Search NASA⌕ Search

SEARCH · Search NASA

Results for “Requirement 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 325 records · Page 18
All sourcesNASA NTRSDOE OSTI
DOE OSTI2012

Application of the JENDL-4.0 nuclear data set for uncertainty analysis of the prototype FBR Monju

This paper deals with uncertainty analysis of the Monju reactor using JENDL-4.0 and the ERANOS code 1. In 2010 the Japan Atomic Energy Agency - JAEA - released the JENDL-4.0 nuclear data set. This new evaluation contains improved values of cross-sections and emphasizes accurate covariance matrices. Also in 2010, JAEA restarted the sodium-cooled fast reactor prototype Monju after about 15 years of shutdown. The long shutdown time resulted in a build-up of {sup 241}Am by natural decay from the initially loaded Pu. As well as improved covariance matrices, JENDL-4.0 is announced to contain improved data for minor actinides 2. The choice of Monju reactor as an application of the new evaluation seems then even more relevant. The uncertainty analysis requires the determination of sensitivity coefficients. The well-established ERANOS code was chosen because of its integrated modules that allow users to perform sensitivity and uncertainty analysis. A JENDL-4.0 cross-sections library is not available for ERANOS. Therefor a cross-sections library had to be made from the original ENDF files for the ECCO cell code (part of ERANOS). For confirmation of the newly made library, calculations of a benchmark core were performed. These calculations used the MZA and MZB benchmarks and showed consistent results with other libraries. Calculations for the Monju reactor were performed using hexagonal 3D geometry and PN transport theory. However, the ERANOS sensitivity modules cannot use the resulting fluxes, as these modules require finite differences based fluxes, obtained from RZ SN-transport or 3D diffusion calculations. The corresponding geometrical models have been made and the results verified with Monju restart experimental data 4. Uncertainty analysis was performed using the RZ model. JENDL-4.0 uncertainty analysis showed a significant reduction of the uncertainty related to the fission cross-section of Pu along with an increase of the uncertainty related to the capture cross-section of {sup 238}U compared with the previous JENDL-3.3 version. Covariance data recently added in JENDL-4.0 for {sup 241}Am appears to have a non-negligible contribution. (authors)

21 SPECIFIC NUCLEAR REACTORS AND ASSOCIATED PLANTS↗
DOE OSTI2025

Evaluating FRI3D for Cost Savings in Fire Hazard Analysis at DOE Sites

A fire hazard analysis, required for many U.S. Department of Energy (DOE) facilities, is a complex, cumbersome, and costly process. Fire hazard analyses may be viewed as a checkbox, but ideally and in spirit with the DOE-STD-1066, the fire hazard analysis (FHA) should be a part of the workflow and used to help in modifications, maintenance, and improving operational safety. With current FHA development processes, it is both time and cost prohibitive for true integration. A tool called Fire Risk Investigation in 3D or FRI3D was developed under the DOE Light Water Reactor Sustainability program to simplify and automate many aspects of a fire probabilistic risk analysis for existing nuclear power plants. The FRI3D tool automates fire scenarios by combining approved fire simulation codes, U.S. Nuclear Regulatory Commission fire calculations methods, 3D modeling and visualization, and probabilistic risk analysis models into a single workflow supported with a user interface. FRI3D was initially designed for used in combination with a PRA, this case study, evaluated using FRI3D for a plant modification, determined the benefits that detailed fire modeling can have for U.S. Department of Energy facilities with or without a PRA model. It also looked at what tasks from DOE requirements could be reduced using the tool and what is needed to integrate fire hazard analysis into site workflow.

