MACS J0717.5+3745

We report our analysis of MACS J0717.5+3745 using 140 and 268 GHz Bolocam data collected at the Caltech Submillimeter Observatory. We detect extended Sunyaev-Zel'dovich (SZ) effect signal at high significance in both Bolocam bands, and we employ Herschel-SPIRE observations to subtract the signal from dusty background galaxies in the 268 GHz data. We constrain the two-band SZ surface brightness toward two of the sub-clusters of MACS J0717.5+3745: the main sub-cluster (named C), and a sub-cluster identified in spectroscopic optical data to have a line-of-sight velocity of +3200 km s{sup –1} (named B). We determine the surface brightness in two separate ways: via fits of parametric models and via direct integration of the images. For both sub-clusters, we find consistent surface brightnesses from both analysis methods. We constrain spectral templates consisting of relativistically corrected thermal and kinetic SZ signals, using a jointly-derived electron temperature from Chandra and XMM-Newton under the assumption that each sub-cluster is isothermal. The data show no evidence for a kinetic SZ signal toward sub-cluster C, but they do indicate a significant kinetic SZ signal toward sub-cluster B. The model-derived surface brightnesses for sub-cluster B yield a best-fit, line-of-sight velocity of v{sub z} = +3450 ± 900 km s{sup –1}, with (1 – Prob[v{sub z} ≥ 0]) = 1.3 × 10{sup –5} (4.2σ away from 0 for a Gaussian distribution). The directly integrated sub-cluster B SZ surface brightnesses provide a best-fit v{sub z} = +2550 ± 1050 km s{sup –1}, with (1 – Prob[v{sub z} ≥ 0]) = 2.2 × 10{sup –3} (2.9σ).

79 ASTRONOMY AND ASTROPHYSICS↗