Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “data structures design and analysis”

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 865 records · Page 48
All sourcesNASA NTRSDOE OSTI
NASA NTRS1985

System technology analysis of aeroassisted orbital transfer vehicles. Moderate lift/drag (0.75-1.5). Volume 1A, part 1: Executive summary, phase 1

Activities and significant results of Phase 1 of a study to access aeroassisted orbit transfer vehicle (AOTV) system technology are summarized. Broad concept evaluations were performed and the technology requirements and sensitivities for ground based AOTV's over a range of vehicle hypersonic lift/drag (L/D) from 0.75 to 1.5 were systematically identified and assessed. The four major task areas included systems analysis, system/subsystem trades, technology payoff assessment and plan, and cost analysis. Findings indicate that substantial performance improvements and hence cost benefit can be obtained by developing enhanced technologies such as: (1) low thrust advanced expander LOX-hydrogen engines with specific impulse of 480 to 490 sec; (2) reducing the external thermal protection system weight and increasing the maximum allowable bond/structure temperature; and (3) reducing the structural shell weight by improving the quality of the design allowable data, or use of advanced structural materials. Results also show that use of mid L/D AOTV's provide significant aerodynamic plane change capability and control authority over trajectory dispersions and off-nominal atmospheres.

Source record↗
NASA NTRS2019

Photogrammetry-Based Analysis of the On-Orbit Structural Dynamics of the Roll-Out Solar Array

The Roll-Out Solar Array (ROSA) flight experiment was launched to the International Space Station (ISS) on June 3rd, 2017. ROSA is an innovative, lightweight solar array with a flexible substrate that makes use of the stored strain energy in its composite structural members to provide deployment without the use of motors. This paper will discuss the results of various structural dynamics experiments conducted on the ISS during the weeks following launch. Data gathered from instrumentation on the solar array wing during the experiments was previously compared with pre-flight predictions from two different Finite Element Modeling (FEM) efforts. In this paper, data generated from photogrammetry is compared with accelerometer data and used to extend previous conclusions. Whereas previous analyses were only able to track the accelerations of six discrete points on the structure and photovoltaic (PV) blanket of ROSA, the photogrammetry analysis makes available displacements for dozens of points distributed throughout the array. This larger data set makes it possible to compare higher-order PV blanket modes with FEM predictions, in addition to verifying conclusions reached using accelerometer data. The goal in this effort was to better understand the performance of ROSA and to improve modeling efforts for future designs of similar solar arrays.

Chamberlain, Matthew K.↗
NASA NTRS2002

CAPRI: Using a Geometric Foundation for Computational Analysis and Design

CAPRI (Computational Analysis Programming Interface) is a software development tool intended to make computerized design, simulation and analysis faster and more efficient. The computational steps traditionally taken for most engineering analysis (Computational Fluid Dynamics (CFD), structural analysis, etc.) are: Surface Generation, usually by employing a Computer Aided Design (CAD) system; Grid Generation, preparing the volume for the simulation; Flow Solver, producing the results at the specified operational point; Post-processing Visualization, interactively attempting to understand the results. It should be noted that the structures problem is more tractable than CFD; there are fewer mesh topologies used and the grids are not as fine (this problem space does not have the length scaling issues of fluids). For CFD, these steps have worked well in the past for simple steady-state simulations at the expense of much user interaction. The data was transmitted between phases via files. In most cases, the output from a CAD system could go IGES files. The output from Grid Generators and Solvers do not really have standards though there are a couple of file formats that can be used for a subset of the gridding (i.e. PLOT3D) data formats and the upcoming CGNS). The user would have to patch up the data or translate from one format to another to move to the next step. Sometimes this could take days. Instead of the serial approach to analysis, CAPRI takes a geometry centric approach. CAPRI is a software building tool-kit that refers to two ideas: (1) A simplified, object-oriented, hierarchical view of a solid part integrating both geometry and topology definitions, and (2) programming access to this part or assembly and any attached data. The connection to the geometry is made through an Application Programming Interface (API) and not a file system.

