Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “Edge computing”

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 1,135 records · Page 63

TransformerG2G: Adaptive time-stepping for learning temporal graph embeddings using transformers

Dynamic graph embedding has emerged as a very effective technique for addressing diverse temporal graph analytic tasks (i.e., link prediction, node classification, recommender systems, anomaly detection, and graph generation) in various applications. Such temporal graphs exhibit heterogeneous transient dynamics, varying time intervals, and highly evolving node features throughout their evolution. Hence, incorporating long-range dependencies from the historical graph context plays a crucial role in accurately learning their temporal dynamics. In this paper, we develop a graph embedding model with uncertainty quantification, TransformerG2G, by exploiting the advanced transformer encoder to first learn intermediate node representations from its current state (t) and previous context (over timestamps [t–1,t–l], l is the length of context). Moreover, we employ two projection layers to generate lower-dimensional multivariate Gaussian distributions as each node's latent embedding at timestamp t. We consider diverse benchmarks with varying levels of "novelty" as measured by the TEA (Temporal Edge Appearance) plots. Here, our experiments demonstrate that the proposed TransformerG2G model outperforms conventional multi-step methods and our prior work (DynG2G) in terms of both link prediction accuracy and computational efficiency, especially for high degree of novelty. Furthermore, the learned time-dependent attention weights across multiple graph snapshots reveal the development of an automatic adaptive time stepping enabled by the transformer. Importantly, by examining the attention weights, we can uncover temporal dependencies, identify influential elements, and gain insights into the complex interactions within the graph structure. For example, we identified a strong correlation between attention weights and node degree at the various stages of the graph topology evolution.

97 MATHEMATICS AND COMPUTING↗

Exploration, Prediction, and Experimental Verification of Structure and Optoelectronic Properties in I 2 -Eu-IV-X 4 (I = Li, Cu, Ag; IV = Si, Ge, Sn; X = S, Se) Chalcogenide Semiconductors

Recently, there has been extensive research into photovoltaic, thermoelectric, and nonlinear optical applications of chalcogenide semiconductors within the large set of defect-resistant I 2 -II-IV-X4 (I = Li, Cu, Ag; II = Ba, Sr, Eu, Pb; IV = Si, Ge, Sn; X = S, Se) compounds. Five Eu-including compounds have previously been reported within this family, but a comparative study of possible structures and electronic properties of all 18 Eu-based combinations is still absent. Herein, we use hybrid density functional theory to study rare-earth-including I 2 -II-IV-X 4 semiconductors with Eu on the II site, in order to further understand this family and test the geometric tolerance factor approach (reported in our previous work) as a tool for predicting potential stable structures. We investigate how the exchange mixing parameter of the HSE06 density functional, α, affects the energetic positions of electronic levels, especially of the localized f-electron orbitals near the band edges of the extended semiconductor structures, using literature photoemission and band gap data of EuS for comparison. Lowest-energy quaternary structure candidates, energy band structures, and densities of states are computationally predicted for all 18 materials. Based on its predicted photovoltaics-relevant band gap, the previously unknown compound Cu 2 EuSnSe 4 was selected and synthesized. Furthermore, the experimental structure, lattice parameters, and band gap of Cu 2 EuSnSe 4 are consistent with the computational predictions, confirming a 1.55 eV band gap.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

Local lattice distortions drive the transition of BaIrO 3 into a ferromagnetic insulator state

Using variable temperature total and resonant x-ray scattering at the K edge of Ir species, we study the “bad metal” to insulator transition in BaIrO 3 , a canonical third transition series oxide. The usage of advanced experimental techniques and large-scale computer modeling helps us show that, contrary to the widely accepted view, charge disproportionation leading to the formation of Ir-trimers with a different number of 5d valence electrons already exists at room temperature. The charge disbalance between the trimers does not evolve much with decreasing temperature while local lattice distortions do, suggesting that the latter and not the former make a key contribution to the emergence of the enigmatic ferromagnetic insulator state of BaIrO 3 . The conclusion is supported by DFT calculations based on unmodified experimental structure data. Our work calls for a reconsideration of the role of lattice distortions in determining the electronic properties of third transition series oxides. It also charts a path to assessing these properties on a realistic and not assumed crystal structure basis.

