Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “Network partitioning”

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 253 records · Page 14

Memory embedded non-intrusive reduced order modeling of non-ergodic flows

Generating a digital twin of any complex system requires modeling and computational approaches that are efficient, accurate, and modular. Traditional reduced order modeling techniques are targeted at only the first two, but the novel nonintrusive approach we present in this study is an attempt at taking all three into account effectively compared to their traditional counterparts. Based on dimensionality reduction using proper orthogonal decomposition (POD), we introduce a long short-term memory neural network architecture together with a principal interval decomposition (PID) framework as an enabler to account for localized modal deformation. As an effective partitioning tool for breaking the Kolmogorov barrier, our PID framework, therefore, can be considered a key element in the accurate reduced order modeling of convective flows. Our applications for convection-dominated systems governed by Burgers, Navier-Stokes, and Boussinesq equations demonstrate that the proposed approach yields significantly more accurate predictions than the POD-Galerkin method and could be a key enabler toward near real-time predictions of unsteady flows.

42 ENGINEERING↗

Crystalline Ceramic Waste Forms: Comparison Of Reference Process For Ceramic Waste Form Fabrication

The research conducted in this work package is aimed at taking advantage of the long term thermodynamic stability of crystalline ceramics to create more durable waste forms (as compared to high level waste glass) in order to reduce the reliance on engineered and natural barrier systems. Durable ceramic waste forms that incorporate a wide range of radionuclides have the potential to broaden the available disposal options and to lower the storage and disposal costs associated with advanced fuel cycles. Assemblages of several titanate phases have been successfully demonstrated to incorporate radioactive waste elements, and the multiphase nature of these materials allows them to accommodate variation in the waste composition. Recent work has shown that they can be produced from a melting and crystallization process. The objective of this report is to explore the phase formation and microstructural differences between lab scale melt processing in varying gas environments with alternative densification processes such as Hot Pressing (HP) and Spark Plasma Sintering (SPS). The waste stream used as the basis for the development and testing is a simulant derived from a combination of the projected Cs/Sr separated stream, the Trivalent Actinide - Lanthanide Separation by Phosphorous reagent Extraction from Aqueous Komplexes (TALSPEAK) waste stream consisting of lanthanide fission products, the transition metal fission product waste stream resulting from the transuranic extraction (TRUEX) process, and a high molybdenum concentration with relatively low noble metal concentrations. Melt processing as well as solid state sintering routes SPS and HP demonstrated the formation of the targeted phases; however differences in microstructure and elemental partitioning were observed. In SPS and HP samples, hollandite, pervoskite/pyrochlore, zirconolite, metallic alloy and TiO 2 and Al 2 O 3 were observed distributed in a network of fine grains with small residual pores. The titanate phases that incorporate M +3 rare earth elements were observed to be distinct phases (ex. Nd 2 Ti 2 O 7 with less degree of substitution as compared to the more homogeneous melt processed samples where a high degree of substitution and variation of composition within grains was observed. Liquid phase sintering was enhanced in reducing gas environments and resulted in large (10-200 microns) irregular shaped grains along with large voids associated with the melt process; SPS and HP samples exhibited finer grain size with smaller voids. Metallic alloys were observed in the bulk of the sample for SPS and HP samples, but were found at the bottom of the crucible in melt processed trials. These results indicate that for a first melter trial, the targeted phases can be formed in air by utilizing Ti/TiO 2 additives which aid phase formation and improve the electrical conductivity. Ultimately, a melter run in reducing gas environments would be beneficial to study differences in phase formation and elemental partitioning.