Haimes, Robert↗
DOE OSTI2025

EV-ELM (Electric Vehicle Policies with the Energy Language Model) [SWR-25-156]

Electric Vehicle Policies with the Energy Language Model (EV-ELM) leverages previous work using Large Language Models (LLMs) to find, download, and parse policy information related to energy infrastructure. In this application, we use LLMs to find policy documents related to the permitting and installation of electric vehicle charging infrastructure. This software contains the code to find, download, and parse these documents, while a related data record in the Open Energy Data Initiative (OEDI) will include the resulting output dataset that can be used for downstream analysis. The EV-ELM repository contains code for the EV-ELM project, which focuses on retrieving and processing EV permitting processes using large language models. The project is composed of two pipelines: (1) a web scraping pipeline for discovering and downloading EV permitting documents, and (2) a document parsing and extraction pipeline that processes the downloaded files to produce structured data. The web scraping pipeline is designed to extract relevant information from various websites, while the document parsing pipeline processes and analyzes the extracted documents to derive meaningful insights. Both pipelines depend on the NLR elm repository, which provides essential tools and functionalities for handling and processing the data. The web scraping pipeline is a modified version of the ordinance_gpt example within the elm repository. It has been adapted to fit the specific requirements of the EV-ELM project, ensuring that it effectively captures and processes the necessary information related to EV permitting.

Olson, Reid [National Laboratory of the Rockies (N↗
DOE OSTI2020

Experimentally Validated Computational Modeling of Creep and Creep-Cracking for Nuclear Concrete Structures

In a Nuclear Power Plant, one of the most important components is the concrete nuclear reactor cavity, which serves both a structural and protective function as the biological radiation shield. Given that creep has been identified as a major knowledge gap in the assessment of nuclear structures (NUREG/CR-7153), this work helps to further the understanding of creep behavior of massive concrete containment structures for decades to enable safe and long-term operation of these facilities. This project has developed a robust, experimentally validated model to predict creep in nuclear concrete structures for up 60 years using short-term creep data thereby enabling a longer service life of critical facilities and early detection of structural failure. The work presented in this report is a pairing of computational and experimental methods. For the first time, the time temperature superposition (TTS) principle was successfully used to generate a uniaxial creep compliance master curve to predict mortar creep response for up to 22,500 days (nearly 60 years) at a reference temperature of 20°C. These data were used as input into finite element analysis (FEA) codes that use highly realistic random, 3D concrete microstructures from reconstructed coarse limestone aggregates. Finite element analysis performed provides the ability to quickly upscale mortar viscoelastic behavior to long-term concrete creep/relaxation data. A master creep compliance curve, constructed from the TTS principle, spanning 27 years, was used to validate two and a half decades of simulated concrete creep. Concurrently, three different simulated wall specimens were designed to mimic the behavior of post-tensioned concrete nuclear containment facility vessel walls over time as a result of concrete creep. The specimens were designed with different thicknesses, transverse and longitudinal reinforcement ratios, and level of post-tensioning stress. Each specimen contained various instrumentation to measure internal concrete temperature, concrete strain, and post-tensioning strain hourly for over 3 years. The concrete creep model developed in this project, based on the FEA concrete simulations, was applied to simulate the structural-scale experiments of prestressed concrete walls conducted in this project using the Grizzly code. These models can represent the effects of reinforcing and prestressing. Although there are some discrepancies with the experimental data, the model can predict the overall trends of the creep response in these experiments. One of these experimental models was also applied to an extended time to demonstrate how the findings from this study can be used to predict the behavior of actual structures of interest that have been in service for extended periods of time.

22 GENERAL STUDIES OF NUCLEAR REACTORS↗
NASA NTRS1986

The 15 meter hoop-column antenna dynamics: Test and analysis

A 15 meter model of the hoop-column antenna concept has been vibration tested for model characterization and analytical model verification. Linear finite element analysis predicted the global vibration frequencies accurately. Good agreement between analysis and test data was obtained only after the analytical model was refined using static test data. As structures become more flexible, structural properties determined from static data become more accurate and should be used to update analytical models. Global vibration modes are not significantly affected by the surface mesh which permits simplified analytical models to be used for prediction of global behavior. These reduced models are believed sufficient for preliminary design and controls simulations where only global behavior is desired. The mesh modes were highly damped due to the knit mesh used for the reflector surface. These modes were also highly coupled and very difficult to measure in the laboratory. The inability to fully characterize the antenna mesh modes in the laboratory indicates robust methods for active surface vibration suppression will be needed. Fortunately, the surface mesh exhibits high passive damping which should be beneficial to active control systems.