36 MATERIALS SCIENCE↗

How turbulence spreading improves power handling in quiescent high confinement fusion plasmas

Viable magnetic fusion devices necessitate combining good confinement with effective power flux handling. A major concern for ITER, and devices beyond, is the divertor heat load width, which sets peak boundary heat loads on the plasma-facing materials. Current estimates of the heat flux width are narrow for future reactors. Here, we demonstrate how pedestal turbulence can expand into, or entrain, the stable scrape-off-layer and so broaden the heat flux width beyond these neoclassical predictions. Employing combined theoretical, computational, and experimental approaches, we focus on quiescent high confinement discharges on the DIII-D tokamak, but the results are of broader significance. Our findings uncover common trends in the edge turbulence intensity flux, the pressure perturbation skewness, and the turbulence mixing length, which together determine the heat flux width. This research demonstrates the physics of scrape-off-layer broadening by turbulence and highlights the promise of a turbulent pedestal for successful core-edge integration in ITER and future fusion devices.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

Ionization energy shift of characteristic K x-ray lines from high-Z materials for plasma diagnostics

The energy of the characteristic x-rays emitted by high atomic number atoms in a plasma that contains energetic electrons depends on the atom's ionization. For tungsten, the ionization energy shift of the L-lines has recently been used to diagnose the plasma's ionization; the change in energy of a K-line has been measured for iridium and observed for ytterbium. Here, we present detailed computations of the ionization energy shift to K-lines of these and an additional element, dysprosium; for these atoms, some K-lines nearly coincide in energy with K-edges of slightly lower Z atoms so that a change in transmission behind a K-edge filter betrays a change in energy. The ionization energy shift of such high-energy K-lines may enable a unique diagnostic when the plasma is inside an otherwise opaque enclosure such as hohlraums used on the National Ignition Facility.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

The effect of spin-orbit coupling in band structure of few-layer graphene

Topological insulators are electronic materials that have a bulk band gap like an ordinary insulator but have protected conducting states on their edge or surface. This can be happened due to spin-orbit coupling and time-reversal symmetry. Moreover, the edge current flows through their edge or surface depends on its spin orientation and also it is robust against non-magnetic impurities. Therefore, topological insulators are predicted to be useful ranging from spintronics to quantum computation. Graphene was first predicted to be the precursor of topological insulator by Kane-Mele. They developed a Hamiltonian model to describe the gap opening in graphene. In this work, we investigate the band structure of few-layer graphene by using this model with analytical approach. The results of our calculations show that the gap opening occurs at K and K’ point, not only in single layer, but also in bilayer and trilayer graphene.

77 NANOSCIENCE AND NANOTECHNOLOGY↗

THE INFRARED SPECTRA OF VERY LARGE IRREGULAR POLYCYCLIC AROMATIC HYDROCARBONS (PAHs): OBSERVATIONAL PROBES OF ASTRONOMICAL PAH GEOMETRY, SIZE, AND CHARGE

The mid-infrared (IR) spectra of six large, irregular polycyclic aromatic hydrocarbons (PAHs) with formulae (C{sub 84}H{sub 24}-C{sub 120}H{sub 36}) have been computed using density functional theory (DFT). Trends in the dominant band positions and intensities are compared to those of large, compact PAHs as a function of geometry, size, and charge. Irregular edge moieties that are common in terrestrial PAHs, such as bay regions and rings with quartet hydrogens, are shown to be uncommon in astronomical PAHs. As for all PAHs comprised solely of C and H reported to date, mid-IR emission from irregular PAHs fails to produce a strong CC{sub str} band at 6.2 {mu}m, the position characteristic of the important, class A astronomical PAH spectra. Earlier studies showed that inclusion of nitrogen within a PAH shifts this to 6.2 {mu}m for PAH cations. Here we show that this band shifts to 6.3 {mu}m in nitrogenated PAH anions, close to the position of the CC stretch in class B astronomical PAH spectra. Thus, nitrogenated PAHs may be important in all sources and the peak position of the CC stretch near 6.2 {mu}m appears to directly reflect the PAH cation to anion ratio. Large irregular PAHs exhibit features at 7.8 {mu}m but lack them near 8.6 {mu}m. Hence, the 7.7 {mu}m astronomical feature is produced by a mixture of small and large PAHs while the 8.6 {mu}m band can only be produced by large compact PAHs. As with the CC{sub str}, the position and profile of these bands reflect the PAH cation to anion ratio.

