Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “System Identification”

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 1,639 records · Page 91

Analyzing Impact of Distributed PV Generation on Integrated Transmission & Distribution System Voltage Stability—A Graph Trace Analysis Based Approach

The use of a Graph Trace Analysis (GTA)-based power flow for analyzing the voltage stability of integrated Transmission and Distribution (T&D) networks is discussed in the context of distributed Photovoltaic (PV) generation. The voltage stability of lines and the load carrying capability of buses is analyzed at various PV penetration levels. It is shown that as the PV generation levels increase, an increase in the steady state voltage stability of the system is observed. Moreover, within certain regions of stability margin changes, changes in voltage stability margins of transmission lines are shown to be linearly related to changes in the loading of the lines. Two case studies are presented, where one case study involves a model with eight voltage levels and 784,000 nodes. In one case study, a voltage-stability heat map is used to demonstrate the identification of weak lines and buses.

14 SOLAR ENERGY↗

The International Experimental Thermal Hydraulic Systems database – TIETHYS: A new NEA validation tool

Nuclear reactor codes require validation with appropriate data representing the plant for specific scenarios. The thermal-hydraulic data is scattered in different locations and in different formats. Some of the data is in danger of being lost. A relational database is being developed to organize the international thermal hydraulic test data for various reactor concepts and different scenarios. At the reactor system level, that data is organized to include separate effect tests and integral effect tests for specific scenarios and corresponding phenomena. The database relies on the phenomena identification sections of expert developed PIRTs. The database will provide a summary of appropriate data, review of facility information, test description, instrumentation, references for the experimental data and some examples of application of the data for validation. The current database platform includes scenarios for PWR, BWR, VVER, and specific benchmarks for CFD modelling data and is to be expanded to include references for molten salt reactors. There are place holders for high temperature gas cooled reactors, CANDU and liquid metal reactors. This relational database is called The International Experimental Thermal Hydraulic Systems (TIETHYS) database and currently resides at Nuclear Energy Agency (NEA) of the OECD and is freely open to public access. Going forward the database will be extended to include additional links and data as they become available. https://www.oecd-nea.org/tiethysweb/

11 NUCLEAR FUEL CYCLE AND FUEL MATERIALS↗

RESOLVING THE sin(I) DEGENERACY IN LOW-MASS MULTI-PLANET SYSTEMS

Long-term orbital evolution of multi-planet systems under tidal dissipation often converges to a stationary state, known as the tidal fixed point. The fixed point is characterized by a lack of oscillations in the eccentricities and apsidal alignment among the orbits. Quantitatively, the nature of the fixed point is dictated by mutual interactions among the planets as well as non-Keplerian effects. We show that if a roughly coplanar system hosts a hot, sub-Saturn mass planet, and is tidally relaxed, separation of planet-planet interactions and non-Keplerian effects in the equations of motion leads to a direct determination of the true masses of the planets. Consequently, a 'snap-shot' observational determination of the orbital state resolves the sin(I) degeneracy and opens up a direct avenue toward identification of the true lowest-mass exoplanets detected. We present an approximate, as well as a general, mathematical framework for computation of the line-of-sight inclination of secular systems, and apply our models illustratively to the 61 Vir system. We conclude by discussing the observability of planetary systems to which our method is applicable and we set our analysis into a broader context by presenting a current summary of the various possibilities for determining the physical properties of planets from observations of their orbital states.

79 ASTRONOMY AND ASTROPHYSICS↗

High entropy alloys with hexagonal close-packed structure derived from thin film combinatorial approach

A thin film combinatorial synthesis approach was utilized to develop a compositional and structural library for the OsRuWCo alloy system. The survey of this alloy space produced 24 unique compositions and their corresponding crystal structures, which represent a phase map where amorphous and single-phase hexagonal close-packed alloys were identified. Based on the identification of a promising candidate film region, a new high entropy alloy based on OsRuWCo, but with non-equiatomic composition, was synthesized in bulk form. The alloy exhibited a single-phase hexagonal close-packed structure in the as-cast state. Three derivatives from this system were also developed by considering the heat of mixing, atomic size, and binary solubility. These new alloys are based on OsRuWCoIr and OsRuWCoFe, both of which exhibit single-phase hexagonal close-packed as-cast structures, as determined by X-ray diffraction and electron microscopy. Additionally, this large compositional space was utilized to evaluate conventional parameters that describe high entropy alloys (ΔS mix , ΔH mix , Ω, δ, and VEC), and these were found to elucidate the criteria for solid solution formation in this system. Additionally, trends illustrating the evolution from amorphous to crystalline phases are demonstrated and discussed. It is concluded that in this particular case, the dominant factors promoting crystallinity are thermodynamic or electronic, rather than geometric.