Belvin, W. Keith↗
NASA NTRS1992

Commercial turbofan engine exhaust nozzle flow analyses using PAB3D

Recent developments of a three-dimensional (PAB3D) code have paved the way for a computational investigation of complex aircraft aerodynamic components. The PAB3D code was developed for solving the simplified Reynolds Averaged Navier-Stokes equations in a three-dimensional multiblock/multizone structured mesh domain. The present analysis was applied to commercial turbofan exhaust flow systems. Solution sensitivity to grid density is presented. Laminar flow solutions were developed for all grids and two-equation k-epsilon solutions were developed for selected grids. Static pressure distributions, mass flow and thrust quantities were calculated for on-design engine operating conditions. Good agreement between predicted surface static pressures and experimental data was observed at different locations. Mass flow was predicted within 0.2 percent of experimental data. Thrust forces were typically within 0.4 percent of experimental data.

Abdol-Hamid, Khaled S.↗
NASA NTRS1992

Design definition of the Laser Atmospheric Wind Sounder (LAWS), phase 2. Volume 2: Final report

Lockheed personnel, along with team member subcontractors and consultants, have performed a preliminary design for the LAWS Instrument. Breadboarding and testing of a LAWS class laser have also been performed. These efforts have demonstrated that LAWS is a feasible Instrument and can be developed with existing state-of-the-art technology. Only a commitment to fund the instrument development and deployment is required to place LAWS in orbit and obtain the anticipated science and operational forecasting benefits. The LAWS Science Team was selected in 1988-89 as were the competing LAWS phase 1/2 contractor teams. The LAWS Science Team developed requirements for the LAWS Instrument, and the NASA/LAWS project office defined launch vehicle and platform design constraints. From these requirements and constraints, the lockheed team developed LAWS Instrument concepts and configurations. A system designed to meet these requirements and constraints is outlined. The LAWS primary subsystem and interfaces - laser, optical, and receiver/processor - required to assemble a lidar are identified. Also identified are the support subsystems required for the lidar to function from space: structures and mechanical, thermal, electrical, and command and data management. The Lockheed team has developed a preliminary design of a LAWS Instrument System consisting of these subsystems and interfaces which will meet the requirements and objectives of the Science Team. This final report provides a summary of the systems engineering analyses and trades of the LAWS. Summaries of the configuration, preliminary designs of the subsystems, testing recommendations, and performance analysis are presented. Environmental considerations associated with deployment of LAWS are discussed. Finally, the successful LAWS laser breadboard effort is discussed along with the requirements and test results.

Wilson, D. J.↗
NASA NTRS1996

Multidisciplinary Design Of Hot Composite Structures

Unified computer code developed to implement multidisciplinary approach to design and analysis of composite-material structures that must withstand high temperatures. Code modular: includes executive module communicating with and coordinating other modules performing calculations pertaining to traditionally separate disciplines like those of acoustics, structural vibrations, structural loads, and thermal effects. Essential feature, finite-element numerical simulation of relevant physical phenomena according to applicable disciplines. Same finite-element mesh used in thermal, vibrational, and structural analyses; minimizing data-preparation time and eliminating errors incurred in transforming temperatures from one finite-element mesh to another.

Chamis, Christos C.↗
NASA NTRS2001

Development and Validation of High Performance Unshrouded Centrifugal Impeller

The feasibility of using a two-stage unshrouded impeller turbopump to replace the current three-stage reusable launch vehicle engine shrouded impeller hydrogen pump has been evaluated from the standpoint of turbopump weight reduction and overall payload improvement. These advantages are a by-product of the higher tip speeds that an unshrouded impeller can sustain. The issues associated with the effect of unshrouded impeller tip clearance on pump efficiency and head have been evaluated with one-dimensional tools and full three-dimensional rotordynamic fluid reaction forces and coefficients have been established through time dependent computational fluid dynamics (CFD) simulation of the whole 360 degree impeller with different rotor eccentricities and whirling ratios. Unlike the shrouded impeller, the unshrouded impeller forces are evaluated as the sum of the pressure forces on the blade and the pressure forces on the hub using the CFD results. The turbopump axial thrust control has been optimized by adjusting the first stage impeller backend wear ring seal diameter and diverting the second stage backend balance piston flow to the proper location. The structural integrity associated with the high tip speed has been checked by analyzing a 3D-Finite Element Model at maximum design conditions (6% higher than the design speed). This impeller was fabricated and tested in the NASA Marshall Space Flight Center water-test rig. The experimental data will be compared with the analytical predictions and presented in another paper. The experimental data provides validation data for the numerical design and analysis methodology. The validated numerical methodology can be used to help design different unshrouded impeller configurations.