74 ATOMIC AND MOLECULAR PHYSICS↗

THE MISSING CAVITIES IN THE SEEDS POLARIZED SCATTERED LIGHT IMAGES OF TRANSITIONAL PROTOPLANETARY DISKS: A GENERIC DISK MODEL

Transitional circumstellar disks around young stellar objects have a distinctive infrared deficit around 10 {mu}m in their spectral energy distributions, recently measured by the Spitzer Infrared Spectrograph (IRS), suggesting dust depletion in the inner regions. These disks have been confirmed to have giant central cavities by imaging of the submillimeter continuum emission using the Submillimeter Array (SMA). However, the polarized near-infrared scattered light images for most objects in a systematic IRS/SMA cross sample, obtained by HiCIAO on the Subaru telescope, show no evidence for the cavity, in clear contrast with SMA and Spitzer observations. Radiative transfer modeling indicates that many of these scattered light images are consistent with a smooth spatial distribution for {mu}m-sized grains, with little discontinuity in the surface density of the {mu}m-sized grains at the cavity edge. Here we present a generic disk model that can simultaneously account for the general features in IRS, SMA, and Subaru observations. Particularly, the scattered light images for this model are computed, which agree with the general trend seen in Subaru data. Decoupling between the spatial distributions of the {mu}m-sized dust and mm-sized dust inside the cavity is suggested by the model, which, if confirmed, necessitates a mechanism, such as dust filtration, for differentiating the small and big dust in the cavity clearing process. Our model also suggests an inwardly increasing gas-to-dust ratio in the inner disk, and different spatial distributions for the small dust inside and outside the cavity, echoing the predictions in grain coagulation and growth models.

79 ASTRONOMY AND ASTROPHYSICS↗

Modeling, analysis, and code/data validation of DIII-D tokamak divertor experiments on ELM and non-ELM plasma tungsten sputtering erosion

Abstract We analyzed recent DIII-D tokamak tungsten divertor probe experiments using advanced, coupled, sputter erosion/redeposition, plasma, and surface response code packages. Modeling is done for ELMing H-mode, and L-mode plasmas, impinging on various size tungsten deposits on Divertor Material Evaluation System (DiMES) carbon probes. The simulations compute 3D, full kinetic, sub-gyromotion, impurity sputtering and transport, including changes in tungsten surface composition and response due to mixed deuterium and carbon ions irradiation. Per our analysis, ELM (edge localized mode) plasma sputtering in DIII-D mostly involves free-streaming high energy (∼500–1000 eV) D + and C +6 ions, with high near-surface plasma density. L-Mode sputtering is due to impurity sputtering (C, W) only, with lower density. All cases show complete redeposition of tungsten on the divertor, with significant redeposition on the tungsten spots themselves, and low self-sputtering. Comparison of ELM plasma gross tungsten erosion simulation results with in-situ spectroscopic data is good, as are code/data comparisons of net erosion using post-exposure Rutherford backscattering (RBS) data for the L-mode probes. The analysis, extrapolated to a full tungsten divertor, implies low net erosion and negligible plasma contamination from sputtering. These results support the use of high-Z plasma facing surfaces in ITER and beyond.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

EuCd 2 ⁢As 2 : A Magnetic Semiconductor

EuCd 2 ⁢As 2 is now widely accepted as a topological semimetal in which a Weyl phase is induced by an external magnetic field. Here, we challenge this view through firm experimental evidence using a combination of electronic transport, optical spectroscopy, and excited-state photoemission spectroscopy. We show that the EuCd 2 ⁢As 2 is in fact a semiconductor with a gap of 0.77 eV. We show that the externally applied magnetic field has a profound impact on the electronic band structure of this system. This is manifested by a huge decrease of the observed band gap, as large as 125 meV at 2 T, and, consequently, by a giant redshift of the interband absorption edge. However, the semiconductor nature of the material remains preserved. EuCd 2 ⁢As 2 is therefore a magnetic semiconductor rather than a Dirac or Weyl semimetal, as suggested by ab initio computations carried out within the local spin-density approximation.

