Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “simulation kernel”

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 235 records · Page 13
All sourcesNASA NTRSDOE OSTI
DOE OSTI2023

Numerical and Experimental Study of an Aircraft Igniter Plasma Jet Discharge

The spark discharge of an aircraft plasma jet igniter is studied using high-fidelity numerical simulations and X-ray radiography measurements. The target problem here features the thermal expansion of hot gas introduced by the electric spark within a confined igniter cavity, which eventually evolves into a pulsed jet of a high-temperature kernel. A comprehensive set of models adapted from existing strategies for internal combustion engine spark plug discharge is extended to the target problem, including the modeling of energy deposition, plasma reactions, thermodynamic properties, and heat losses. A series of validation and parameter studies are performed and presented. The kernel size is found to be sensitive to heat losses arising from radiation and hot gas remained within the discharge cavity, rather than heat conduction to the wall in the discharge cavity. Depending on the enforced shape of the post-breakdown electric arc, the spark kernel can be off-centered, tilted, and considerably asymmetric. These features have been previously not considered when studying such igniter configurations and may have a first-order impact on the ignition process. Provided a proper setup of the heat loss models and electric arc shape, the numerical results are quantitatively comparable to the experimental results in terms of the kernel size, shape, and velocity throughout different stages after the spark discharge.

Tang, Yihao↗
DOE OSTI2024

On-the-fly machine learned force fields for the study of warm dense matter: Application to diffusion and viscosity of CH

We develop a framework for on-the-fly machine learned force field (MLFF) molecular dynamics (MD) simulations of warm dense matter (WDM). In particular, we employ an MLFF scheme based on the kernel method and Bayesian linear regression, with the training data generated from the Kohn–Sham density functional theory (DFT) using the Gauss spectral quadrature method, within which we calculate energies, atomic forces, and stresses. We verify the accuracy of the formalism by comparing the predicted properties of warm dense carbon with recent Kohn–Sham DFT results in the literature. In so doing, we demonstrate that ab initio MD simulations of WDM can be accelerated by up to three orders of magnitude, while retaining ab initio accuracy. We apply this framework to calculate the diffusion coefficients and shear viscosity of CH at a density of 1 g/cm3 and temperatures in the range of 75 000–750 000 K. We find that the self- and inter-diffusion coefficients and the viscosity obey a power law with temperature, and that the diffusion coefficient results suggest a weak coupling between C and H in CH. In addition, we find agreement within standard deviation with previous results for C and CH but disagreement for H, demonstrating the need for ab initio calculations as presented here.

75 CONDENSED MATTER PHYSICS, SUPERCONDUCTIVITY AND↗
DOE OSTI2017

Uncertainty analysis of wavelet-based feature extraction for isotope identification on NaI gamma-ray spectra

Low-resolution isotope identifiers are widely deployed for nuclear security purposes, but these detectors currently demonstrate problems in making correct identifications in many typical usage scenarios. While there are many hardware alternatives and improvements that can be made, performance on existing low resolution isotope identifiers should be able to be improved by developing new identification algorithms. We have developed a wavelet-based peak extraction algorithm and an implementation of a Bayesian classifier for automated peak-based identification. The peak extraction algorithm has been extended to compute uncertainties in the peak area calculations. To build empirical joint probability distributions of the peak areas and uncertainties, a large set of spectra were simulated in MCNP6 and processed with the wavelet-based feature extraction algorithm. Kernel density estimation was then used to create a new component of the likelihood function in the Bayesian classifier. Furthermore, identification performance is demonstrated on a variety of real low-resolution spectra, including Category I quantities of special nuclear material.

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗
DOE OSTI2017

Pillars of the Mantle: Imaging the Interior of the Earth with Adjoint Tomography. In: PEARC17 Proceedings of the Practice and Experience in Advanced Research Computing 2017 on Sustainability, Success and Impact, Article No. 75

In this work, we investigate global seismic tomographic models obtained by spectral-element simulations of seismic wave propagation and adjoint methods. Global crustal and mantle models are obtained based on an iterative conjugate-gradient type of optimization scheme. Forward and adjoint seismic wave propagation simulations, which result in synthetic seismic data to make measurements and data sensitivity kernels to compute gradient for model updates, respectively, are performed by the SPECFEM3D_GLOBE package [1] [2] at the Oak Ridge Leadership Computing Facility (OLCF) to study the structure of the Earth at unprecedented levels. Using advances in solver techniques that run on the GPUs on Titan at the OLCF, scientists are able to perform large-scale seismic inverse modeling and imaging. Using seismic data from global and regional networks from global CMT earthquakes, scientists are using SPECFEM3D_GLOBE to understand the structure of the mantle layer of the Earth. Visualization of the generated data sets provide an effective way to understand the computed wave perturbations which define the structure of mantle in the Earth.