Chen, Wei-Chung↗
NASA NTRS2000

An architecture for a brain-image database

The widespread availability of methods for noninvasive assessment of brain structure has enabled researchers to investigate neuroimaging correlates of normal aging, cerebrovascular disease, and other processes; we designate such studies as image-based clinical trials (IBCTs). We propose an architecture for a brain-image database, which integrates image processing and statistical operators, and thus supports the implementation and analysis of IBCTs. The implementation of this architecture is described and results from the analysis of image and clinical data from two IBCTs are presented. We expect that systems such as this will play a central role in the management and analysis of complex research data sets.

NASA Discipline Neuroscience↗
NASA NTRS2007

A Conceptual Aerospace Vehicle Structural System Modeling, Analysis and Design Process

A process for aerospace structural concept analysis and design is presented, with examples of a blended-wing-body fuselage, a multi-bubble fuselage concept, a notional crew exploration vehicle, and a high altitude long endurance aircraft. Aerospace vehicle structures must withstand all anticipated mission loads, yet must be designed to have optimal structural weight with the required safety margins. For a viable systems study of advanced concepts, these conflicting requirements must be imposed and analyzed early in the conceptual design cycle, preferably with a high degree of fidelity. In this design process, integrated multidisciplinary analysis tools are used in a collaborative engineering environment. First, parametric solid and surface models including the internal structural layout are developed for detailed finite element analyses. Multiple design scenarios are generated for analyzing several structural configurations and material alternatives. The structural stress, deflection, strain, and margins of safety distributions are visualized and the design is improved. Over several design cycles, the refined vehicle parts and assembly models are generated. The accumulated design data is used for the structural mass comparison and concept ranking. The present application focus on the blended-wing-body vehicle structure and advanced composite material are also discussed.

Mukhopadhyay, Vivek↗
NASA NTRS2014

Flutter Analysis for Turbomachinery Using Volterra Series

The objective of this paper is to describe an accurate and efficient reduced order modeling method for aeroelastic (AE) analysis and for determining the flutter boundary. Without losing accuracy, we develop a reduced order model based on the Volterra series to achieve significant savings in computational cost. The aerodynamic force is provided by a high-fidelity solution from the Reynolds-averaged Navier-Stokes (RANS) equations; the structural mode shapes are determined from the finite element analysis. The fluid-structure coupling is then modeled by the state-space formulation with the structural displacement as input and the aerodynamic force as output, which in turn acts as an external force to the aeroelastic displacement equation for providing the structural deformation. NASA's rotor 67 blade is used to study its aeroelastic characteristics under the designated operating condition. First, the CFD results are validated against measured data available for the steady state condition. Then, the accuracy of the developed reduced order model is compared with the full-order solutions. Finally the aeroelastic solutions of the blade are computed and a flutter boundary is identified, suggesting that the rotor, with the material property chosen for the study, is structurally stable at the operating condition, free of encountering flutter.

computational fluid dynamics↗
NASA NTRS2015

The Impact of Radiation Damage on Photon Counting with an EMCCD for the WFIRST-AFTA Coronagraph

WFIRST-AFTA is a 2.4m class NASA observatory designed to address a wide range of science objectives using two complementary scientific payloads. The Wide Field Instrument (WFI) offers Hubble quality imaging over a 0.28 square degree field of view, and will gather NIR statistical data on exoplanets through gravitational microlensing. The second instrument is a high contrast coronagraph that will carry out the direct imaging and spectroscopic analysis of exoplanets, providing a means to probe the structure and composition of planetary systems. The coronagraph instrument is expected to operate in low photon flux for long integration times, meaning all noise sources must be kept to a minimum. In order to satisfy the low noise requirements, the Electron Multiplication (EM)-CCD has been baselined for both the imaging and spectrograph cameras. The EMCCD was selected in comparison with other candidates because of its low effective electronic read noise at sub-electron values with appropriate multiplication gain setting. The presence of other noise sources, however, such as thermal dark signal and Clock Induced Charge (CIC), need to be characterised and mitigated. In addition, operation within a space environment will subject the device to radiation damage that will degrade the Charge Transfer Efficiency (CTE) of the device throughout the mission lifetime. Here we present our latest results from pre- and post-irradiation testing of the e2v CCD201-20 BI EMCCD sensor, baselined for the WFIRST-AFTA coronagraph instrument. A description of the detector technology is presented, alongside considerations for operation within a space environment. The results from a room temperature irradiation are discussed in context with the nominal operating requirements of AFTA-C and future work which entails a cryogenic irradiation of the CCD201-20 is presented.