36 MATERIALS SCIENCE↗

The first full orbit of η Carinae seen by Fermi

The binary system η Carinae has completed its first 5.54 y orbit since the beginning of science operation of the Fermi Large Area Telescope (LAT). We are now able to investigate the high-energy γ-ray source at the position of η Carinae over its full orbital period. By this, we can address and confirm earlier predictions for temporal and spectral variability. Here, newer versions of the LAT datasets, instrument response functions and background models allow for a more accurate analysis. Therefore it is important to re-evaluate the previously analyzed time period along with the new data to further constrain location, spectral shape, and flux time history of the γ-ray source. As a result, we confirm earlier predictions of increasing flux values above 10 GeV toward the next periastron passage. For the most recent part of the data sample, flux values as high as those before the first periastron passage in 2008 are recorded. A comparison of spectral energy distributions around periastron and apastron passages reveals strong variation in the high-energy band. This is due to a second spectral component that is present only around periastron. In conclusion, improved spatial consistency with the γ-ray source at the position of η Carinae along with the confirmation of temporal variability above 10 GeV in conjunction with the orbital period strengthens the argument for unambiguous source identification. Spectral variability provides additional constraints for future modeling of the particle acceleration and γ-ray emission in colliding-wind binary systems.

79 ASTRONOMY AND ASTROPHYSICS↗

Association Mapping of Cell Wall Synthesis Regulatory Genes and Cell Wall Quality in Switchgrass