Pugmire, David↗
DOE OSTI2024

Blueprints for Training Information Bottlenecks for Collider Analyses

Dimensionality reduction is a crucial aspect of data analysis in high energy physics, even if accompanied by information loss. Several methods, including histogram- and kernel-based analyses, are only computationally feasible for low-dimensional data. Furthermore, simulation models used in HEP can often only be validated for low-dimensional data. We provide several blueprints for using machine learning to create low-dimensional data representations (continuous event variables and discrete classification labels) for use in signal discovery and parameter estimation tasks. We also describe how to design the learned representation to facilitate a) searches with unknown model parameters and b) validation of simulation models in data control regions.

43 PARTICLE ACCELERATORS↗
DOE OSTI2015

SU-E-T-36: A GPU-Accelerated Monte-Carlo Dose Calculation Platform and Its Application Toward Validating a ViewRay Beam Model

Purpose: To build a fast, accurate and easily-deployable research platform for Monte-Carlo dose calculations. We port the dose calculation engine PENELOPE to C++, and accelerate calculations using GPU acceleration. Simulations of a Co-60 beam model provided by ViewRay demonstrate the capabilities of the platform. Methods: We built software that incorporates a beam model interface, CT-phantom model, GPU-accelerated PENELOPE engine, and GUI front-end. We rewrote the PENELOPE kernel in C++ (from Fortran) and accelerated the code on a GPU. We seamlessly integrated a Co-60 beam model (obtained from ViewRay) into our platform. Simulations of various field sizes and SSDs using a homogeneous water phantom generated PDDs, dose profiles, and output factors that were compared to experiment data. Results: With GPU acceleration using a dated graphics card (Nvidia Tesla C2050), a highly accurate simulation – including 100*100*100 grid, 3×3×3 mm3 voxels, <1% uncertainty, and 4.2×4.2 cm2 field size – runs 24 times faster (20 minutes versus 8 hours) than when parallelizing on 8 threads across a new CPU (Intel i7-4770). Simulated PDDs, profiles and output ratios for the commercial system agree well with experiment data measured using radiographic film or ionization chamber. Based on our analysis, this beam model is precise enough for general applications. Conclusions: Using a beam model for a Co-60 system provided by ViewRay, we evaluate a dose calculation platform that we developed. Comparison to measurements demonstrates the promise of our software for use as a research platform for dose calculations, with applications including quality assurance and treatment plan verification.

60 APPLIED LIFE SCIENCES↗
DOE OSTI2017

Behavior-based approach to misuse detection of a simulated SCADA system - 339

This paper presents the initial findings in applying a behavior-based approach for detection of unauthorized activities in a simulated Supervisory Control and Data Acquisition (SCADA) system. Misuse detection of this type utilizes fault-free system telemetry to develop empirical models that learn normal system behavior. Future monitored telemetry sources that show statistically significant deviations from this learned behavior may indicate an attack or other unwanted actions. The experimental test bed consists of a set of Linux based enterprise servers that were isolated from a larger university research cluster. All servers are connected to a private network and simulate several components and tasks seen in a typical SCADA system. Telemetry sources included kernel statistics, resource usages and internal system hardware measurements. For this study, the Auto Associative Kernel Regression (AAKR) and Auto Associative Multivariate State Estimation Technique (AAMSET) are employed to develop empirical models. Prognostic efficacy of these methods for computer security used several groups of signals taken from available telemetry classes. The Sequential Probability Ratio Test (SPRT) is used along with these models for intrusion detection purposes. The different intrusion types shown include host/network discovery, DoS, brute force login, privilege escalation and malicious exfiltration actions. For this study, all intrusion types tested displayed alterations in the residuals of much of the monitored telemetry and were able to be detected in all signal groups used by both model types. The methods presented can be extended and implemented to industries besides nuclear that use SCADA or business-critical networks. (authors)

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗
DOE OSTI2015

SU-E-J-175: Comparison of the Treatment Reproducibility of Tumors Affected by Breathing Motion