Peddada, Pavani↗
NASA NTRS1990

Theoretical hypervelocity ballistic limit for single or double plates using nonlinear modal analysis

The original research on the use on nonlinear vibration technique to solve for the hypervelocity ballistic limit for double plates is examined. Such structure is commonly found in typical Space Station design where the incoming space or man-made debris would be fragmented upon hitting the outer plate (shield) and the subsequent impact on the main wall would result in a much reduced damge of the space station or spacecraft. The existing few theoretical impact equations do not agree well with each other. The existing computer code bumper used at NASA-Johnson Space Center appears to predict an unconservative ballistic limit when compared with experimental data where the velocity ranges from 3 km/s to 8 km/s. Such unconservative prediction is unacceptable from a practical safe design point of view. The bumper code is based on Wilkinson's (1968) paper and his equations have not been improved nor modified even though they are viewed with suspicion due to lack of agreement with experiments. To make matters worse, there is not another theory which is better than Wilkinson's equation and the designers are forced to use purely empirical Nysmith (1969) or semiempirical equations developed by Cour-Palais in 1969. The Cour-Palais equations were later modified empirically in 1989. The purpose of the present investigation is to examine the many assumptions of the Wilkinson equation. An attempt is made to present design charts based on the modified-Wilkinson equation so that the designer can get a feel of the ranges of the parameters which are of interest and discard a huge range of parameters, thus, significantly reducing the number of test shots required. The analysis is based on a solution of the governing nonlinear differential equations for a plate, assuming axisymmetric behavior using polar coordinates.

Hui, David↗
DOE OSTI2025

High-count-rate effects in event processing for the XRISM/Resolve X-ray microcalorimeter. II. Energy scale and resolution in orbit

The Resolve instrument on the X-ray Imaging and Spectroscopy Mission (XRISM) uses a 36 pixel microcalorimeter designed to deliver high-resolution, non-dispersive X-ray spectroscopy. Although it is optimized for extended sources with low count rates, Resolve observations of bright point sources are still able to provide unique insights into the physics of these objects, as long as high-count-rate effects are addressed in the analysis. These effects include the loss of exposure time for each pixel, changes in the energy scale, and changes in the energy resolution. To investigate these effects under realistic observational conditions, we observed the bright X-ray source, the Crab Nebula, with XRISM at several offset positions with respect to the Resolve field of view and with continuous illumination from 55 Fe sources on the filter wheel. For the spectral analysis, we excluded data where exposure-time loss was too significant to ensure reliable spectral statistics. The energy scale at 6 keV shows a slight negative shift in the high-count-rate regime. The energy resolution at 6 keV worsens as the count rate in electrically neighboring pixels increases, but can be restored by applying a nearest-neighbor coincidence cut (“cross-talk cut”). We examined how these effects influence the observation of bright point sources, using GX 13+1 as a test case, and identified an eV-scale energy offset at 6 keV between the inner (brighter) and outer (fainter) pixels. Users who seek to analyze velocity structures on the order of tens of km s–1 should account for such high-count-rate effects. These findings will aid in the interpretation of Resolve data from bright sources and provide valuable considerations for designing and planning for future microcalorimeter missions.