Inefficient conversion of biomass to biofuels is one of the main barriers for biofuel production from such materials. Approximately half of polysaccharides in biomass remain unused by typical biochemical conversion methods. Conversion efficiency is influenced by the composition and structure of cell walls of biomass. Grasses such as wheat, maize, and rice, as well as dedicated perennial bioenergy crops, like switchgrass, make up ~55% of biomass that can be produced in the United States. Grass cell walls have a different composition and patterning compared with dicotyledonous plants, including the well-studied model plant, Arabidopsis. This project identified genetic determinants of cell wall composition in grasses using both naturally occurring genetic variation of switchgrass and gene network reconstruction and functional assays in rice. In addition, the project linked functional data in rice and other species to switchgrass improvement efforts through curation of the most abundant class of regulators in the switchgrass genome. Characterizing natural diversity of switchgrass for variation in cell wall composition and properties, also known as quality, provides an unbiased avenue for identifying biologically viable diversity in switchgrass cell walls. To characterizing natural diversity, this project generated cell wall composition and enzymatic deconstruction data for ~450 genotypes of the Switchgrass Southern Association Collection (SSAC), a diverse collection composed of 36 switchgrass accessions from the southern U.S. distribution of switchgrass. Comparing these data with other measures of cell wall quality for the same samples demonstrated the complementary nature of the diverse characterization platforms now being used for biomass characterization. Association of the composition data with ~3.2K single nucleotide variant markers identified six significant single nucleotide variant markers co-associated with digestibility and another compositional trait. These markers might be used to select switchgrass genotypes with improved composition in breeding programs for biofuel and forage production. Because the SSAC continues to be characterized by collaborators in the bioenergy community, the data generated will be used to identify additional markers in higher resolution genotyping data to approach identifying the genes and alleles that cause natural variation in switchgrass cell wall quality. For example, these markers can be surveyed in the 2100-member Oklahoma Southern and Northern Lowland switchgrass collections that this project also characterized. An orthogonal approach to biodiversity studies, using comparative functional genomics permits systematic querying of how much regulatory information is likely to be transferable from dicots to grasses and use of accumulated functional genomics resources for better-characterized grass species, such as rice, itself a biomass source in global agriculture and in certain regions. The project generated and tested a number of specific hypotheses regarding cell wall transcription factors and enzymes of grasses. To aid identification of cell wall regulators, the project assembled a novel, highdepth and -quality gene association network using a general linearized model scoring system to combine rice gene network data. Using known or putative orthologs of Arabidopsis cell wall biosynthesis genes and regulators, the project pulled from this network a cell wall sub-network that includes 96 transcription factors. Reverse genetics of a co-ortholog of the Arabidopsis MYB61 transcription factor in rice revealed that this regulatory node has evolved the ability to regulate grass-specific cell wall synthesis enzymes. A transcription factor with such activity has not been previously characterized to our knowledge, representing a major conclusion of this work. Changes in gene expression in a protoplast-based assay demonstrated positive or negative roles in cell wall regulation for eleven other transcription factors from the rice gene network. Eight of fifteen (53%) of these have not previously been examined for this function. Some of these may represent novel grass-diverged cell wall regulators, while others are likely to have this function across angiosperms. A parallel effort of this project to expand knowledge of enzymes that have evolved to function in grass cell wall synthesis, revealed that a grass-diverged enzyme in rice, OsAT 5 , ferulates monolignols that are naturally incorporated into grass cell walls. This finding opens potential natural selection avenues for improving biomass composition for downstream processing by weak base pretreatment. Thus, this project has significantly expanded knowledge of cell wall synthesis and regulation in rice, information that can be used in reverse genetics and synthetic biology approaches to re-engineer cell walls for improved production of biofuel and high-value products. To lay the foundation for translating these results directly for switchgrass improvement, the project employed a comparative phylogenetic analysis of the major group of cell wall transcription factors that have been found to function in cell wall regulation, the R 2 R 3 MYBs. This analysis concluded that known cell wall regulators are largely conserved across switchgrass, rice, maize, poplar, and Arabidopsis. This interpretation is also largely consistent with the gene network analysis described above, though both approaches provide evidence that some co-orthologs of Arabidopsis regulators have diminished or increased in importance based on gene expression patterns. Also, several clades containing dicot cell wall regulators have expanded, consistent with the evolution of new cell wall regulators. This latter result is supported by functional analysis of the R 2 R 3 MYB protein SWAM 1 in a collaboration between this project and the DOE-funded group of Dr. S. Hazen at the University of Massachusettes. The curation of the switchgrass genome through this project provides specific targets for future engineering of switchgrass cell wall regulation and may also facilitate identification of regulators that underlie the molecular markers that are genetically linked to differences in cell wall quality. With the goal of spurring further research and technological developments in lignocellulosic biofuel production, this work has been communicated to the bioenergy and cell wall communities though various presentations and publications. To date, three manuscripts have been published, two others are near to publication, three others are in an advanced state, and two to four more are likely to be written based on analyses still in progress. In addition, project participants have presented thirteen posters and talks at regional, national, and international meetings about aspects of this project. In sum, the work supported by this funding has made and communicated significant progress in identifying the genes that grasses use for cell wall synthesis and regulation, information that will be used by project participants and others to improve the efficiency of conversion of lignocellulosic biomass to biofuels.

09 BIOMASS FUELS↗

Managing a Cray supercomputer as a git branch

Summary Modern software engineering/DevOps techniques applied to Cray XC Series systems management enable higher fidelity translation from test to production environments, reduce administration costs by avoiding duplicate efforts, and increase reliability. This can be done by using a git repository to track and manage all system configurations on the SMW(s) (System Management Workstation) and then adapt a gitflow‐like development methodology. Using git improves change management with peer review of changes, identification of who made a change, knowing why a change was made, facilitating easier reversion of bad changes, and enables a workflow using the ideas from gitflow with individual branches for preparing, maintaining, and recording releases. Using branches enables multiple contributors to work on different types of changes without getting unexpected changes from another feature area. We will describe a process similar to that used at NERSC (National Energy Research Scientific Computing Center) to extract/abstract configuration data, manage it using git, and then re‐apply changes to the SMW(s).

Jacobsen, Douglas M.↗

Characterization of morphology and hydration products of high-volume fly ash paste by monochromatic scanning x-ray micro-diffraction (μ-SXRD)

The present study focuses on identification and micro-structural characterization of the hydration products formed in high-volume fly ash (HVFA)/portland cement (PC) systems using monochromatic scanning x-ray micro-diffraction (μ-SXRD) and SEM-EDS. Pastes with up to 80% fly ash replacement were studied. Phase maps for HVFA samples using μ-SXRD patterns prove that μ-SXRD is an effective method to identify and visualize the distribution of phases in the matrix. μ-SXRD and SEM-EDS analysis shows that the C-S-H formed in HVFA system containing 50% or more of fly ash has a similar structure as C-S-H(I) with comparatively lower Ca/Si ratio than the one produced in PC system. Moreover, coexistence of C-S-H(I) and strätlingite is observed in the system containing 80% of fly ash, confirming that the amount of alumina and silicate phases provided by the fly ash is a major factor for the formation of C-S-H(I) and strätlingite in HVFA system. - Highlights: • High-volume fly ash (HVFA) paste was studied by scanning x-ray micro-diffraction. • Coexistence of C-S-H(I) and strätlingite in the HVFA system is clearly shown. • The distribution of minor phases in the HVFA system is shown. • Differences between inner and outer products of fly ash are observed by SEM-EDS.