Purpose: The aim of the dose distribution simulations was to form a global idea of intensity-modulated radiation therapy (IMRT) realization, by its comparison to three-dimensional conformal radiation therapy (3DCRT) delivery for tumors affected by respiratory motion. Methods: In the group of 10patients both 3DCRT and IMRT plans were prepared.For each field the motion kernel was generated with the largest movement amplitude of 4;6 and 8mm.Additionally,the sets of reference measurements were made in no motion conditions(0 mm).The evaluation of plan delivery,using a diode array placed on moving platform,was based on the Gamma Index analysis with distance to agreement of 3mm and dose difference of 3%. Results: IMRT plans tended to spare doses delivered to lungs compared to 3DCRT.Nonetheless,analyzed volumes showed no significant difference between the static and dynamic techniques,except for the volumes of both lungs receiving 10 and 15Gy.After adding the components associated with the respiratory movement,all IMRT lung parameters evaluated for the ipsilateral,contralateral and both lungs together,revealed considerable differences between the 0vs.6, 0vs.8 and 4vs.8-mm amplitudes.Similar results were obtained for the 3DCRT lung measurements,but without significance between the 0vs.6-mm amplitude.Taking into account the CTV score parameter in 3DCRT and IMRT plans,there was no statistically significant difference between the motion patterns with the smallest amplitudes.The differences were found for the 8-mm amplitude when it was compared both with static conditions and 4-mm amplitude (for 3DCRT) and between 0vs.6, 0vs.8 and 4vs.8-mm amplitudes (for IMRT).All accepted and measured 3DCRT and IMRT doses to spinal cord,esophagus and heart were always below the QUANTEC limits. Conclusion: The application of IMRT technique in lung radiotherapy affords possibilities for reducing the lung doses.For maximal amplitudes of breathing trajectory below 4mm,the disagreement between CTV planned and delivered dose distribution can be neglected.Whereas respiratory motion with an amplitude above 5mm leads to significant changes in IMRT delivery.

60 APPLIED LIFE SCIENCES↗
DOE OSTI2018

Simulation of Properties of Images with Atomic Resolution in a Scanning Probe Microscope

A method for simulation of images in a scanning probe microscope (SPM) using simultaneous wavelet transform and median filtering is proposed. The wavelet transform with the fourth-order Daubechies kernel is used. Such a transform makes it possible to select details of different scales in the SPM image and, hence, study fractal properties of surfaces. Simulation is used to show that ultrahigh (atomic) resolution is possible in SPM provided that the size of the contact region in the probe–sample system is significantly greater than atomic size and the lattice atoms are randomly distributed. Contrast inversion in the SPM images in the multiscan mode is interpreted.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗
DOE OSTI2026

Simulation of electron Bernstein waves using FullWave with a 2D non-local hot plasma model

Hot plasma wave simulation capability is expanded in the FullWave code by updating the hybrid iterative solver in the code with a semi-implicit time stepping method. The new approach is used to simulate Electron Bernstein Wave (EBW) heating in over-dense spherical tokamak plasmas. The code’s hybrid iterative solver circumvents the prohibitive memory cost of direct methods by combining a time evolution of Maxwell’s equations with frequency-domain relaxation, while the conductivity kernel, calculated via 3D particle tracking, captures the essential non-local wave–particle interactions. One-dimensional EBW simulations verify the algorithm’s accuracy by demonstrating mode conversion from X-mode wave to EBW at the upper hybrid resonance and a strong cyclotron damping near the plasma core. Two-dimensional simulation reproduces the predicted short EBW wavelength and quantitatively matches the hot-plasma dispersion relation. This study demonstrates the fidelity of the hybrid solver for the electron cyclotron frequency range.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗
DOE OSTI2015