12 MANAGEMENT OF RADIOACTIVE AND NON-RADIOACTIVE W↗

Mesoscopic-microscopic spatial stochastic simulation with automatic system partitioning

The reaction-diffusion master equation (RDME) is a model that allows for efficient on-lattice simulation of spatially resolved stochastic chemical kinetics. Compared to off-lattice hard-sphere simulations with Brownian dynamics or Green’s function reaction dynamics, the RDME can be orders of magnitude faster if the lattice spacing can be chosen coarse enough. However, strongly diffusion-controlled reactions mandate a very fine mesh resolution for acceptable accuracy. It is common that reactions in the same model differ in their degree of diffusion control and therefore require different degrees of mesh resolution. This renders mesoscopic simulation inefficient for systems with multiscale properties. Mesoscopic-microscopic hybrid methods address this problem by resolving the most challenging reactions with a microscale, off-lattice simulation. However, all methods to date require manual partitioning of a system, effectively limiting their usefulness as “black-box” simulation codes. In this paper, we propose a hybrid simulation algorithm with automatic system partitioning based on indirect a priori error estimates. We demonstrate the accuracy and efficiency of the method on models of diffusion-controlled networks in 3D.

97 MATHEMATICS AND COMPUTING↗

Adaptive Visual Sort and Summary of Micrographic Images of Nanoparticles for Forensic Analysis

Forensic analysis of nanoparticles is often conducted through the collection and identifi- cation of electron microscopy images to determine the origin of suspected nuclear material. Each image is carefully studied by experts for classification of materials based on texture, shape, and size. Manually inspecting large image datasets takes enormous amounts of time. However, automatic classification of large image datasets is a challenging problem due to the complexity involved in choosing image features, the lack of training data available for effective machine learning methods, and the availability of user interfaces to parse through images. Therefore, a significant need exists for automated and semi-automated methods to help analysts perform accurate image classification in large image datasets. We present INStINCt, our Intelligent Signature Canvas, as a framework for quickly organizing image data in a web based canvas framework. Images are partitioned using small sets of example images, chosen by users, and presented in an optimal layout based on features derived from convolutional neural networks.

active canvas↗

RingX: Scalable Parallel Attention for Long-Context Learning on HPC

The attention mechanism has become foundational for remarkable AI breakthroughs since the introduction of the Transformer, driving the demand for increasingly longer context to power frontier models such as large-scale reasoning language models and high-resolution image/video generators. However, its quadratic computational and memory complexities present substantial challenges. Current state-of-the-art parallel attention methods, such as ring attention, are widely adopted for long-context training but utilize a point-to-point communication strategy that fails to fully exploit the capabilities of modern HPC network architectures. In this work, we propose ringX, a scalable family of parallel attention methods optimized explicitly for HPC systems. By enhancing workload partitioning, refining communication patterns, and improving load balancing, ringX achieves up to 3.4 × speedup compared to conventional ring attention on the Frontier supercomputer. Optimized for both bi-directional and causal attention mechanisms, ringX demonstrates its effectiveness through training benchmarks of a Vision Transformer (ViT) on a climate dataset and a Generative Pre-Trained Transformer (GPT) model, Llama3 8B. Our method attains an end-to-end training speedup of approximately 1.5 × in both scenarios. To our knowledge, the achieved 38% model FLOPs utilization (MFU) for training Llama3 8B with a 1M-token sequence length on 4,096 GPUs represents one of the highest training efficiencies reported for long-context learning on HPC systems. Our code implementation is available at https://github.com/jqyin/ringX-attention.

Yin, Junqi [ORNL] (ORCID:0000000338435520)↗

Use of networked workstations for parallel nonlinear structural dynamic simulations of rotating bladed-disk assemblies

The principal objective of this research is to investigate, develop and demonstrate coarse-grained, parallel-processing strategies for nonlinear dynamic simulations for rotating bladed-disk assemblies. The parallel -processing strategies addressed include numerical algorithms for parallel nonlinear solutions and techniques to effect load balancing among processors. The parallel environment employed is a distributed-memory, coarse-grained one consisting of networked workstations. A parallel explicit time integration method has been implemented for transient nonlinear solutions of rotationg bladed-disk assemblies. Automatic domain partitioning techniques have been investigated for load balancing among processors. Advanced computing environments, data structures and interactive computer graphics all contribute to an integrated parallel finite element analysis system to facilitate more efficient and powerful dynamic simulations.