36 MATERIALS SCIENCE↗

Identification of a High-Risk Group Among Patients With Oral Cavity Squamous Cell Carcinoma and pT1-2N0 Disease

Purpose: In the American Joint Committee on Cancer 2010 classification system, pT1-2N0 oral cavity squamous cell carcinoma (OSCC) is considered an early-stage cancer treatable with surgery alone (National Comprehensive Cancer Network 2010 guidelines). Our aim was to evaluate the feasibility of surgery alone for pT1-2N0 OSCC patients. Methods and Materials: Among 1279 previously untreated OSCC patients referred to our hospital between January 1996 and May 2008, we identified 457 consecutive patients with pT1-2N0 disease. All had radical tumor excision with neck dissection. A total of 387 patients showing pathologic margins greater than 4 mm and treated by surgery alone were included in the final analysis. All were followed up for at least 24 months after surgery or until death. The 5-year rates of control, distant metastasis, and survival were the main outcome measures. Results: The 5-year rates in the entire group of pT1-2N0 patients were as follows: local control, 91%; neck control, 92%; distant metastases, 1%; disease-free survival, 85%; disease-specific survival, 93%; and overall survival, 84%. Multivariate analysis identified poor differentiation and pathologic tumor depth of 4 mm or greater as independent risk factors for neck control, disease-free survival, and disease-specific survival. A scoring system using poor differentiation and tumor depth was formulated to define distinct prognostic groups. The presence of both poorly differentiated tumors and a tumor depth of 4 mm or greater resulted in significantly poorer 5-year neck control (p < 0.0001), disease-free (p < 0.0001), disease-specific (p < 0.0001), and overall survival (p = 0.0046) rates. Conclusion: The combination of poor differentiation and pathologic tumor depth of 4 mm or greater identified a subset of pT1-2N0 OSCC patients with poor outcome, who may have clinical benefit from postoperative adjuvant radiotherapy.

62 RADIOLOGY AND NUCLEAR MEDICINE↗

Finding a Needle in the Haystack: Identification of Functionally Important Minority Phases in an Operating Battery

The in-depth understanding of the minority phases’ roles in functional materials, e.g., batteries, is critical for optimizing the system performance and the operational efficiency. Although the visualization of battery electrode under operating conditions has been demonstrated, the development of advanced data-mining approaches is still needed in order to identify minority phases and to understand their functionalities. The present study uses nanoscale X-ray spectromicroscopy to study a functional LiCoO 2 /Li battery pouch cell. The data-mining approaches developed herein were used to search through over 10 million X-ray absorption spectra that cover more than 100 active cathode particles. Two particles with unanticipated chemical fingerprints were identified and further analyzed, providing direct evidence and valuable insight into the undesired side reactions involving the cation dissolution and precipitation as well as the local overlithiation-caused subparticle domain deactivation. As a result, the data-mining approach described in this work is widely applicable to many other structurally complex and chemically heterogeneous systems, in which the secondary/minority phases could critically affect the overall performance of the system, well beyond battery research.

25 ENERGY STORAGE↗

A primer on selecting grain boundary sets for comparison of interfacial fracture properties in molecular dynamics simulations

In this study, all grain boundaries are not equal in their predisposition for fracture due to the complex coupling between lattice geometry, interfacial structure, and mechanical properties. The ability to understand these relationships is crucial to engineer materials resilient to grain boundary fracture. Here, a methodology is presented to isolate the role of grain boundary structure on interfacial fracture properties, such as the tensile strength and work of separation, using atomistic simulations. Instead of constructing sets of grain boundary models within the misorientation/structure space by simply varying the misorientation angle around a fixed misorientation axis, the proposed method creates sets of grain boundary models by means of isocurves associated with important fracture-related properties of the adjoining lattices. Such properties may include anisotropic elastic moduli, the Schmid factor for primary slip, and the propensity for simultaneous slip on multiple slip systems. This approach eliminates the effect of lattice properties from the comparative analysis of interfacial fracture properties and thus enables the identification of structure-property relationships for grain boundaries. As an example, this methodology is implemented to study crack propagation along Ni grain boundaries. Segregated H is used as a means to emphasize differences in the selected grain boundary structures while keeping lattice properties fixed.