36 MATERIALS SCIENCE↗

MDDC Multi-Length Scale Data Architecture Contribution Report – PNNL, INL, ANL, LANL and ORNL

This report offers a comprehensive view of data streams currently generated at Pacific Northwest National Laboratory, Idaho National Laboratory, Argonne National Laboratory, Los Alamos National Laboratory, and Oak Ridge National Laboratory set to integrate into the evolving Multi-Dimensional Data Correlation framework at Oak Ridge National Laboratory. Developed by the Advanced Materials and Manufacturing Technologies program, the Multi-Dimensional Data Correlation framework serves as a cutting-edge software to manage data relevant to advanced manufacturing and material behavior in advanced reactors. The report defines data streams, highlights their generation methods and visualization methods both for experimental and computational aspects relevant to the Advanced Materials and Manufacturing Technologies project. A logical next step for this work is to integrate the MDDC framework into PNNL’s, INL’s, ANL’s, LANL’s and ORNL’s fabrication, experimentation, and modelling workflows. This would require setting up the MDDC framework at PNNL, INL, ANL, and LANL and integrating it into the data collection and storage for these different activities.

36 MATERIALS SCIENCE↗

Block-structured adaptive mesh refinement - theory, implementation and application

Structured adaptive mesh refinement (SAMR) techniques can enable cutting-edge simulations of problems governed by conservation laws. Focusing on the strictly hyperbolic case, these notes explain all algorithmic and mathematical details of a technically relevant implementation tailored for distributed memory computers. An overview of the background of commonly used finite volume discretizations for gas dynamics is included and typical benchmarks to quantify accuracy and performance of the dynamically adaptive code are discussed. Large-scale simulations of shock-induced realistic combustion in non-Cartesian geometry and shock-driven fluid-structure interaction with fully coupled dynamic boundary motion demonstrate the applicability of the discussed techniques for complex scenarios.

99 GENERAL AND MISCELLANEOUS↗

Techno Economic Model

The Technoeconomic model is a computational model of a lignocellulosic biorefinery that can be used by industry to establish benchmarks of performance and risk-benefit analysis in order to assess the potential impact of cutting edge technologies. The model can be used to evaluate, guide, and optimize research efforts, biorefinery design, and process operation. The model will help to reduce the risk of commercial investment and development of biorefineries and help steer future research to those parts of the refining process in need of further developments for biofuels to be cost competitive. We have now aded modules for the following sections: feed handling, pretreatment, fermentation, product and water recovery, waste treatment, and steam/electricity generation. We have incorporated a kinetic model for microorganism growth and production of ethanol, inclouding toxin inhibition. For example, the feed handling section incorporates information regarding feedstock transport distance-dependent costs. The steam and electricity generation section now includes a turbogenerator that supplies power to be used by other unit operations and contains equations for efficiency calculations.

Source record↗

SDN-NGenIA, a Software Defined Next Generation Integrated Architecture for HEP and Data Intensive Science

This project aims to enable the LHC and other leading programs in high energy physics and other global science domains funded by the DOE to operate with a new level of efficiency and control, through the development of a Next Generation Integrated Architecture (NGenIA) based on intelligent software defined network (SDN)-driven network systems coupled to high throughput applications. While the initial focus is on the challenging LHC use case, the systems and products being developed are general, and apply to many fields of data intensive science ranging from astrophysical sky surveys to bioinformatics and earth observation, as well as other organizations facing the challenges of extracting knowledge from distributed multi-petabyte data stores. A central concept in this development program is a new paradigm ‘consistent network operations’ among widely distributed computing and storage facilities, where stable high throughput flows, at set rates, cross load-balanced network paths, up to flexible high water marks that are adjusted in real time to accommodate other network traffic. The large smooth flows are launched and managed by SDN services that act in concert with the experiments’ site-resident data distribution and management systems, to meet the expanding needs of the science programs. The technical goals include the construction of autonomous, intelligent site-resident services that interact dynamically with network-resident services, and with the science programs’ principal data distribution and management tools, to request or command network resources in support of high throughput petascale to exascale workflow Specific work items include: 1. Developing compact Data Transfer Nodes (DTNs) with auto-tuning functions that support data transfer rates in the 100 Gbps to the 1 Tbps range when used with high throughput data transfer applications 2. Deep site orchestration among virtualized clusters, storage subsystems and subnets to successfully coschedule CPU, storage and network resources 3. Science-program designed site architectures, operational modes, and policy and resource usage priorities, adjudicated across multiple network domains and virtual organizations, using the orchestration functions and methods 4. Seamlessly extending end-to-end operation across both extra-site and intra-site boundaries through the use of Open vSwitch (OVS), FireQoS + next generation Science DMZs 5. Novel methods of system integration that enable granular control of extreme scale long distance transfers through flow matching of scattered source-destination address pairs to multi-domain dynamic circuits 6. Funneling massive sets of streams to DTNs at the site edge hosting petascale buffer pools configured for flows of 100 Gbps and up, exploiting state of the art data transfers 7. Adaptive scheduling based on pervasive end-to-end monitoring, including DTN or compute-node resident agents providing comprehensive end-system profiling 8. Developing unsupervised and supervised machine learning and modeling methods to optimize the workflow involving terabyte to multi-petabyte datasets.

