Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “A. Microstructures”

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 127 records · Page 7
All sourcesNASA NTRSDOE OSTI
DOE OSTI2023

Automated, high-accuracy classification of textured microstructures using a convolutional neural network

Crystallographic texture is an important descriptor of material properties but requires time-intensive electron backscatter diffraction (EBSD) for identifying grain orientations. While some metrics such as grain size or grain aspect ratio can distinguish textured microstructures from untextured microstructures after significant grain growth, such morphological differences are not always visually observable. This paper explores the use of deep learning to classify experimentally measured textured microstructures without knowledge of crystallographic orientation. A deep convolutional neural network is used to extract high-order morphological features from binary images to distinguish textured microstructures from untextured microstructures. The convolutional neural network results are compared with a statistical Kolmogorov–Smirnov tests with traditional morphological metrics for describing microstructures. Results show that the convolutional neural network achieves a significantly improved classification accuracy, particularly at early stages of grain growth, highlighting the capability of deep learning to identify the subtle morphological patterns resulting from texture. The results demonstrate the potential of a convolutional neural network as a tool for reliable and automated microstructure classification with minimal preprocessing.

36 MATERIALS SCIENCE↗
DOE OSTI2020

Transient Microstructure Evolutions and Local Properties of Dual-Phase 980 MPa Grade Steel Via Friction Stir Spot Processing

Friction stir processing is a novel solid-state process to modify microstructures and their properties by intense, localized plastic deformation. However, little research has been reported for microstructure evolutions of advanced high-strength steels during the process. The present work focuses on the study of transient microstructure changes and local mechanical properties for friction stir spot processed dual-phase (DP) 980 MPa grade steel (DP980) under different peak temperatures. A pinless silicon nitride ceramic tool was used to produce relatively simple material deformation and flow near the tool. Friction stir spot processed steel samples were characterized by optical and electron microscopies. Furthermore, Vickers microhardness and nano-indentation measurements were used to study local mechanical properties for correlation with microstructures. A swallow layer of refined grains (<0.6 µm) was obtained with a low peak temperature (under 400 °C), whereas higher peak temperatures (>Ac 1 ) led to a change in grain size with different microstructures (fine-grained DP or martensite). Electron back-scattered diffraction characterizations revealed a large deformation in the as-received microstructures (mixture of ferrite and tempered martensite) induced by friction stir spot processing, leading to recrystallization and grain refinement around the stirred zone. Also, nano-indentation measurements showed a higher hardness than the hardness of the as-received DP980. Friction stir processing with different process conditions effectively changed microstructures and local mechanical properties.

36 MATERIALS SCIENCE↗
DOE OSTI2019

Atom Probe Tomography for understanding microstructural and chemical changes in nuclear materials

Atom Probe Tomography (APT) is routinely employed to determine 3D microstructure and chemistry changes in structural materials used in reactor to understand irradiation effect. These findings help to correlate changes to mechanical and physical properties of nuclear materials which are subjected to extreme reactor conditions. With safety and economics of a reactor in mind, it is essential to be able to predict the changes in the fuel composition and microstructure as a function of burnup. Also, there is current need to develop advanced fuels with improved properties and performance which requires the experimental data on changes in microstructure, physical and chemical properties after irradiation for feedback process. A novel approach is presented to determine the local burnup in irradiated fuels using isotopic quantification obtained by Atom Probe Tomography (APT). Considering the volume of sample used (<100?µm3) for APT experiments using the lift-out process in a scanning electron microscope equipped with a Focused Ion Beam (FIB), the presented method determines the local burnup from a nuclear fuel, where a minimal amount of waste is produced. In this work, three samples were analyzed with different burnup conditions to quantify the isotopes of 235U, 236U, and 238U for burnup calculation in the irradiated metallic U–7Mo dispersion fuel. The calculated values were found in excellent agreement with MCNP calculations. With regard to structural materials, one of the primary factors limiting the utilization of fuel rods of Zr alloys is hydrogen pick up during oxidation. Past experimental studies have shown that the hydrogen pickup in Zirconium alloys is influenced by the alloy composition as well by irradiation temperature and neutron dose. Oftentimes, it is not feasible to carry out comprehensive studies on understanding the microstructural evolution in neutron irradiated materials since in reactor irradiation programs are extremely costly, lengthy and logistically complicated. The use of ion irradiation as a substitute has grown in recent years for several reasons including less radioactivity, faster irradiations and lower costs. In the framework of the MUZIC 3 project, samples of Zircaloy-4 have been proton and Zr2+ irradiated at the Michigan Ion Beam Lab (MIBL) facility with aim to create a microstructure that is similar to that observed after in-reactor service. The temperature of the environment is adjusted to compensate for the higher dose rate, as has been done previously. As a first step in this project the irradiated microstructures have been characterized using Transmission Electron Microscopy and Atom Probe Tomography to elucidate the microstructural changes, including precipitate amorphization and dissolution and the formation of dislocation loops. This presentation plans to highlight initial finding of this project.