97 - MATHEMATICS AND COMPUTING↗
DOE OSTI2025

Improving Cyber Situational Understanding

Effective cybersecurity operations require the ability to analyze large amounts of information to assess security risks and formulate defensive strategies against adversaries. This has become more complex in recent years as the sprawl and interconnectivity of devices grows through implementation of virtualization, cloud computing, and Internet of Things (IoT). The amount of data and analysis required for effective cybersecurity command and control decisions far exceeds humans’ capacity to perform manually. We characterize the analysis problem as cyber situational understanding. The research presented to improve cyber situational understanding focuses on vulnerability analysis and threat intelligence. Regarding vulnerabilities, entities must analyze and plan work for between thousands and tens of thousands of software vulnerabilities annually. Entities heavily use network firewalls to limit vulnerability exposure. As a result, some of these vulnerabilities permit exposure to adversarial exploitation, whereas others are inaccessible and therefore present negligible risk of exploitation. Distinguishing between high and low risk software vulnerabilities requires a deep understanding of the vulnerability, network firewall protection, and characteristics of the targeted device. This problem is solved by extracting network service features from vulnerability data features using both machine-learning and natural language processing. Then, the network firewall topology is parsed to determine which vulnerabilities are reachable by adversaries. Ultimately, a state-based safety analysis ascertains which vulnerabilities are unsafe. A related vulnerability analysis problem occurs in cybersecurity operations when associating an entity’s hardware and software assets to public vulnerability databases. Assets often reveal hardware and software through installation artifacts and network service identification, and entities store these artifacts in inventory databases. However, software and hardware vendors apply a standard Common Platform Enumeration (CPE) naming convention when publicly reporting vulnerabilities. Associating these two datasets often requires many hours to days of manual inspection. The proposed solution automates the mapping approach of human analysts using fuzzy matching techniques, natural language processing, and, ultimately, machine learning to present a small set of recommendations for mapping the two datasets. The result significantly reduces human analysis time and reduces the occurrence of false positives in vulnerability notifications. Finally, cyber threat intelligence (CTI) requires associating cyber observable artifacts, such as IP addresses, URIs, and file hashes, with cyber threat tactics, techniques, and procedures. Unfortunately, most CTI data is compartmentalized across multiple organizations and cannot be shared due to the legal and reputational risk with cyber threat being associated with the entity. The approach to solving this problem inovlves using a distributed ledger with anonymous token spending and authentication. This allows a consortium of semi-trusted entities to share the workload of curating CTI for a threat sharing community’s cooperative benefit.

Huff, Philip↗
DOE OSTI2011

Including Shielding Effects in Application of the TPCA Method for Detection of Embedded Radiation Sources

Conventional full spectrum gamma spectroscopic analysis has the objective of quantitative identification of all the radionuclides present in a measurement. For low-energy resolution detectors such as NaI, when photopeaks alone are not sufficient for complete isotopic identification, such analysis requires template spectra for all the radionuclides present in the measurement. When many radionuclides are present it is difficult to make the correct identification and this process often requires many attempts to obtain a statistically valid solution by highly skilled spectroscopists. A previous report investigated using the targeted principal component analysis method (TPCA) for detection of embedded sources for RPM applications. This method uses spatial/temporal information from multiple spectral measurements to test the hypothesis of the presence of a target spectrum of interest in these measurements without the need to identify all the other radionuclides present. The previous analysis showed that the TPCA method has significant potential for automated detection of target radionuclides of interest, but did not include the effects of shielding. This report complements the previous analysis by including the effects of spectral distortion due to shielding effects for the same problem of detection of embedded sources. Two examples, one with one target radionuclide and the other with two, show that the TPCA method can successfully detect shielded targets in the presence of many other radionuclides. The shielding parameters are determined as part of the optimization process using interpolation of library spectra that are defined on a 2D grid of atomic numbers and areal densities.