SU-F-BRD-06: Robust Dose Calculation in Intensity Modulated Proton Therapy

Purpose: Commissioning data for intensity modulated proton therapy (IMPT) must be post-processed by fits to ad-hoc functions to derive the dose calculation kernel parameters in a treatment planning system (TPS). Whether from experimental measurement or Monte Carlo simulation, the limited and noisy nature of such data makes this task very challenging. We present a method to improve the modeling of the lateral dose distribution of clinical energy proton beams in water to commission an in-house IMPT dose calculation engine. Methods: A linear sum of three Gaussian distribution functions was fitted to the lateral dose data in logarithmic scale. Starting values of fitting solutions were determined from the Generalized Highland Approximation. We exhaustively optimized the combinations of data weights with upper bounds of the fitting solutions to minimize confidence intervals of the fitting solutions while maintaining the coefficient of determination (R{sup 2}). Results: Across all energies, average confidence bounds improved 72.88% [Max: 88.28%, Min: 55.05%] for small angle coulomb scattering, 114.25% [409.13%, 66.72%,] for nuclear scattering, and 68.66% [141.09%, 33.27%] for large angle coulomb scattering, while the coefficients of determination of the fits (R{sup 2}) remained comparable. On average R {sup 2} only changed 0.18% and were very close to 1 (approx. 0.999). Wilcoxon signed rank tests comparing unweighted/unbounded fits with weighted/bounded fits averaged 0.0146 (Max: 0.177, Min: 7.05×10−{sup 7}) for small angle Coulomb, 0.0903 (0.945, 7.05×10−{sup 7}) for nuclear, and 0.254 (0.871, 1.86×10−{sup 6}) for large angle Coulomb scattering. This allows rejection of the null hypothesis for small angle Coulomb scattering at the 0.015 level and nuclear interaction at the 0.1 level. Conclusion: Optimal weights assigned to IMPT lateral dose data minimized fitting to stochastic noise in the tail region. Optimizing the upper bounds of fitting parameters improved the robustness of fitting solutions. Both led to a more physically meaningful dose calculation kernel for the TPS. NIH/NCI K25CA168984, Eagles Cancer Research Career Development, The Lawrence W. and Marilyn W. Matteson Fund for Cancer Research, Mayo ASU Seed Grant, and The Kemper Marley Foundation.

60 APPLIED LIFE SCIENCES↗
DOE OSTI2017

Computational Approach for Improving Three-Dimensional Sub-Surface Earth Structure for Regional Earthquake Hazard Simulations in the San Francisco Bay Area

In our Exascale Computing Project (ECP) we seek to simulate earthquake ground motions at much higher frequency than is currently possible. Previous simulations in the SFBA were limited to 0.5-1 Hz or lower (Aagaard et al. 2008, 2010), while we have recently simulated the response to 5 Hz. In order to improve confidence in simulated ground motions, we must accurately represent the three-dimensional (3D) sub-surface material properties that govern seismic wave propagation over a broad region. We are currently focusing on the San Francisco Bay Area (SFBA) with a Cartesian domain of size 120 x 80 x 35 km, but this area will be expanded to cover a larger domain. Currently, the United States Geologic Survey (USGS) has a 3D model of the SFBA for seismic simulations. However, this model suffers from two serious shortcomings relative to our application: 1) it does not fit most of the available low frequency (< 1 Hz) seismic waveforms from moderate (magnitude M 3.5-5.0) earthquakes; and 2) it is represented with much lower resolution than necessary for the high frequency simulations (> 5 Hz) we seek to perform. The current model will serve as a starting model for full waveform tomography based on 3D sensitivity kernels. This report serves as the deliverable for our ECP FY2017 Quarter 4 milestone to FY 2018 “Computational approach to developing model updates”. We summarize the current state of 3D seismic simulations in the SFBA and demonstrate the performance of the USGS 3D model for a few selected paths. We show the available open-source waveform data sets for model updates, based on moderate earthquakes recorded in the region. We present a plan for improving the 3D model utilizing the available data and further development of our SW4 application. We project how the model could be improved and present options for further improvements focused on the shallow geotechnical layers using dense passive recordings of ambient and human-induced noise.

58 GEOSCIENCES↗
DOE OSTI2015

Mapping the three-dimensional “x-shaped structure” in models of the Galactic bulge

Numerical simulations have shown that the X-shaped structure in the Milky Way bulge can naturally arise from bar instability and buckling instability. To understand the influence of buckling amplitude on the morphology of the X-shape, we analyze three self-consistent numerical simulations of barred galaxies with different buckling amplitudes (strong, intermediate, and weak). We derive the three-dimensional density using an adaptive kernel smoothing technique. The face-on isodensity surfaces are all elliptical, while in the edge-on view, the morphology of buckled bars transitions with increasing radius from a central boxy core to a peanut bulge and then to an extended, thin bar. Based on these isodensity surfaces at different density levels, we find no clear evidence for a well-defined structure shaped like the letter X. The X-shaped structure is more peanut-like, the visual perception of which is probably enhanced by the pinched inner concave isodensity contours. The peanut bulge can reproduce qualitatively the observed bimodal distributions that were used as evidence for the discovery of the X-shape. The central boxy core is shaped like an oblong tablet, extending ∼500 pc above and below the Galactic plane (|b|∼4{sup ∘}). From the solar perspective, lines of sight passing through the central boxy core do not show bimodal distributions. This generally agrees with the observations that double peaks merge at |b| ∼ 4°–5° from the Galactic plane, indicating the presence of a possibly similar structure in the Galactic bulge.