11 NUCLEAR FUEL CYCLE AND FUEL MATERIALS↗
DOE OSTI2021

SOC Synthetic Microstructure Bank

QUICK START: Start with property_library.html (can be found by typing the filename into the query box) and use the interactive table to filter, sort, and select a microstructure with the desired properties. Search for the alphabetic code to obtain the corresponding dataset. Full description: This is a bank of 1,970 unique 3-phase electrode microstructure files. When you account for reassigning phase IDs (e.g. declare that 1=Ni and 2=pore, instead of 1=pore and 2=Ni), it actually represents 5,910 unique electrode microstructures, each of which could be considered to be either an air or a fuel electrode (e.g. declare that the phase IDs correspond to pore, Ni, and YSZ; or that they correspond to pore, LSCF, and GDC; or whatever electron-conductor and ion-conductor combination is being studied). The voxel size is 50 nm and each electrode file contains a 4x4 grid of (12.5 micron)^3 sub-volumes. If placed together in a grid, they comprise a 50x50x12.5 micron electrode (note that the interfaces between sub-volumes will be sharp; this can be mitigated via simulating annealing/relaxation). The sub-volumes can also be used individually for a reasonably sized 12.5 micron cubic region-of-interest. A user can simply consider the voxel size to be a different value to rescale the volumes (and all of their morphological features, including particle size) as desired. These microstructures were generated using DREAM3D. The general procedure is outlined in https://doi.org/10.1016/j.jpowsour.2018.03.025 The file names are an alphabetic code having to do with the input parameters used in DREAM3D when they were generated. Most users would be best served by starting with the file property_library.html or property_library_subvols.html (which lists properties for each individual subvolume). These files contain a catalogue of the actual, measured properties of every microstructure in the database. Any combination of property values can be filtered and sorted until a desired electrode is found, at which point the user can find the file corresponding to that alphabetic code. The properties in the catalogue include connected TPB density, and for each phase: phase fraction, average particle size, polydispersity of particle size, tortuosity, and connected pair-wise interfacial area. They also include what fraction of each property is connected through to the interfaces of the volume. If the desired combination of properties is not found at first, remember that the phase IDs can be re-assigned arbitrarily, e.g. swapping 1s and 2s. In fact, the database was generated with this in mind so as not to generate redundant microstructures. If the database does not contain the desired property combinations, try to search for the other possible permutations of those properties with re-assigned phase IDs. Please cite https://doi.org/10.1149/10301.0909ecst for use. Please contact the maintainer, William K. Epting, for additional information or assistance.

3D microstructure↗
DOE OSTI2023

The effect of microstructure on the dynamic shock response of 1045 steel

Herein, the effect of microstructure on the shock response of 1045 steel is investigated via plate impact experiments and postmortem characterization. Three unique microstructures are explored: ferrite-pearlite, martensite, and ferrite with spheroidal cementite (i.e. spheroidized). Two spall recovery experiments, at approximate peak pressures of 3.2 and 3.5 GPa, are conducted to assess the Hugoniot elastic limit (HEL), spall strength, and damage morphology of the various microstructures. The ferrite-pearlite and martensite microstructures exhibit continuous yielding at both quasi-static and dynamic rates, while the spheroidized condition displays discontinuous yielding. Discontinuous yielding of the spheroidized microstructure is attributed to a combined low initial dislocation density coupled with a low dislocation nucleation rate. The spall strength of ferrite-pearlite is consistently lower than the spheroidized microstructure, attributed to elongated cementite that is more susceptible to cracking than more spherical cementite precipitates. Despite a high density of boundaries, martensite exhibits the highest spall strength. A large percentage of the boundaries within the martensite microstructure are found to be low energy (i.e. Σ3 or low angle), and are thus less susceptible to spall damage. Overall, the high spall strength of martensite is likely linked to traditional strengthening mechanisms that limit dislocation motion.