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗
NASA NTRS1998

Microcracking of Materials for Space

The effect of thermal-cycling-induced microcracking in fiber-reinforced polymer matrix composites is studied. Specific attention is focused on microcrack density as a function of the number of thermal cycles, and the effect of microcracking on the dimensional stability of composite materials. Changes in laminate coefficient of thermal expansion (CTE) and laminate stiffness are of primary concern. Included in the study are materials containing four different Thornel fiber types: a PAN-based T50 fiber and three pitch-based fibers, P55, P75, and P120. The fiber stiffnesses range from 55 Msi to 120 Msi. The fiber CTE's range from -0.50 x 1O(exp -6)/degrees F to -0.80 x 10(exp -6)/degrees F. Also included are three matrix types: Fiberite's 934 epoxy, Amoco's ERL1962 toughened epoxy, and YLA's RS3 cyanate ester. The lamination sequences of the materials considered include a cross-ply configuration, [0/90](sub 2s), and two quasi-isotropic configurations, [0/+45/-45/90](sub s), and [0/+45/90/-45](sub s). The layer thickness of the materials range from a nominal 0.001 in. to 0.005 in. In addition to the variety of materials considered, three different thermal cycling temperature ranges are considered. These temperature ranges are +/-250 degrees F, +/-l5O degrees F, and +/-50 degrees F. The combination of these material and geometric parameters and temperature ranges, combined with thermal cycling to thousands of cycles, makes this one of the most comprehensive studies of thermal-cycling-induced microcracking to date. Experimental comparisons are presented by examining the effect of layer thickness, fiber type, matrix type, and thermal cycling temperature range on microcracking and its influence on the laminates. Results regarding layer thickness effects indicate that thin-layer laminates microcrack more severely than identical laminates with thick layers. For some specimens in this study, the number of microcracks in thin-layer specimens exceeds that in thick-layer specimens by more than a factor of two. Despite the higher number of microcracks in the thin-layer specimens, small changes in CTE after thousands of cycles indicate that the thin-layer specimens are relatively unaffected by the presence of these cracks compared to the thick-layer specimens. Results regarding fiber type indicate that the number of microcracks and the change in CTE after thousands of cycles in the specimens containing PAN-based fibers are less than in the specimens containing comparable stiffness pitch-based fibers. Results for specimens containing the different pitch-based fibers indicate that after thousands of cycles, the number of microcracks in the specimens does not depend on the modulus or CTE of the fiber. The change in laminate CTE does, however, depend highly on the stiffness and CTE of the fiber. Fibers with higher stiffness and more negative CTE exhibit the lowest change in laminate CTE as a result of thermal cycling. The overall CTE of these specimens is, however, more negative as a result of the more negative CTE of the fiber. Results regarding matrix type based on the +/-250 degree F temperature range indicate that the RS3 cyanate ester resin system exhibits the greatest resistance to microcracking and the least change in CTE, particularly for cycles numbering 3000 and less. Extrapolations to higher numbers of cycles indicate, however, that the margin of increased performance is expected to decrease with additional thermal cycling. Results regarding thermal cycling temperature range depend on the matrix type considered and the layer thickness of the specimens. For the ERL1962 resin system, microcrack saturation is expected to occur in all specimens, regardless of the temperature range to which the specimens are exposed. By contrast, the RS3 resin system demonstrates a threshold effect such that cycled to less severe temperature ranges, microcracking does not occur. For the RS3 specimens with 0.005 in. layer thickness, no microcracking or changes in CTE are observed in specimens cycled between +/-150 degree F or +/- 50 degree F. For the RS3 specimens with 0.002 in. layer thickness, no microcracking or changes in CTE are observed in specimens cycled between +/-50 degree F. Results regarding laminate stiffness indicate negligible change in laminate stiffness due to thermal cycling for the materials and geometries considered in this investigation. The study includes X-ray examination of the specimens, showing that cracks observed at the edge of the specimens penetrate the entire width of the specimen. Glass transition temperatures of the specimens are measured, showing that resin chemistry is not altered as a result of thermal cycling. Results are also presented based on a one-dimensional shear lag analysis developed in the literature. The analysis requires material property information that is difficult to obtain experimentally. Using limited data from the present investigation, material properties associated with the analysis are modified to obtain reasonable agreement with measured microcrack densities. Based on these derived material properties, the analysis generally overpredicts the change in laminate CTE. Predicted changes in laminate stiffness show reasonable correlation with experimentally measured values.

Brown, Timothy L.↗
NASA NTRSDate not supplied

Ramdb: The NASA Raman Spectral Database (version 1.00).

Given that, in most instances, minimal sample preparation is required and due to its contactless instrument design, Raman spectroscopy is one of the most versatile vibrational spectroscopic techniques for the chemical analysis of environmental and biological specimens. The diversity of applications of Raman spectroscopy ranges anywhere from art [1] to planetary science missions [2]. The advancement in the use of Raman spectroscopy in Solar System missions, notably in post-mission sample return analysis, requires a spectral library holding the broad range of specimens that could be found in Solar System sources. For this purpose, we have initiated the development of a Raman spectral database (Ramdb) at NASA Ames Research Center. Currently, the database includes experimental and theoretical Raman spectra of PAHs [3, 4], as well as laboratory Raman spectra of amino acids, carbon allotropes, minerals, and analogs relevance to Earth Sciences [5], Exobiology [6], Planetary [7], and Astrochemistry [8] to name just a few examples. Ramdb can be found on the web at www.astrochemistry.org/ramdb, where raw and processed Raman spectra can be downloaded in CSV format. The laboratory Raman spectra are measured using a laser Raman spectrometer (JASCO NRS-5500-532QRI). The Raman instrument is equipped with three excitation lasers, with wavelengths of 405, 532, and 785 nm. A clean silicon substrate is used as the internal standard for wavenumber calibration. Powdered samples were prepared (microscopic >10 um, grounded microscopic < 10 um) on glass slides. Some raw data exhibited a background signal arising as a combination of laser-induced fluorescence from the sample. To correct this background, we developed a Python pipeline that uses open-source Python libraries. Ramdb provides both raw and processed (using Python pipeline) data, which includes tabulated Raman shift transitions and other measurement details. The theoretical Raman band positions of PAHs (pyrene monomers and tetramer clusters) were computed using density functional theory (DFT) with the help of the Gaussian 16 suite of programs [9]. In the near future, Ramdb will serve as a repository of Raman spectral data from Laboratory Astrophysics and Planetary Science experiments involving the irradiation of organic compounds under simulated space and planetary conditions. In addition, online and offline tools will be developed for utilising the database for comparison to the user’s sample.