Hsieh, Shang-Hsien↗

Software defined grid energy storage

Today, consumer battery installations are isolated, physical devices. Virtual power plants (VPPs) allow consumer devices to aggregate for grid services, but they are are vertically integrated, vendor controlled systems (e.g., Tesla’s VPP). Consumer batteries are therefore unable to participate in energy markets or other grid services outside what their vendor provides. We describe a software system that provides software control of multiple, networked battery energy storage systems in the electric grid. The system introduces two new ideas that enable flexible and dependable management of energy storage. The first is a virtual battery, which can either partition a battery or aggregate multiple batteries. The second is a reservation-based API which allows asynchronous control of batteries to meet contractual guarantees in a safe and dependable manner. Virtual batteries and a reservation-based API address the unique challenges of achieving high and efficient utilization of energy storage systems, including heterogeneity of battery systems such as varying C-rates, participation in energy markets, utility bill management systems, community resource sharing, and reliability. Using a testbed comprised of sonnen Inc. storage units installed in several homes and a lab, we demonstrate that virtualized batteries can seamlessly replace physical batteries, flexibly manage energy storage resources, isolate multiple clients using a shared battery, and create new energy storage applications.

25 ENERGY STORAGE↗

A lightweight network anomaly detection technique

While the network anomaly detection is essential in network operations and management, it becomes further challenging to perform the first line of detection against the exponentially increasing volume of network traffic. In this paper, we develop a technique for the first line of online anomaly detection with two important considerations: (i) availability of traffic attributes during the monitoring time, and (ii) computational scalability for streaming data. The presented learning technique is lightweight and highly scalable with the beauty of approximation based on the grid partitioning of the given dimensional space. With the public traffic traces of KDD Cup 1999 and NSL-KDD, we show that our technique yields 98.5% and 83% of detection accuracy, respectively, only with a couple of readily available traffic attributes that can be obtained without the help of post-processing. Finally, the results are at least comparable with the classical learning methods including decision tree and random forest, with approximately two orders of magnitude faster learning performance.

97 MATHEMATICS AND COMPUTING↗

Phenolic Polymer Interactions with Water and Ethylene Glycol Solvents

Interactions between pre-cured phenolic polymer chains and a solvent have a significant impact on the structure and properties of the final post-cured phenolic resin. Developing an understanding of the nature of these interactions is important and will aid in the selection of the proper solvent that will lead to the desired final product. Here, we investigate the role of the phenolic chain structure and the solvent type on the overall solvation performance of the system through ab initio techniques and molecular dynamics simulations. Two types of solvents are considered: ethylene glycol (EGL) and H2O. Three phenolic chain structures are considered, including two novolac-type chains with either an ortho-ortho (OON) or an ortho-para (OPN) backbone network and a resole-type (RES) chain with an ortho-ortho network. Each system is characterized through a structural analysis of the solvation shell and the hydrogen-bonding environment as well as through a quantification of the solvation free energy along with partitioned interaction energies between specific molecular species. The combination of simulations and the analyses indicate that EGL provides a higher solvation free energy than H2O due to more energetically favorable hydrophilic interactions as well as favorable hydrophobic interactions between CH element groups. In addition, the phenolic chain structure significantly affects the solvation performance, with OON having limited intermolecular hydrogen-bond formations, while OPN and RES interact more favorably with the solvent molecules. The results suggest that a resole-type phenolic chain with an ortho-para network should have the best solvation performance in EGL, H2O, and other similar solvents.

Haskins, Justin B.↗

Optimized Lie–Trotter–Suzuki decompositions for two and three non-commuting terms

Lie–Trotter–Suzuki decompositions are an efficient way to approximate operator exponentials exp ( t H ) when H is a sum of n (non-commuting) terms which, individually, can be exponentiated easily. They are employed in time-evolution algorithms for tensor network states, digital quantum simulation protocols, path integral methods like quantum Monte Carlo, and splitting methods for symplectic integrators in classical Hamiltonian systems. Here, we provide optimized decompositions up to order t 6 . The leading error term is expanded in nested commutators (Hall bases) and we minimize the 1-norm of the coefficients. For n = 2 terms, several of the optima we find are close to those in McLachlan (1995). Generally, our results substantially improve over unoptimized decompositions by Forest, Ruth, Yoshida, and Suzuki. We explain why these decompositions are sufficient to efficiently simulate any one- or two-dimensional lattice model with finite-range interactions. This follows by solving a partitioning problem for the interaction graph.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Traffic study of a computer system.