36 MATERIALS SCIENCE↗
DOE OSTI2022

Schmid factor crack propagation and tracking crystallographic texture markers of microstructural condition in direct energy deposition additive manufacturing of Ti-6Al-4V

Metallic additive manufacturing (AM) provides a customizable and tailorable manufacturing process for new engineering designs and technologies. The greatest challenge currently facing metallic AM is maintaining control of microstructural evolution during solidification and any solid state phase transformations during the build process. Ti-6Al-4V has been extensively surveyed in this regard, with the potential solid state and solidification microstructures explored at length. In this study, we evaluate the applicability of previously determined crystallo- graphic markers of microstructural condition observed in electron beam melting powder bed fusion (PBF-EB) builds of Ti-6Al-4V in a directed energy deposition (DED) build process. The aim of this effort is to elucidate whether or not these specific crystallographic textures are useful tools for indicating microstructural conditions in AM variants beyond PBF-EB. Parent β-Ti grain size was determined to be directly related to α-Ti textures in the DED build process, and the solid state microstructural condition could be inferred from the intensity of specific α-Ti orientations. Qualitative trends on the as-solidified β-Ti grain size were also determined to be related to the presence of a fiber texture, and proposed as a marker for as-solidified grain size in any cubic metal melted by AM. Analysis of the DED Ti-6Al-4V build also demonstrated a near complete fracture of the build volume, suspected to originate from accumulated thermal stresses in the solid state. Crack propagation was found to only appreciably occur in regions of slow cooling with large α +β colonies. Schmid factors for the basal and prismatic α-Ti systems explained the observed crack pathway, including slower bifurcation in colonies with lower Schmid factors of both slip systems. Colony morphologies and localized equiaxed β-Ti solidification were also found to originate from build pauses during production and uneven heating of the build edges during deposition. Tailoring of DED Ti-6Al-4V microstructures with the insight gained here is proposed, along with cautionary insight on preventing unplanned build pauses to maintain an informed and controlled thermal environment for microstructural control.

36 MATERIALS SCIENCE↗
DOE OSTI2022

Deep Neural Network Algorithm for CMC Microstructure Characterization and Variability Quantification

Microstructure characterization and variability quantification are crucial for understanding ceramic matrix composites (CMCs) mechanical behavior and deformation mechanisms across length scales. Traditionally, analyses of the micrographs obtained from microscopy are labor-intensive. However, with the vast improvement in computer vision (CV) and deep learning (DL), an automated algorithm can be designed to extract essential microstructure variability from micrographs which can then be used to construct a statistically representative volume element (SRVE). The DL-based algorithm spans the taxonomy of microstructure analyses, including semantic segmentation of microstructure constituents, secondary phases, matrix/fiber interface, and defects, and quantifying the microstructure variability in terms of probability distributions. In this work, C/SiNC and SiC/SiNC CMCs microstructures are semantically segmented through a deep convolutional neural network, followed by variability quantification through the implementation of a fully connected regression layer, hence forming a deep regression network. The deep regression network operates in a feedforward regime, in which the neuron output signal traverses through the network in a unidirectional manner. The weight tensor associated with each layer is updated through a backpropagation stochastic gradient descent approach. The input gray-scale image obtained through in-house scanning electron microscope and confocal microscope micrographs is augmented through affine transformations to increase the training set size, which is then processed through four strided convolutional layers. This compresses the image resolution by half at each layer while increasing the image depth by applying different filters (image encoding). The class activation maps (CAMs) corresponding to the applied filters highlight the key architectural features and assist with the semantic segmentation of the microstructure.

Hamza, Mohamed H.↗
DOE OSTI2020

Carbide Particle Redistribution in U-10Mo Alloy during Hot and Cold Rolling Processes: Synthetic Microstructure-Based Finite-Element Approach

The influence of initial carbide particle distribution features on particle fracture and the evolution of stringer distribution within an alloy of uranium with 10 weight percent molybdenum during cold rolling was investigated in this study. This work will provide guidance to casting optimization work and identify actionable process changes to eliminate the surface features directly attributed to carbides. In a previous study, various synthetic microstructure-based finite-element (FE) models were generated with consideration of different parameters (i.e., particle shape, particle volume fraction, and particle distribution), and used to examine the effects of hot rolling on the microstructure evolution. In this study, microstructures after ~80% hot rolling were used as the initial microstructures for the subsequent cold rolling simulation. Plane-strain compression loading was employed in the FE models to simulate cold rolling with reductions from ~80% to ~94%. Stringer analyses with the consideration of particle fracture were then performed with the simulated cold rolling results. The results of void evolution analyses show that microstructures with elliptical particles incorporate a much larger (almost double) void volume fraction than those with circular particles, indicating that the elliptical particles are fractured more easily than the circular ones. The results from the fracture mode analyses show that, under the adopted simple compression loading assumption, almost all the particles (i.e., >99%) were fractured by tensile separation. The results of stringer analyses show that, during cold rolling, the stringer volume fractions increase substantially due to the particle fracture, especially for the microstructures with the elliptical particles, and that the stringer volume fractions from stringers comprising many particles tend to increase faster. The maximum stringer lengths for the microstructures with circular particles tend to remain constant or decrease during cold rolling, whereas those for elliptical particles tend to slightly increase after a slight decrease. The number of particles at the surface increases very rapidly due to particle fracture during cold rolling.