X-rays: general↗
NASA NTRS1998

A Geometry Based Infra-Structure for Computational Analysis and Design

The computational steps traditionally taken for most engineering analysis suites (computational fluid dynamics (CFD), structural analysis, heat transfer and etc.) are: (1) Surface Generation -- usually by employing a Computer Assisted Design (CAD) system; (2) Grid Generation -- preparing the volume for the simulation; (3) Flow Solver -- producing the results at the specified operational point; (4) Post-processing Visualization -- interactively attempting to understand the results. For structural analysis, integrated systems can be obtained from a number of commercial vendors. These vendors couple directly to a number of CAD systems and are executed from within the CAD Graphical User Interface (GUI). It should be noted that the structural analysis problem is more tractable than CFD; there are fewer mesh topologies used and the grids are not as fine (this problem space does not have the length scaling issues of fluids). For CFD, these steps have worked well in the past for simple steady-state simulations at the expense of much user interaction. The data was transmitted between phases via files. In most cases, the output from a CAD system could go to Initial Graphics Exchange Specification (IGES) or Standard Exchange Program (STEP) files. The output from Grid Generators and Solvers do not really have standards though there are a couple of file formats that can be used for a subset of the gridding (i.e. PLOT3D data formats). The user would have to patch up the data or translate from one format to another to move to the next step. Sometimes this could take days. Specifically the problems with this procedure are:(1) File based -- Information flows from one step to the next via data files with formats specified for that procedure. File standards, when they exist, are wholly inadequate. For example, geometry from CAD systems (transmitted via IGES files) is defined as disjoint surfaces and curves (as well as masses of other information of no interest for the Grid Generator). This is particularly onerous for modern CAD systems based on solid modeling. The part was a proper solid and in the translation to IGES has lost this important characteristic. STEP is another standard for CAD data that exists and supports the concept of a solid. The problem with STEP is that a solid modeling geometry kernel is required to query and manipulate the data within this type of file. (2) 'Good' Geometry. A bottleneck in getting results from a solver is the construction of proper geometry to be fed to the grid generator. With 'good' geometry a grid can be constructed in tens of minutes (even with a complex configuration) using unstructured techniques. Adroit multi-block methods are not far behind. This means that a million node steady-state solution can be computed on the order of hours (using current high performance computers) starting from this 'good' geometry. Unfortunately, the geometry usually transmitted from the CAD system is not 'good' in the grid generator sense. The grid generator needs smooth closed solid geometry. It can take a week (or more) of interaction with the CAD output (sometimes by hand) before the process can begin. One way Communication. (3) One-way Communication -- All information travels on from one phase to the next. This makes procedures like node adaptation difficult when attempting to add or move nodes that sit on bounding surfaces (when the actual surface data has been lost after the grid generation phase). Until this process can be automated, more complex problems such as multi-disciplinary analysis or using the above procedure for design becomes prohibitive. There is also no way to easily deal with this system in a modular manner. One can only replace the grid generator, for example, if the software reads and writes the same files. Instead of the serial approach to analysis as described above, CAPRI takes a geometry centric approach. This makes the actual geometry (not a discretized version) accessible to all phases of the analysis. The connection to the geometry is made through an Application Programming Interface (API) and NOT a file system. This API isolates the top-level applications (grid generators, solvers and visualization components) from the geometry engine. Also this allows the replacement of one geometry kernel with another, without effecting these top-level applications. For example, if UniGraphics is used as the CAD package then Parasolid (UG's own geometry engine) can be used for all geometric queries so that no solid geometry information is lost in a translation. This is much better than STEP because when the data is queried, the same software is executed as used in the CAD system. Therefore, one analyzes the exact part that is in the CAD system. CAPRI uses the same idea as the commercial structural analysis codes but does not specify control. Software components of the CAD system are used, but the analysis suite, not the CAD operator, specifies the control of the software session. This also means that the license issues (may be) minimized and individuals need not have to know how to operate a CAD system in order to run the suite.

Haimes, Robert↗
DOE OSTI2025

Robust Spectral Anomaly Detection in EELS Spectral Images via 3D Convolutional Variational Autoencoders

Abstract A 3D Convolutional Variational Autoencoder (3D‐CVAE) is introduced for automated anomaly detection in electron energy‐loss spectroscopy spectrum imaging (EELS‐SI) data. This approach leverages the full 3D structure of EELS‐SI data to detect subtle spectral anomalies while preserving both spatial and spectral correlations across the datacube. By employing cross‐entropy loss and training on bulk spectra, the model learns to reconstruct bulk features characteristic of the defect‐free material. In exploring methods for anomaly detection, both the 3D‐CVAE approach and principal component analysis (PCA) are evaluated, testing their performance using FeL‐edge ΔEpeak shifts designed to simulate material defects. These results show that 3D‐CVAE achieves superior anomaly detection and maintains consistent performance across various shift magnitudes. The method demonstrates clear bimodal separation between bulk and anomalous spectra, enabling reliable classification. Further analysis verifies that lower‐dimensional representations are robust to anomalies in the data. While performance advantages over PCA diminish with decreasing anomaly concentration, our method maintains high reconstruction quality even in challenging, noise‐dominated spectral regions. This approach provides a robust framework for unsupervised automated detection of spectral anomalies in EELS‐SI data, particularly valuable for analyzing complex material systems.

Chemistry↗