A study which may guide the operations of existing computer installations, as well as the design of future networks, is described. Performance data and evaluations are considered with reference to interarrival time, users' habits, waiting time for execution, time spent in a partition, figures of merit, and states of the system. The analysis of the variables proceeds from examination of typical data with appropriate statistical tests to conclusions about the possible state of nature.

Cramer, R. L.↗

A differentiable approach to the maximum independent set problem using dataless neural networks

The success of machine learning solutions for reasoning about discrete structures has brought attention to its adoption within combinatorial optimization algorithms. Such approaches generally rely on supervised learning by leveraging datasets of the combinatorial structures of interest drawn from some distribution of problem instances. Reinforcement learning has also been employed to find such structures. Here, in this paper, we propose a different approach in that no data is required for training the neural networks that produce the solution. In this sense, what we present is not a machine learning solution, but rather one that is dependent on neural networks and where backpropagation is applied to a loss function defined by the structure of the neural network architecture as opposed to a training dataset. In particular, we reduce the popular combinatorial optimization problem of finding a maximum independent set to a neural network and employ a dataless training scheme to refine the parameters of the network such that those parameters yield the structure of interest. Additionally, we propose a universal graph reduction procedure to handle large-scale graphs. The reduction exploits community detection for graph partitioning and is applicable to any graph type and/or density. Experimental results on both real and synthetic graphs demonstrate that our proposed method performs on par or outperforms state-of-the-art learning-based methods in terms of the size of the found set without requiring any training data.

97 MATHEMATICS AND COMPUTING↗

A Modular and Transferable Reinforcement Learning Framework for the Fleet Rebalancing Problem

Mobility on demand (MoD) systems show great promise in realizing flexible and efficient urban transportation. However, significant technical challenges arise from operational decision making associated with MoD vehicle dispatch and fleet rebalancing. For this reason, operators tend to employ simplified algorithms that have been demonstrated to work well in a particular setting. To help bridge the gap between novel and existing methods, we propose a modular framework for fleet rebalancing based on model-free reinforcement learning (RL) that can leverage an existing dispatch method to minimize system cost. In particular, by treating dispatch as part of the environment dynamics, a centralized agent can learn to intermittently direct the dispatcher to reposition free vehicles and mitigate against fleet imbalance. We formulate RL state and action spaces as distributions over a grid partitioning of the operating area, making the framework scalable and avoiding the complexities associated with multiagent RL. Numerical experiments, using real-world trip and network data, demonstrate that RL reduces waiting time by 28% to 38% for the same-day evaluation, 17% to 44% for cross-day evaluation, and 22% to 25% for cross-season evaluation compared with no rebalancing scenarios. This approach has several distinct advantages over baseline methods including: improved system cost; high degree of adaptability to the selected dispatch method; and the ability to perform scale-invariant transfer learning between problem instances with similar vehicle and request distributions.

33 ADVANCED PROPULSION SYSTEMS↗

Fracture Network Prediction Using Physics-based Machine Learning Algorithms

In recent years, systematic CO2 injection into geological reservoirs across the U.S. has gained traction as a strategy to mitigate greenhouse gas emissions. This approach necessitates precise monitoring to ensure secure containment, minimize risks, and optimize storage management. Our study leverages machine learning (ML) techniques to advance the understanding of CO2 injection processes, focusing on the Illinois Basin. Over a three-year injection period, we analyzed microseismic data, identifying 19 temporal intervals with significant bottom-hole pressure changes. By partitioning microseismic events into these intervals and estimating b-values, we revealed over 100 clusters of events related to fracture initiation or reactivation. Advanced spatial analysis highlighted horizontally-oriented fractures along the NNW-SSE axis. This quantification of fracture networks informs dynamic injection scheduling, work-over strategies, and risk assessments, enhancing carbon capture, utilization, and storage (CCUS) operations. Additionally, our methodology offers valuable insights for oil and gas operations and geothermal development, supporting fracture-based monitoring and risk mitigation.

