SEARCH · Search NASA
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.
The Employment of Defects (EoD) approach to measuring and analyzing defects seeks to identify and capture trends and phenomena that are critical to managing software quality in the iterative software development lifecycle at JPL.
Explore the source record for details and available documents.
Explore the source record for details and available documents.
The multi-scale and complex process of printing additively manufactured (AM) parts can have unexpected, but predictable, build conditions that result in material microstructure variability. In this work, we describe a fully parallel reduced order modeling approach that has been developed to evaluate the evolution of AM processes, termed the AM moment measure method. This method couples the known sequence of the AM process with a physically informed nearest neighbors’ calculation to map the conditions of a part-scale build. The result is a map of the build that is derived directly from build files or in-situ process monitoring sensors. The methodology and terminology of the approach will be described, and computed build maps will be calculated and compared for various laser powder bed fusion (LPBF) builds of Ti-6Al-4V. Such comparative results develop understanding of how the sequential process actions can affect the LPBF-AM build quality and microstructure variability.
In powder bed fusion additive manufacturing, machines are often equipped with in-situ sensors to monitor the build environment as well as machine actuators and subsystems. The data from these sensors offer rich information about the consistency of the fabrication process within a build and across builds. This information may be used for process monitoring and defect detection; however, little has been done to leverage this data from the machines for more than just coarse-grained process monitoring. In this work we demonstrate how these inherently temporal data may be mapped spatially by leveraging scan path information. We then train a XGBoost machine learning model to predict localized defects—specifically soot–using only the mapped process data of builds from a laser powder bed fusion process as input features. The XGBoost model offers a feature importance metric that will help to elucidate possible relationships between the process data and observed defects. Finally, we analyze the model performance spatially and rationalize areas of greater and lesser performance.
Understanding the relationship between the evolution of microstructures of irradiated LiAlO2pellets and tritium diffusion, retention and release could improve predictions of tritium performance. Given expert-labeled segmented images of irradiated and unirradiated pellets, we trained Deep Convolutional Neural Networks to segment images into defect, grain, and boundary classes. Qualitative microstructural information was calculated from these segmented images to facilitate the comparison of unirradiated and irradiated pellets. We tested modifications to improve the sensitivity of the model, including incorporating meta-data into the model and utilizing uncertainty quantification. The predicted segmentation was similar to the expert-labeled segmentation for most methods of microstructural qualification, including pixel proportion, defect area, and defect density. Overall, the high performance metrics for the best models for both irradiated and unirradiated images shows that utilizing neural network models is a viable alternative to expert-labeled images.
Explore the source record for details and available documents.
The purpose of health monitoring systems is to detect failures or defects for increased safety and performance and to provide on-condition maintenance with reduced costs. The problems associated with health monitoring systems include high rates of false alarms and missed failures, which make monitoring an unreliable and costly task. The reason for this is that unaccounted variations invalidate signal modeling assumptions. Our approach was to focus on vibration monitoring of rotating components. We analyzed baseline signals to determine statistical variations, identify and model factors that influence vibrations (pre-production vs. post-production variations), determine hit and false alarm rates with baseline flight data, model and predict effects of defects and variations on vibrations, and develop algorithms and metrics for failure and anomaly detection in the presence of variations.
During reactor operation, nuclear fuels are subject to extreme temperature and irradiation conditions which can significantly degrade the fuel's thermal transport properties. The reduction in thermal conductivity of the fuel as a result of irradiation-induced lattice defects is arguably the most important fuel performance metric in regard to reactor efficiency and safety. Because thorium dioxide (ThO2) is suitable as a model system for more complex materials such as UO2 and its mixed oxides, we present a theoretical investigation of thermal conductivity reduction seen in defect-bearing thorium dioxide and compare directly to experimental measurements. Phonon-mediated thermal transport of the fuel is modeled by a solution to the Boltzmann transport equation (BTE) for phonons. A cluster dynamics (CD) model for lattice defect evolution during irradiation predicts defect densities which are used as input to the BTE for modeling phonon-defect scatterings. Phonon scatterings by lattice defects include those from point defects and vacancy clusters and interstitial clusters of various sizes. The CD model is benchmarked against structural defect characterization of irradiated thorium dioxide using electron microscopy. Thermal conductivity predicted by the BTE model is compared to measured values for irradiated thorium dioxide specimens below room temperature to isolate effects of phonon-defect scattering from intrinsic 3-phonon processes, which dominate at higher temperatures. The computed conductivity values are in partial agreement at temperatures close to room temperature while slight deviations are observed at the lowest measured temperatures, suggesting that implemented phonon-defect scattering cross-section expressions may not be adequate for low temperatures. The presented work provides a necessary investigation of the influence of irradiation induced defects on fuel performance and represents a first step toward a full characterization of phonon mediated thermal transport in irradiated materials with complex defect microstructure.