N Punnakayathil↗
NASA NTRS1982

Galileo spacecraft pointing accuracy analysis

This paper describes the analysis required to evaluate the pointing performance of the Galileo spacecraft science instruments and antenna. The main contribution lies in the development of the models needed for pointing analysis. Pointing control mechanisms of the science instruments, as well as the spacecraft autonomous attitude determination process and the pointing control mechanisms for the high gain antenna, are provided in order to identify the various error sources involved. A covariance analysis method is then used to obtain the pointing capabilities of the instruments and the antenna. The results are depicted in tabular form to compare the capabilities against the levied requirements.

Hayati, S. A.↗
DOE OSTI2013

Relationship between dislocations and residual stresses in cold-drawn pearlitic steel analyzed by energy-dispersive X-ray diffraction

We analyzed the dislocation distribution of cold-drawn pearlitic-steel wire by using the line-profile analysis based on the energy dispersive X-ray diffraction (EDXD). Although this line-profile analysis requires a high resolution in reciprocal space, the resolution for EDXD is generally poor due to the energy resolution of the detector. Our analysis demonstrated that the resolution in the reciprocal space can be maximized at small scattering angles. Using the line-profile analysis based on the EDXD, the microstructural parameters such as the crystallite size and the dislocation density of the ferrite phase in the pearlitic steel were successfully analyzed. In addition, the distribution of the residual stress of the ferrite phase of a pearlitic steel wire was also analyzed using the EDXD measurement. - Highlights: • Energy dispersive X-ray diffraction is applied to the line-profile analysis. • Distribution of dislocations in ferrite in the pearlitic steel wire is analyzed. • Relationship between dislocations and residual stress is discussed.

36 MATERIALS SCIENCE↗
DOE OSTI2016

Advanced Computational Framework for Environmental Management ZEM, Version 1.x

Typically environmental management problems require analysis of large and complex data sets originating from concurrent data streams with different data collection frequencies and pedigree. These big data sets require on-the-fly integration into a series of models with different complexity for various types of model analyses where the data are applied as soft and hard model constraints. This is needed to provide fast iterative model analyses based on the latest available data to guide decision-making. Furthermore, the data and model are associated with uncertainties. The uncertainties are probabilistic (e.g. measurement errors) and non-probabilistic (unknowns, e.g. alternative conceptual models characterizing site conditions). To address all of these issues, we have developed an integrated framework for real-time data and model analyses for environmental decision-making called ZEM. The framework allows for seamless and on-the-fly integration of data and modeling results for robust and scientifically-defensible decision-making applying advanced decision analyses tools such as Bayesian- Information-Gap Decision Theory (BIG-DT). The framework also includes advanced methods for optimization that are capable of dealing with a large number of unknown model parameters, and surrogate (reduced order) modeling capabilities based on support vector regression techniques. The framework is coded in Julia, a state-of-the-art high-performance programing language (http://julialang.org). The ZEM framework is open-source and can be applied to any environmental management site. The framework will be open-source and released under GPL V3 license.

Vesselinov, VelimirV.↗
NASA NTRS1972

Modular space station, phase B extension. Information management advanced development. Volume 4: Data processing assembly

The computation and logical functions which are performed by the data processing assembly of the modular space station are defined. The subjects discussed are: (1) requirements analysis, (2) baseline data processing assembly configuration, (3) information flow study, (4) throughput simulation, (5) redundancy study, (6) memory studies, and (7) design requirements specification.