47 OTHER INSTRUMENTATION↗

Manifold projection image segmentation for nano-XANES imaging

As spectral imaging techniques are becoming more prominent in science, advanced image segmentation algorithms are required to identify appropriate domains in these images. We present a version of image segmentation called manifold projection image segmentation (MPIS) that is generally applicable to a broad range of systems without the need for training because MPIS uses unsupervised machine learning with a few physically motivated hyperparameters. We apply MPIS to nanoscale x-ray absorption near edge structure (XANES) imaging, where XANES spectra are collected with nanometer spatial resolution. We show the superiority of manifold projection over linear transformations, such as the commonly used principal component analysis (PCA). Moreover, MPIS maintains accuracy while reducing computation time and sensitivity to noise compared to the standard nano-XANES imaging analysis procedure. Finally, we demonstrate how multimodal information, such as x-ray fluorescence data and spatial location of pixels, can be incorporated into the MPIS framework. We propose that MPIS is adaptable for any spectral imaging technique, including scanning transmission x-ray microscopy, where the length scale of domains is larger than the resolution of the experiment.

36 MATERIALS SCIENCE↗

Nondestructive geochemical characterization of fossil hominin taphonomy and burial history

To date, only three Homo habilis specimens have been discovered that have associated craniodental and postcranial elements, providing a limited fossil record of the ontogeny and morphology of early members of the genus Homo. Recently, a nearly complete dentition, likely attributable to H. habilis, was discovered and excavated from early Pleistocene-age fluvial-lacustrine sediments of the upper Burgi Member of the Koobi Fora Formation at site F25787 in Area 13, near Ileret, Kenya. On the surface less than 15 m away, at site F25966, postcranial elements were found, which, if from the same individual as the nearby dentition, would represent the fourth associated craniodental and postcranial assemblage of this species. We developed a geochemical taphonomic history of these ca. 2 Ma hominin fossils using nondestructive X-ray based microanalytical tools (synchrotron and benchtop X-ray fluorescence chemical imaging and micro- and nano-computed tomography volumetric reconstruction), bulk analyses of sediments and paleosols at the excavation sites, and sedimentologic and stratigraphic observations. We integrate the chemical and physical taphonomic histories to test whether teeth (excavated in situ) and postcranial bones (eroded onto the outcrop surface) derive from a single individual. Minor differences in taphonomic history are attributable to the different biomineral properties of the dental and osseous components and to differences in physical damage during early post-mortem scavenging, dispersal, and burial in adjacent depositional settings. Microscale geochemical mapping enabled the temporal ordination of chemical and physical events in the specimens’ chemical taphonomic histories. Specifically, authigenic Fe- and K-bearing clays and Y, U, and Sr uptake occurred in post-burial fractures in bones and were also incorporated pervasively throughout dentin in teeth. Barite mineralization occurred along the latest fractures in both materials, and as a coating on tooth roots. The stratigraphic, taphonomic, and geochemical evidence supports the interpretation that the hominin fossils represent a single individual. Finally, successful application of these nondestructive sample characterization methods demonstrates capabilities for thorough interrogation of the taphonomic histories of other potential hominin fossil associations, enabling more robust and accurate palaeontologic constraints using those relationships.