Components fabricated using the powder bed fusion laser beam metallic (PBF-LB/M) additive manufacturing (AM) process comprise a multitude of sequential weld passes. Qualifying materials for components in critical applications requires a thorough understanding of pertinent processing-microstructure-property relationships. Porosity defects have a strong adverse effect on the mechanical properties of a material. The morphology of porosity defects created by lack of fusion in PBF-LB/M often exhibit high aspect ratios that can act as stress concentrators and become crack initiation sites. The occurrence of spatter induced porosity is studied using synchronized AM model-based process metrics and porosity measured by high-resolution X-ray computed tomography. A novel AM process metric is introduced to quantify the relative exposure to spatter ejecta impacts and incorporation throughout the sequence of the PBF-LB/M process. Further characterization using optical metallography and scanning electron microscopy indicated that both keyhole and lack of fusion porosity correlated statistically with the spatter ejecta impact field of the PBF-LB/M process. Both keyhole and lack of fusion porosity can be predicted by the spatter exposure process metric, which can be exploited to control the fusion sequence.
Defect diffusion is a key process in materials science and catalysis, but as migration mechanisms are often too complex to enumerate a priori, calculation of transport tensors typically have no measure of convergence and require significant end-user intervention. These two bottlenecks prevent high-throughput implementations essential to propagate model-form uncertainty from interatomic interactions to predictive simulations. In order to address these issues, we extend a massively parallel accelerated sampling scheme, autonomously controlled by Bayesian estimators of statewide sampling completeness, to build atomistic kinetic Monte Carlo models on a state-space irreducible under exchange and space group symmetries. Focusing on isolated defects, we derive analytic expressions for drift and diffusion coefficients, providing a convergence metric by calculating the Kullback–Leibler divergence across the ensemble of diffusion processes consistent with the sampling uncertainty. The autonomy and efficacy of the method is demonstrated on surface trimers in tungsten and Hexa-interstitials in magnesium oxide, both of which exhibit complex, correlated migration mechanisms.
Polymer Electrolyte Membrane (PEM) fuel cells are a promising source of alternative energy. However, their production is limited by a lack of well-established methods for quality control of their constituent materials like the membrane-electrode assembly during roll-to-roll manufacturing. One potential solution is the implementation of deep learning methods to detect unwanted defects through their detection in scanned images. Here we explore the detection of defects like scratches, pinholes, and scuffs in a sample dataset of PEM optical images using two deep learning algorithms: Patch Distribution Modeling (PaDiM) for unsupervised anomaly detection and Faster-RCNN for supervised object detection. Both methods achieve scores on performance metrics (ROC-AUC and PRO-AUC for PaDiM and AP for Faster-RCNN) that are comparable to their scores on benchmark datasets. These methods also have the potential to detect a wider range of defects compared to IR thermography and previous optical inspection methods. Overall, deep learning shows promise at detecting relevant defects of interest and has the potential to achieve real-time defect detection.
Nuclear fuel operates in an extreme environment that induces complex multiphysics phenomena occurring over distances ranging from inter-atomic spacing to meters, and times scales ranging from microseconds to years. To simulate this behavior requires a wide variety of material models that are often complex and nonlinear. The recently developed BISON code represents a powerful fuel performance simulation tool based on its material and physical behavior capabilities, finite-element versatility of spatial representation, and use of parallel computing. The code can operate in full three dimensional (3D) mode, as well as in reduced two dimensional (2D) modes, e.g., axisymmetric radial-axial (R-Z) or plane radial-circumferential (R-θ), to suit the application and to allow treatment of global and local effects. A BISON case study was used in this paper to illustrate analysis of Pellet Clad Mechanical Interaction failures from manufacturing defects using combined 2D and 3D analyses. The analysis involved commercial fuel rods and demonstrated successful computation of metrics of interest to fuel failures, including cladding peak hoop stress and strain energy density. Finally, in comparison with a failure threshold derived from power ramp tests, results corroborate industry analyses of the root cause of the pellet-clad interaction failures and illustrate the importance of modeling 3D local effects around fuel pellet defects, which can produce complex effects including cold spots in the cladding, stress concentrations, and hot spots in the fuel that can lead to enhanced cladding degradation such as hydriding, oxidation, CRUD formation, and stress corrosion cracking.
A great deal of effort is now being devoted to the study, analysis, prediction, and minimization of software maintenance expected cost, long before software is delivered to users or customers. It has been estimated that, on the average, the effort spent on software maintenance is as costly as the effort spent on all other software costs. Software design methods should be the starting point to aid in alleviating the problems of software maintenance complexity and high costs. Two aspects of maintenance deserve attention: (1) protocols for locating and rectifying defects, and for ensuring that noe new defects are introduced in the development phase of the software process; and (2) protocols for modification, enhancement, and upgrading. This article focuses primarily on the second aspect, the development of protocols to help increase the quality and reduce the costs associated with modifications, enhancements, and upgrades of existing software. This study developed parsimonious models and a relative complexity metric for complexity measurement of software that were used to rank the modules in the system relative to one another. Some success was achieved in using the models and the relative metric to identify maintenance-prone modules.