Gerber, C. R.↗
NASA NTRS1998

Modeling of the Magnetosphere-Ionosphere-Atmosphere System

This grant covered the transition period between the ISTP/GGS main mission MAMI work (NAGW5-1097) to the extended mission of GGS/SolaxMax. During this period we concentrated on the following subjects: (1) Ionospheric modeling and in particular the coupling of the ionosphere and the magnetosphere require knowledge of the ionospheric conductances. We have an algorithm to derive these conductances from POLAR UVI or POLAR VIS images. The image analysis requires carefully calibrated measurements, separation of auroral brightness from scattered solar EUV and emissions caused by photoelectrons, and modeling of the auroral conditions, Under this grant we evaluated the various sources of uncertainty that contribute to the conductance calculation by comparing different model approaches and we conducted sensitivity studies. A joint modeling effort between our ionospheric models and a global MHD magnetospheric model with S. Slinker and J. Fedder lead to improved parameterizations for the MHD model. The predictions of that new model were validated using data from ground based instruments, POLAR UVI images, and DMSP satellites. These findings were presented at the GEM meeting in Snowmass, and at the Huntsville modeling workshop in Guntersville. A manuscript for publication is in preparation. A study to evaluate the thermospheric effects of large substorms and storms was con- ducted with the TIE-GCM model. Of particular interest are compositional changes of the thermosphere in the southern hemisphere. Southern hemisphere effects are stronger than those on the northern hemisphere because of the larger offset of the magnetic dipole from the rotational axis. These results were presented at the Huntsville modeling workshop in Guntersville.

Source record↗
NASA NTRS1992

Geometric requirements for multidisciplinary analysis of aerospace-vehicle design

The geometric requirements for creating surfaces and grids for multidisciplinary analysis and optimization of aerospace-vehicle designs are described. Geometric surface representations are outlined and compared. Directions for future designs are proposed. High-speed civil transport aircraft configurations are targeted to demonstrate the processes.

Smith, Robert E.↗
DOE OSTI2011

NID Copper Sample Analysis

The current focal point of the nuclear physics program at PNNL is the MAJORANA DEMONSTRATOR, and the follow-on Tonne-Scale experiment, a large array of ultra-low background high-purity germanium detectors, enriched in 76Ge, designed to search for zero-neutrino double-beta decay (0νββ). This experiment requires the use of germanium isotopically enriched in 76Ge. The DEMONSTRATOR will utilize 76Ge from Russia, but for the Tonne-Scale experiment it is hoped that an alternate technology under development at Nonlinear Ion Dynamics (NID) will be a viable, US-based, lower-cost source of separated material. Samples of separated material from NID require analysis to determine the isotopic distribution and impurities. The MAJORANA DEMONSTRATOR is a DOE and NSF funded project with a major science impact. DOE is funding NID through an SBIR grant for development of their separation technology for application to the Tonne-Scale experiment. The Environmental Molecular Sciences facility (EMSL), a DOE user facility at PNNL, has the required mass spectroscopy instruments for making these isotopic measurements that are essential to the quality assurance for the MAJORANA DEMONSTRATOR and for the development of the future separation technology required for the Tonne-Scale experiment. A sample of isotopically separated copper was provided by NID to PNNL for isotopic analysis as a test of the NID technology. The results of that analysis are reported here.

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗
DOE OSTI2011

NID Copper Sample Analysis

The current focal point of the nuclear physics program at PNNL is the MAJORANA DEMONSTRATOR, and the follow-on Tonne-Scale experiment, a large array of ultra-low background high-purity germanium detectors, enriched in 76Ge, designed to search for zero-neutrino double-beta decay (0νββ). This experiment requires the use of germanium isotopically enriched in 76Ge. The MAJORANA DEMONSTRATOR is a DOE and NSF funded project with a major science impact. The DEMONSTRATOR will utilize 76Ge from Russia, but for the Tonne-Scale experiment it is hoped that an alternate technology, possibly one under development at Nonlinear Ion Dynamics (NID), will be a viable, US-based, lower-cost source of separated material. Samples of separated material from NID require analysis to determine the isotopic distribution and impurities. DOE is funding NID through an SBIR grant for development of their separation technology for application to the Tonne-Scale experiment. The Environmental Molecular Sciences facility (EMSL), a DOE user facility at PNNL, has the required mass spectroscopy instruments for making isotopic measurements that are essential to the quality assurance for the MAJORANA DEMONSTRATOR and for the development of the future separation technology required for the Tonne-Scale experiment. A sample of isotopically separated copper was provided by NID to PNNL in January 2011 for isotopic analysis as a test of the NID technology. The results of that analysis are reported here. A second sample of isotopically separated copper was provided by NID to PNNL in August 2011 for isotopic analysis as a test of the NID technology. The results of that analysis are also reported here.