36 MATERIALS SCIENCE↗
DOE OSTI2024

Update on Activities Related to the Library of Graphite Microstructures

This report provides an overview and update of the ongoing efforts to create a comprehensive library of microstructures for nuclear graphite and carbon-based materials under consideration for nuclear applications. The library includes data on microstructural characterization of unirradiated graphite materials, a guide to the techniques used to analyze graphite (which complements the ASME guidelines and ASTM standards), a summary of characterization data for neutron-irradiated or oxidized material, and a compendium of microstructural information for carbon-based materials. These efforts are being conducted at various length scales for the filler and binder phases in graphite to better understand graphite’s local structure and property relationships. This report is a follow-up to the previous milestone report titled Report on initial development of a database of nuclear graphite characteristics based on microstructural characterization, ORNL/TM/-2023/2992, published in July 2023.The effort to develop the library of microstructures supports the US Department of Energy Office of Advanced Reactor Technologies program objectives of aiding the material selection, licensing, management, and core assessments of a graphite core by documenting the unirradiated microstructure of relevant grades or characterizing the microstructure’s evolution under the reactor environment. Additionally, this project aims to provide (1) information and guidelines for the characterizing of graphite and (2) a protocol to assess a nuclear graphite grade.

36 MATERIALS SCIENCE↗
DOE OSTI2023

Microstructure and Temperature Dependent Indentation Response of Additively Manufactured Precipitation-Strengthened Al 0.3 Ti 0.2 Co 0.7 CrFeNi 1.7 High Entropy Alloy

The temperature-dependent [from room temperature (RT) to 500°C] nanoindentation behavior of a precipitation-strengthened Al 0.3 Ti 0.2 Co 0.7 CrFeNi 1.7 high-entropy alloy (HEA) processed via two different additive manufacturing (AM) techniques was investigated in the as-deposited and annealed conditions. The hierarchically heterogeneous microstructures were achieved via simple one-step annealing treatments, exploiting the residual stresses in the AM-processed HEA to partially recrystallize the microstructure, performed remarkably better than the nearly homogeneous microstructures in the as-deposited state. Further, the one-step annealed conditions revealed < 6.6% reduction in hardness values at 500°C compared to RT, while the as-deposited conditions showed a > 18% reduction in the hardness. The one-step annealed conditions also exhibited significantly higher hardness than the as-deposited conditions owing to their L1 2 -strengthened FCC microstructures. Furthermore, serrated yielding or the Portevin-Le Chatlier effect indicative of microstructural instability was observed during nanoindentation deformation (at 500°C) for the as-deposited conditions but not for the one-step annealed conditions. This, therefore, signifies the robustness of the hierarchically heterogeneous microstructures at elevated temperatures presenting a strong avenue for tuning the HEAs for future nuclear reactor applications.

36 MATERIALS SCIENCE↗
DOE OSTI2024

Automated Grain Boundary (GB) Segmentation and Microstructural Analysis in 347H Stainless Steel Using Deep Learning and Multimodal Microscopy

Austenitic 347H stainless steel offers superior mechanical properties and corrosion resistance required for extreme operating conditions such as high temperature. The change in microstructure due to composition and process variations is expected to impact material properties. Identifying microstructural features such as grain boundaries thus becomes an important task in the process-microstructure-properties loop. Applying convolutional neural network (CNN)-based deep learning models is a powerful technique to detect features from material micrographs in an automated manner. In contrast to microstructural classification, supervised CNN models for segmentation tasks require pixel-wise annotation labels. However, manual labeling of the images for the segmentation task poses a major bottleneck for generating training data and labels in a reliable and reproducible way within a reasonable timeframe. Microstructural characterization especially needs to be expedited for faster material discovery by changing alloy compositions. Here, in this study, we attempt to overcome such limitations by utilizing multimodal microscopy to generate labels directly instead of manual labeling. We combine scanning electron microscopy images of 347H stainless steel as training data and electron backscatter diffraction micrographs as pixel-wise labels for grain boundary detection as a semantic segmentation task. The viability of our method is evaluated by considering a set of deep CNN architectures. We demonstrate that despite producing instrumentation drift during data collection between two modes of microscopy, this method performs comparably to similar segmentation tasks that used manual labeling. Additionally, we find that naïve pixel-wise segmentation results in small gaps and missing boundaries in the predicted grain boundary map. By incorporating topological information during model training, the connectivity of the grain boundary network and segmentation performance is improved. Finally, our approach is validated by accurate computation on downstream tasks of predicting the underlying grain morphology distributions which are the ultimate quantities of interest for microstructural characterization.