79 ASTRONOMY AND ASTROPHYSICS↗
DOE OSTI2015

Ion aggregation in high salt solutions. III. Computational vibrational spectroscopy of HDO in aqueous salt solutions

The vibrational frequency, frequency fluctuation dynamics, and transition dipole moment of the O—D stretch mode of HDO molecule in aqueous solutions are strongly dependent on its local electrostatic environment and hydrogen-bond network structure. Therefore, the time-resolved vibrational spectroscopy the O—D stretch mode has been particularly used to investigate specific ion effects on water structure. Despite prolonged efforts to understand the interplay of O—D vibrational dynamics with local water hydrogen-bond network and ion aggregate structures in high salt solutions, still there exists a gap between theory and experiment due to a lack of quantitative model for accurately describing O—D stretch frequency in high salt solutions. To fill this gap, we have performed numerical simulations of Raman scattering and IR absorption spectra of the O—D stretch mode of HDO in highly concentrated NaCl and KSCN solutions and compared them with experimental results. Carrying out extensive quantum chemistry calculations on not only water clusters but also ion-water clusters, we first developed a distributed vibrational solvatochromic charge model for the O—D stretch mode in aqueous salt solutions. Furthermore, the non-Condon effect on the vibrational transition dipole moment of the O—D stretch mode was fully taken into consideration with the charge response kernel that is non-local polarizability density. From the fluctuating O—D stretch mode frequencies and transition dipole vectors obtained from the molecular dynamics simulations, the O—D stretch Raman scattering and IR absorption spectra of HDO in salt solutions could be calculated. The polarization effect on the transition dipole vector of the O—D stretch mode is shown to be important and the asymmetric line shapes of the O—D stretch Raman scattering and IR absorption spectra of HDO especially in highly concentrated NaCl and KSCN solutions are in quantitative agreement with experimental results. We anticipate that this computational approach will be of critical use in interpreting linear and nonlinear vibrational spectroscopies of HDO molecule that is considered as an excellent local probe for monitoring local electrostatic and hydrogen-bonding environment in not just salt but also other confined and crowded solutions.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2017

Roofline Analysis in the Intel® Advisor to Deliver Optimized Performance for applications on Intel® Xeon Phi™ Processor

In this session we show, in two case studies, how the roofline feature of Intel Advisor has been utilized to optimize the performance of kernels of the XGC1 and PICSAR codes in preparation for Intel Knights Landing architecture. The impact of the implemented optimizations and the benefits of using the automatic roofline feature of Intel Advisor to study performance of large applications will be presented. This demonstrates an effective optimization strategy that has enabled these science applications to achieve up to 4.6 times speed-up and prepare for future exascale architectures. # Goal/Relevance of Session The roofline model [1,2] is a powerful tool for analyzing the performance of applications with respect to the theoretical peak achievable on a given computer architecture. It allows one to graphically represent the performance of an application in terms of operational intensity, i.e. the ratio of flops performed and bytes moved from memory in order to guide optimization efforts. Given the scale and complexity of modern science applications, it can often be a tedious task for the user to perform the analysis on the level of functions or loops to identify where performance gains can be made. With new Intel tools, it is now possible to automate this task, as well as base the estimates of peak performance on measurements rather than vendor specifications. The goal of this session is to demonstrate how the roofline feature of Intel Advisor can be used to balance memory vs. computation related optimization efforts and effectively identify performance bottlenecks. A series of typical optimization techniques: cache blocking, structure refactoring, data alignment, and vectorization illustrated by the kernel cases will be addressed. # Description of the codes ## XGC1 The XGC1 code [3] is a magnetic fusion Particle-In-Cell code that uses an unstructured mesh for its Poisson solver that allows it to accurately resolve the edge plasma of a magnetic fusion device. After recent optimizations to its collision kernel [4], most of the computing time is spent in the electron push (pushe) kernel, where these optimization efforts have been focused. The kernel code scaled well with MPI+OpenMP but had almost no automatic compiler vectorization, in part due to indirect memory addresses and in part due to low trip counts of low-level loops that would be candidates for vectorization. Particle blocking and sorting have been implemented to increase trip counts of low-level loops and improve memory locality, and OpenMP directives have been added to vectorize compute-intensive loops that were identified by Advisor. The optimizations have improved the performance of the pushe kernel 2x on Haswell processors and 1.7x on KNL. The KNL node-for-node performance has been brought to within 30% of a NERSC Cori phase I Haswell node and we expect to bridge this gap by reducing the memory footprint of compute intensive routines to improve cache reuse. ## PICSAR is a Fortran/Python high-performance Particle-In-Cell library targeting at MIC architectures first designed to be coupled with the PIC code WARP for the simulation of laser-matter interaction and particle accelerators. PICSAR also contains a FORTRAN stand-alone kernel for performance studies and benchmarks. A MPI domain decomposition is used between NUMA domains and a tile decomposition (cache-blocking) handled by OpenMP has been added for shared-memory parallelism and better cache management. The so-called current deposition and field gathering steps that compose the PIC time loop constitute major hotspots that have been rewritten to enable more efficient vectorization. Particle communications between tiles and MPI domain has been merged and parallelized. All considered, these improvements provide speedups of 3.1 for order 1 and 4.6 for order 3 interpolation shape factors on KNL configured in SNC4 quadrant flat mode. Performance is similar between a node of cori phase 1 and KNL at order 1 and better on KNL by a factor 1.6 at order 3 with the considered test case (homogeneous thermal plasma).