copper isotopics↗
NASA NTRS1982

Solution of nonlinear multivariable constrained systems using a gradient projection digital algorithm that is insensitive to the initial state

Optimal digital control of nonlinear multivariable constrained systems was studied. The optimal controller in the form of an algorithm was improved and refined by reducing running time and storage requirements. A particularly difficult system of nine nonlinear state variable equations was chosen as a test problem for analyzing and improving the controller. Lengthy analysis, modeling, computing and optimization were accomplished. A remote interactive teletype terminal was installed. Analysis requiring computer usage of short duration was accomplished using Tuskegee's VAX 11/750 system.

Hargrove, A.↗
NASA NTRS2020

Flutter Analysis with Stabilized Finite Elements Based on the Linearized Frequency-Domain Approach

When designing and certifying aircraft, engineers must take into consideration aeroelastic effects such as flutter. Design and certification of a vehicle may require analysis of thousands of aeroelastic responses. Standard tools in the aerospace industry are based on linear aerodynamic models such as the doublet-lattice method, but these methods can be nonconservative in certain situations such as in the transonic regime. While computational fluid dynamics (CFD) is a higher fidelity alternative, the time-marching approach has a drastically increased computational cost compared to the linear aerodynamic methods. By taking advantage of the periodic nature of flutter, frequency-domain methods offer a more efficient alternative to time-marching CFD. In this work, a linearized frequency-domain method is implemented and verified in the stabilized finite-element solver in FUN3D. The linearized frequency-domain method is demonstrated and compared to other methods for traditional benchmark cases for computational aeroelasticity: the AGARD 445.6 wing, the Benchmark Supercritical Wing, and the Benchmark NACA 0012Wing.

Kevin E Jacobson↗
DOE OSTI2024

Performance Comparison of Machine Learning Models for Ultrasonic Nondestructive Evaluation of Alkali-Silica Reaction in Concrete

Alkali-silica reaction (ASR) causes concrete degradation, leading to cracking, rebar corrosion, and reduced structural integrity, which raises safety concerns. Ultrasonic nondestructive evaluation (NDE) effectively assesses concrete properties and monitors ASR progression. However, its deployment and analysis require specialized expertise and subjective interpretation. As computational power increases, artificial intelligence (AI) and machine learning (ML) algorithms are increasingly being used to automate NDE data analysis across various industries for AI-assisted automation. Regulatory agencies are adapting to this technological shift, prompting a need to evaluate current ML technologies’ capabilities and limitations in assessing concrete material properties and damage. This report presents a comparative analysis of four ML regression models for predicting concrete material damage induced by ASR expansion using long-term ultrasonic data monitoring. The models investigated include linear regression (LR), support vector regression (SVR), shallow neural networks (NN), and deep neural networks (DNN). LR, SVR, and shallow NN models use features extracted from ultrasonic signals, whereas the DNN model processes time-domain ultrasonic signals and frequency spectra directly. The study systematically compared the models’ performance from various perspectives, including model input, prediction performance, and generalization ability. The findings indicate significant variability in model performance, with some ML algorithms achieving very high or very low prediction accuracy depending on the preprocessing and feature engineering (extraction and selection) applied. Key insights include the observation that shallow ML models (LR, SVR, and shallow NNs) require meticulous preprocessing and feature extraction to achieve high accuracy. In contrast, the DNN model, although it bypasses the need for feature engineering, necessitates extensive preprocessing to mitigate noise and computational demands. The SVR model emerged as the top performer among the shallow models, and the DNN model exhibited superior performance on specific datasets but struggled with generalization across specimens from different batches. Additionally, the SVR model is sensitive to temperature variations, whereas the DNN model is robust in this regard. Using recurrent neural networks is recommended for future ASR expansion prediction studies. Recurrent neural networks’ inherent ability to capture temporal dependencies and long-term patterns makes them well suited for analyzing sequential ultrasonic monitoring data. Overall, the results and conclusions of this study could provide insights into the capabilities and effectiveness of ML when applied to ultrasonic NDE data and help identify best practices for using ML for ultrasonic NDE of concrete material properties.

36 MATERIALS SCIENCE↗