We examine the effects of the Rastall parameter on the behaviour of spherically symmetric static distributions of perfect fluid matter. It was claimed by Visser [Physics Letters B, 782, 83, (2018)] that the Rastall proposition is completely equivalent to the Einstein theory. While many authors have raised contrary arguments, our intention is to analyse the properties of Rastall gravity through variation of the Rastall parameter in the context of perfect fluids spheres that may be used to model neutron stars or cold fluid planets. This analysis also serves to counter the claim that Rastall gravity is equivalent to the standard Einstein theory. It turns out that the condition of pressure isotropy is exactly the same as for Einstein gravity and hence that any known solution of the Einstein equations may be used to study the effects of the Rastall dynamical quantities. Moreover, by choosing the well studied Tolman metrics, we discover that in the majority of cases there is substantial deviation from the Einstein case when the Rastall parameter vanishes and in cases where the Einstein model displays defective behaviour, certain Rastall models obey the well known elementary requirements for physical plausibility. These empirical findings do not support the idea that Rastall theory is equivalent to Einstein theory as several deviations in physical behaviour are displayed as counter-examples.
Components fabricated using the powder bed fusion laser beam metallic (PBF-LB/M)additive manufacturing process are the result of a multitude of weld passes conducted sequentially. Qualifying components for aerospace applications requires a thorough understanding of the process-structure-properties relationships. Porosity defects are known to have a strong adverse effect on the mechanical properties of a component. In particular, porosity defects created by lack of fusion have high aspect ratio morphologies leading to stress concentrations that become crack initiation sites. In the present work, the occurrence of spatter induced lack of fusion porosity was studied using synchronized in-situ process monitoring, additive manufacturing model-based process metrics, and high-resolution X-ray computed tomography. The results show that lack of fusion porosity is statistically correlated with unremoved welding spatter ejecta of the PBF-LB/M process and process metrics related to the hatching strategy.
Imaging nanoscale radiation-induced defects using the transmission electron microscope (TEM) is a key factor in the successful implementation of materials for nuclear energy structural applications. Analyzing each defect in a TEM micrograph is currently a manual task. To identify the defects in a single image can take anywhere from 15 min to an hour and a project can require the analysis of anywhere from tens to ≥ 100 images. Here, we use artificial intelligence (AI) models to automate this task. For simplification, we evaluated images with only a single type of defect; helium bubbles. Additionally, we performed semantic segmentation of these helium bubble defects in electron microscopy images of irradiated FeCrAl alloys using a deep learning DefectSegNet model. This model, which was previously used to classify crystal defects, is inspired by the classic DenseNet and U-Net image segmentation models. It claims high spatial resolution, but has poor performance at object boundaries. Our paper improves the DefectSegNet model’s application by adding two new features. First, the DefectSegNet model is applied not only to perform calculation pixel-wise but also object (or feature) wise. Because object-wise metrics are directly relevant to our final goal of detecting bubbles, whereas pixel-wise classification is only an intermediate step, it is an important part of our study. Second, a distance map loss (DML) function has been added to increase its performance at object boundaries. It is crucial to accurately represent defects boundaries, especially bubbles, in order to track the bubble-induced swelling caused by irradiation. The boundary-focused DML function is also compared to other loss functions like Cross-entropy, Weighted Binary Cross Entropy (WBCE), Dice and Intersection over Union (IOU). Finally, by incorporating new features, we found a marked improvement on segmentation quality and better shape preservation at the boundaries and areas of the bubbles.
Components fabricated using the powder bed fusion laser beam metallic (PBF) additive manufacturing (AM) process comprise a multitude of sequential weld passes. Porosity defects resulting from weld pass inconsistencies are a concern for PBF materials and have a strong adverse effect on the mechanical properties. In an effort to understand and avoid spatter induced porosity, this work aims to provide a novel computational technique that can be used to quantify and predict the sequence-sensitive impact of spatter upon the PBF process stability and consequential porosity in as-printed material. A novel spatter impact AM model-based process metric (AM-PM) is introduced to model and quantify the impact that spatter has on the PBF process. The occurrence of spatter induced porosity is studied using thermal rise, lack of fusion, and spatter impact AM-PMs synchronized with porosity measured by high-resolution X-ray computed tomography. Six cylindrical specimens were designed to compare the AM-PM correlations with porosity across two different laser powers and three hatch scanning strategies. A point field based computational approach is shown to be effective for testing the influence of the different AM-PMs and interrogating the physical process conditions underlying the porosity formation. Further, characterization using optical metallography and scanning electron microscopy indicated that both keyhole and lack of fusion porosity correlated with the spatter impact AM-PM, which can be exploited to predict and control spatter related porosity by the hatch strategy. Efforts to predict porosity composition within PBF material should incorporate the PBF process disturbances that can result from spatter. The spatter impact AM-PM can be readily incorporated into the development of defect prediction models at the part scale.