36 MATERIALS SCIENCE↗

Catalytic Oxidation of Propane over Palladium Alloyed with Gold: An Assessment of the Chemical and Intermediate Species

Understanding the catalytic oxidation of propane is important for developing catalysts not only for catalytic oxidation of hydrocarbons in emission systems but also for selective oxidation in the chemical processing industry. For palladium-based catalysts, little is known about the identification of the chemical or intermediate species involved in propane oxidation. We describe herein findings of an investigation of the catalytic oxidation of propane over supported palladium nanoalloys with different compositions of gold (Pd n Au 100–n ), focusing on probing the chemical or intermediate species on the catalysts in correlation with the bimetallic composition and the alloying phase structure. In addition to an enhanced catalytic activity, a strong dependence of the catalytic activity on the bimetallic composition was revealed, displaying an activity maximum at a Pd : Au ratio of 50 : 50 in terms of reaction temperature. This dependence is also reflected by its dependence on the thermochemical treatment conditions. While the activity for nanoalloys with n ~ 50 showed little change after the thermochemical treatment under oxygen, the activities for nanoalloys with n < 50 and n > 50 showed opposite trends. Importantly, this catalytic synergy is linked to the subtle differences of chemical and intermediate species which have been identified for the catalysts with different bimetallic compositions by in situ measurements using diffuse reflectance infrared Fourier transform spectroscopy (DRIFTS). For the catalytic oxidation of propane over the highly-active catalyst with a Pd : Au ratio of 50 : 50, the major species identified include acetate and bicarbonate, showing subtle differences in comparison with the identification of bicarbonate and formate for the catalyst with <50% Au (with a lower activity) and the absence of apparent species for the catalyst with >50% Au (activity is largely absent). The alloying of 50% Au in Pd is believed to increase the oxophilicity of Pd, which facilitates the first carbon–carbon bond cleavage and oxygenation of propane. Finally, the implications of the findings on the catalytic synergy of Pd alloyed with Au and the design of active Pd alloy catalysts are also discussed.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

Is there such a thing as a molecular organic alloy?

In inorganic materials, an alloy is a mixture of two or more substances that generally exhibits electronic and/or physical properties that differ from those of its constituents. In organic systems, the formation of a “molecular alloy” comprising mixtures of molecular organic materials has also been proposed. We test the validity of this concept via the study of the optoelectronic properties of a ternary system that has previously been identified to form a molecular acceptor alloy, namely a blend of a poly(3-hexylthiophene) (P3HT) donor, with two acceptors indene-C 60 bisadduct (ICBA) and phenyl-C 61 -butyric acid methyl ester (PC 61 BM) [R. A. Street, et al., J. Am. Chem. Soc., 2013, 135, 986–989]. Using photoelectron spectroscopy, we find that the ICBA:PC 61 BM blend shows the same highest occupied molecular orbital and exciton energies as that of ICBA, indicating the absence of a new exciton state in the blend. Furthermore, charge transfer state spectra of ternary blends are found to comprise a simple linear superposition of the corresponding binaries. From these results, no evidence of new, emergent electronic states is found to support the existence of a molecular alloy in this system. To our knowledge there is as yet no clear evidence of the existence of an alloy in any organic semiconductor system. Finally, we discuss the criteria that should be met by a molecular organic alloy and procedures needed for their unambiguous identification.

36 MATERIALS SCIENCE↗

IDENTIFICATION OF AN {sup 84}Sr-DEPLETED CARRIER IN PRIMITIVE METEORITES AND IMPLICATIONS FOR THERMAL PROCESSING IN THE SOLAR PROTOPLANETARY DISK