36 MATERIALS SCIENCE↗

Less is More: Bigger Data from Compressive Measurements

Compressive sensing approaches are beginning to take hold in (scanning) transmission electron microscopy (S/TEM) [1,2,3]. Compressive sensing is a mathematical theory about acquiring signals in a compressed form (measurements) and the probability of recovering the original signal by solving an inverse problem [4]. The inverse problem is underdetermined (more unknowns than measurements), so it is not obvious that recovery is possible. Compression is achieved by taking inner products of the signal with measurement weight vectors. Both Gaussian random weights and Bernoulli (0,1) random weights form a large class of measurement vectors for which recovery is possible. The measurements can also be designed through an optimization process. The key insight for electron microscopists is that compressive sensing can be used to increase acquisition speed and reduce dose. Building on work initially developed for optical cameras, this new paradigm will allow electron microscopists to solve more problems in the engineering and life sciences. We will be collecting orders of magnitude more data than previously possible. The reason that we will have more data is because we will have increased temporal/spatial/spectral sampling rates, and we will be able ability to interrogate larger classes of samples that were previously too beam sensitive to survive the experiment. For example consider an in-situ experiment that takes 1 minute. With traditional sensing, we might collect 5 images per second for a total of 300 images. With compressive sensing, each of those 300 images can be expanded into 10 more images, making the collection rate 50 images per second, and the decompressed data a total of 3000 images [3]. But, what are the implications, in terms of data, for this new methodology? Acquisition of compressed data will require downstream reconstruction to be useful. The reconstructed data will be much larger than traditional data, we will need space to store the reconstructions during analysis, and the computational demands for analysis will be higher. Moreover, there will be time costs associated with reconstruction. Deep learning [5] is an approach to address these problems. Deep learning is a hierarchical approach to find useful (for a particular task) representations of data. Each layer of the hierarchy is intended to represent higher levels of abstraction. For example, a deep model of faces might have sinusoids, edges and gradients in the first layer; eyes, noses, and mouths in the second layer, and faces in the third layer. There has been significant effort recently in deep learning algorithms for tasks beyond image classification such as compressive reconstruction [6] and image segmentation [7]. A drawback of deep learning, however, is that training the model requires large datasets and dedicated computational resources (to reduce training time to a few days). A second issue is that deep learning is not user-friendly and the meaning behind the results is usually not interpretable. We have shown it is possible to reduce the data set size while maintaining model quality [8] and developed interpretable models for image classification [9], but the demands are still significant. The key to addressing these problems is to NOT reconstruct the data. Instead, we should design computational sensors that give answers to specific problems. A simple version of this idea is compressive classification [10], where the goal is to classify signal type from a small number of compressed measurements. Classification is a much simpler problem than reconstruction, so 1) much fewer measurements will be necessary, and 2) these measurements will probably not be useful for reconstruction. Other simple examples of computational sensing include determining object volume or the number of objects present in the field of view [11].

Stevens, Andrew↗

Discrete Green’s functions and spectral graph theory for computationally efficient thermal modeling

Here, this work concerns solutions of the heat equation with the spectral graph method, for which the temperature is defined at discrete points in the domain and the spatial relationship among the points is described by a graph. The heat equation on the graph is solved using matrix techniques involving the eigenvectors and eigenvalues of the Laplacian matrix. The spectral graph approach precludes the computationally intensive meshing and numerous time-integration steps of the finite element method. In the present work, the spectral graph method is extended to include heat loss at the boundaries with a generalized boundary condition, and physics-based edge weights are introduced which simplify the calibration process. From this approach a discrete Green’s function is defined which allows for solutions under a variety of heating conditions including: space-varying initial conditions; time-and-space varying internal heating; and, time-and-space-varying heating at boundaries of type 1 (Dirichlet), type 2 (Neumann) and type 3 (Robin). Results are provided for benchmark heat transfer problems in one spatial dimension and in three spatial dimensions, and verification is provided by comparison with exact analytical solutions and finite difference solutions. The spectral graph method converges within 0.4% error of the analytical solution. The practical utility of the approach is demonstrated by thermal simulation of a multilayer additive manufacturing process. The spectral graph results are compared to experimentally-obtained temperature data for two metal parts, with error less than 5% of the experimental measurements, with computation time less than one minute on a desktop computer.

36 MATERIALS SCIENCE↗