Kumar, Abhash↗

Parallel scalability of Hartree–Fock calculations

Quantum chemistry is increasingly performed using large cluster computers consisting of multiple interconnected nodes. For a fixed molecular problem, the efficiency of a calculation usually decreases as more nodes are used, due to the cost of communication between the nodes. This paper empirically investigates the parallel scalability of Hartree–Fock calculations. The construction of the Fock matrix and the density matrix calculation are analyzed separately. For the former, we use a parallelization of Fock matrix construction based on a static partitioning of work followed by a work stealing phase. For the latter, we use density matrix purification from the linear scaling methods literature, but without using sparsity. When using large numbers of nodes for moderately sized problems, density matrix computations are network-bandwidth bound, making purification methods potentially faster than eigendecomposition methods.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

Cation–Ligand Interactions Dictate Salt Partitioning and Diffusivity in Ligand-Functionalized Polymer Membranes

Membranes are an attractive alternative to current thermal separations due to their scalability and energy efficiency in desalinating water. Unfortunately, many of the conventional membrane materials available today are unable to differentiate between ionic solutes, especially alkali cations, compromising their use in ion–ion separations. Inspired by the ion-specific interactions exhibited by biological ion channels, recent research efforts have focused on synthesizing and characterizing new polymeric materials that incorporate ligands into polymer networks to bias solubility and/or diffusivity of one cationic species over another. Despite these efforts, little is known about the influence of incorporating ligands into polymer membranes on solubility and diffusivity of the complexing species. In this study, we first build a qualitative model of salt partitioning, diffusivity, and permeability in generic cation-complexing ligand-functionalized polymer membranes. Next, to validate our model and hypotheses, we perform atomistic molecular dynamics simulations of a 12-crown-4-functionalized membrane in the presence of alkali halide salts at low concentration. Generally, cation complexation enhances cation solubility but decreases diffusivity. Interestingly, the reduction in diffusivity is predicted to be larger than the enhancement in solubility for materials which operate by the mechanisms proposed in our physical picture, ultimately resulting in a reduction in the permeability of the selectively complexing ion.

36 MATERIALS SCIENCE↗

Evaluating the Potential of Cray Gemini Interconnect for PGAS Communication Runtime Systems

The Cray Gemini Interconnect has been recently introduced as the next generation network for building scalable multi-petascale supercomputers. The Cray XE6 systems, which use the Gemini Interconnect are becoming available with Message Passing Interface (MPI) and Partitioned Global Address Space (PGAS) Models such as as Global Arrays, Unified Parallel C, Co-Array Fortran and Cascade High Performance Language. These PGAS models use one-sided communication runtime systems such as MPI-Remote Memory Access, Aggregate Remote Memory Copy Interface and proprietary communication runtime systems. The primary objective of our work is to study the potential of Cray Gemini Interconnect by designing application specific micro-benchmarks using the DMAPP userspace library provided by the Gemini Interconnect. To meet this objective, we design micro-benchmarks to study the performance of simple communication primitives and application specific micro-benchmarks to understand the behavior of Gemini Interconnect at scale. In our experiments, the Gemini Interconnect can achieve a peak bandwidth of 6911 MB/s and a latency of 1us for get communication primitive. Scalability tests for atomic memory operations and shift communication operation up to 65536 processes shows the efficacy of the Gemini Interconnect.

Vishnu, Abhinav↗

Data-driven estimation of energy consumption for electric bus under real-world driving conditions

Reliable and accurate estimation of an electric bus’s instantaneous energy consumption is critical in evaluating energy impacts of planning and control of electric bus operations. In this study, we developed machine learning-based long short-term memory (LSTM) and artificial neural network (ANN) models to estimate 1 Hz energy consumption of electric buses based on continuous monitoring data of electric buses in Chattanooga, Tennessee, in 2019 and 2020. We propose a data-partitioning algorithm to separate energy charging and discharging modes before applying data-driven estimation models. Here, a K-fold cross-validation-based model selection process was conducted to identify the optimal model structure and input variables in terms of prediction accuracy. The estimation results show the predicted mean absolute percentage error rates of LSTM and ANN models were 3% and 5%, respectively. We compared the proposed models with existing models in the literature based on the same testing data to demonstrate the predictability of our models.

Artificial neural network↗