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2018

A generalized framework for reduced-order modeling of a wind turbine wake

A reduced-order model for a wind turbine wake is sought from large eddy simulation data. Fluctuating velocity fields are combined in the correlation tensor to form the kernel of the proper orthogonal decomposition (POD). Proper orthogonal decomposition modes resulting from the decomposition represent the spatially coherent turbulence structures in the wind turbine wake; eigenvalues delineate the relative amount of turbulent kinetic energy associated with each mode. Back-projecting the POD modes onto the velocity snapshots produces dynamic coefficients that express the amplitude of each mode in time. A reduced-order model of the wind turbine wake (wakeROM) is defined through a series of polynomial parameters that quantify mode interaction and the evolution of each POD mode coefficients. The resulting system of ordinary differential equations models the wind turbine wake composed only of the large-scale turbulent dynamics identified by the POD. Tikhonov regularization is used to recalibrate the dynamical system by adding additional constraints to the minimization seeking polynomial parameters, reducing error in the modeled mode coefficients. The wakeROM is periodically reinitialized with new initial conditions found by relating the incoming turbulent velocity to the POD mode coefficients through a series of open-loop transfer functions. The wakeROM reproduces mode coefficients to within 25.2%, quantified through the normalized root-mean-square error. Furthermore, a high-level view of the modeling approach is provided as a platform to discuss promising research directions, alternate processes that could benefit stability and efficiency, and desired extensions of the wakeROM.

17 WIND ENERGY↗
DOE OSTI2025

DONKEY: A Flexible and Accurate Algorithm for Clustering

We propose an accurate clustering algorithm suitable for the varied and multidimensional data sets that correspond to temporal snapshots from on-the-fly nonadiabatic trajectory-based simulations of photoexcited dynamics. The algorithm approximates the underlying probability density function using variable kernel density estimation, with local maxima corresponding to cluster centers. Each data point is then assigned to one of the maxima by employing a maximization procedure. Finally, clusters artificially separated by minor fluctuations in the probability density are merged. The algorithm does not require parameter tuning, which ensures flexibility and reduces the risk of bias. It is tested on several synthetic data sets, where it consistently outperforms conventional clustering algorithms. As a final example, the algorithm is applied to the excited dynamics of the norbornadiene ⇌ quadricyclane (C 7 H 8 ) molecular photoswitch, demonstrating how distinct reaction pathways can be identified.

algorithms↗
DOE OSTI2020

Scalability of high-performance PDE solvers

Performance tests and analyses are critical to effective high-performance computing software development and are central components in the design and implementation of computational algorithms for achieving faster simulations on existing and future computing architectures for large-scale application problems. In this article, we explore performance and space-time trade-offs for important compute-intensive kernels of large-scale numerical solvers for partial differential equations (PDEs) that govern a wide range of physical applications. We consider a sequence of PDE-motivated bake-off problems designed to establish best practices for efficient high-order simulations across a variety of codes and platforms. We measure peak performance (degrees of freedom per second) on a fixed number of nodes and identify effective code optimization strategies for each architecture. In addition to peak performance, we identify the minimum time to solution at 80% parallel efficiency. The performance analysis is based on spectral and p-type finite elements but is equally applicable to a broad spectrum of numerical PDE discretizations, including finite difference, finite volume, and h-type finite elements.

97 MATHEMATICS AND COMPUTING↗