The existence of correlated nucleosynthetic heterogeneities in solar system reservoirs is now well demonstrated for numerous nuclides. However, it has proven difficult to discriminate between the two disparate processes that can explain such correlated variability: incomplete mixing of presolar material or secondary processing of a well-mixed disk. Using stepwise acid-leaching of the Ivuna CI-chondrite, we show that unlike other nuclides such as {sup 54}Cr and {sup 50}Ti, Sr-isotope variability is the result of a carrier depleted in {sup 84}Sr. The carrier is most likely presolar SiC, which is known to have both high Sr-concentrations relative to solar abundances and extremely depleted {sup 84}Sr compositions. Thus, variability in {sup 84}Sr in meteorites and their components can be attributed to varying contributions from presolar SiC. The observed {sup 84}Sr excesses in calcium-aluminum refractory inclusions (CAIs) suggest their formation from an SiC-free gaseous reservoir, whereas the {sup 84}Sr depletions present in differentiated meteorites require their formation from material with an increased concentration of SiC relative to CI chondrites. The presence of a positive correlation between {sup 84}Sr and {sup 54}Cr, despite being hosted in carriers of negative and positive anomalies, respectively, is not compatible with incomplete mixing of presolar material but instead suggests that the solar system's nucleosynthetic heterogeneity reflects selective thermal processing of dust. Based on vaporization experiments of SiC under nebular conditions, the lack of SiC material in the CAI-forming gas inferred from our data requires that the duration of thermal processing of dust resulting in the vaporization of CAI precursors was extremely short-lived, possibly lasting only hours to days.

79 ASTRONOMY AND ASTROPHYSICS↗

Mitigation of Cosmic Rays-Induced Errors in Superconducting Quantum Processors

Environmental radioactivity and cosmic-rays have recently been identified as a source of decoherence in super-conducting quantum bits (qubits). In particular, the absorption of cosmic-ray muons and gamma rays emitted by naturally occurring radioactive isotopes in the qubit substrate leads to correlated errors in superconducting quantum processors, posing significant challenges to quantum error correction. To enable quantum computing to scale, it is therefore necessary the devel-opment of mitigation strategies to prevent, or keep under control, error bursts due to particle impacts in the chip. While most environmental radioactive sources can be effectively suppressed using dedicated shielding, cosmic-ray muons, with their high penetration capability, can only be mitigated by moving the entire facility in a deep underground laboratory. This work explores the potential for developing a novel class of quantum processors equipped with an active veto system to protect superconducting-based quantum computers from the detrimental effects of atmospheric muons. Such a device would enable the identification of an atmospheric muon interaction within the processor and veto all operations performed during the occurrence of such an interaction. By demonstrating high detection efficiency and negligible dead time, we aim to establish that the future of quantum processors can be envisioned in above-around facilities.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗

Nonlinear Multiple Models Adaptive Secondary Voltage Control of Microgrids

This article proposes a model-free secondary voltage control (SVC) for microgrids (MG) using nonlinear multiple models adaptive control. Firstly, a linear robust adaptive controller is designed to guarantee the voltage stability in the bounded-input- bounded-output (BIBO) manner so as to meet the operation requirements of MGs. Secondly, a nonlinear adaptive controller is developed to improve the voltage tracking performance with the help of artificial neural networks (ANNs). A switching mechanism for coordinating such two controllers is designed to guarantee the closed-loop stability while achieving accurate voltage tracking. Here, by an online identification based on the input and output data of MGs, the proposed method does not resort to any apriori information of system model and primary control, thus exhibiting good robustness, ease of deployment and disturbance rejection.

17 WIND ENERGY↗

CEGANN: CRYSTAL EDGE GRAPH ATTENTION NEURAL NETWORK

SF-22-156 Machine learning (ML) models and applications in materials design and discovery typically involve the use of feature representations or descriptors followed by a learning algorithm that maps them to user desired properties of interest. Most popular mathematical formulation-based descriptors are not unique across atomic environments and suffer from transferability issues across different application domains and/or material classes. The CEGANN code provides a unified interface to facilitate material characterization across materials across multiple scales (from atomic to mesoscale) and diverse classes of materials ranging from metals oxides, non-metals, and even hierarchical materials such as zeolites and semi ordered materials such as mesophases. CEGANN implements a Graph Attention Network (GAT) type convolution architecture. The details of network architecture can be found in the paper https://doi.org/10.48550/arXiv.2207.10168. The software comes with pretrained examples and dataset for the classification of the following representative systems: (1) Structure-level representation such as space group (2) Structural dimensionality (e.g., bulk, 2D, clusters etc.) (3) Grain boundary identification (4) Nucleation and growth of a zeolite polymorph (5) Characterization of binary mesophases and their phase transitions (6) Growth of ice. The code is written in python programming language.

CHAN, HENRYT↗