36 MATERIALS SCIENCE↗
DOE OSTI2022

Computing diffraction patterns of microstructures from phase-field simulations

The diffraction pattern of a material contains information not only on the crystal structures of its constituting phases, but also on its mesoscale spatial distributions of phases, grains, and ferroelastic, ferroelectric, and ferromagnetic domains. While diffraction patterns from experiments such as X-ray diffraction are presented in the reciprocal or Fourier space, mesoscale microstructure models such as the phase-field method naturally produce real-space images of spatial distribution of chemical composition, structural, and ferroic domains. Although one could rather readily compute the Fourier amplitudes of chemical and structural domain distributions generated by mesoscale simulations, they only contain information about the length scale and alignment of the real-space chemical and structure domains. Therefore, a direct comparison between diffraction experiments and mesoscale microstructure simulations is not possible. Here, we develop a theoretical approach to directly compute the crystal diffraction patterns of microstructures predicted by phase-field simulations. In particular, we consider five representative examples of microstructure patterns involving purely compositional domains, a single pair of tetragonal twin structures, multiple twin variants in a hexagonal system, ferroelectric polar vortices, and polycrystalline grains. The results are compared with previous experimental observations as well as X-ray diffraction experiments performed in the present study. Furthermore, the theoretical framework allows one to directly connect material microstructures and diffraction patterns predicted from phase-field simulations and the corresponding diffraction patterns from experiments, and thus providing guidance to experimental diffraction characterization and interpretation of microstructures.

36 MATERIALS SCIENCE↗
DOE OSTI2020

Geometry-independent microstructure optimization for electron beam powder bed fusion additive manufacturing

The ability to manipulate the formation of microstructure is one of the potential advantages of additive manufacturing. Yet, the additive manufacturing process is riddled with complex interactions between processing parameters, geometry, and spot melting sequence which makes the task of controlling microstructure challenging. This paper proposes a method to solve the microstructure texture optimization problem by leveraging established optimization techniques and increasingly fast and accurate analytical heat transfer models. Respectively, we apply two different optimization algorithms, genetic algorithm and concurrent greedy line search, in order to induce arbitrarily shaped microstructure regions and we use a heat transfer model to evaluate the fitness of solutions in an iterative approach that is far more efficient than pure trial and error. The objective of the optimization algorithm is to control local thermal conditions – hence microstructure formation – during the melting process by manipulating the order and the dwell time of all the points of the infill pattern sequence. We then analyze the optimized sequence to determine important path characteristics that affect the microstructure formation. Through examples, we show that relative order and frequency, as measured by weighted average and variance of spot order, are important factors that influence the final characteristics.

36 MATERIALS SCIENCE↗
DOE OSTI2023

Leveraging the digital thread for physics-based prediction of microstructure heterogeneity in additively manufactured parts

A major limitation of additive manufacturing (AM) processes is that local conditions of material deposition frequently lead to unintentional heterogeneities in microstructure and properties within a single component, despite nominally uniform process conditions. Up to now, there has been no way to a priori determine the distribution of these heterogeneities, requiring expensive trial-and-error approaches to fabrication, testing, and characterization. Here, a physics-based framework for creating a digital representation of the laser powder bed fusion (PBF) process is proposed to predict the variation in solidification behavior that leads to heterogeneous microstructures in an as-built part. By leveraging in situ process data stored in the part’s digital thread, the scan path and process parameters were input into a heat transfer model which predicted solidification data at the melt pool scale. A two-step unsupervised clustering algorithm was used to first cluster the local solidification conditions (12.5µm 3 voxels) and then to cluster the regional behavior on the scale of multiple scan passes and print layers (250µm 3 super-voxels). This process was used to identify regions with similar solidification characteristics for multiple locations in a Stainless Steel 316-L component. The corresponding as-built part was sectioned and characterized using electron backscatter diffraction (EBSD). Quantitative analysis of the pole figures confirmed that the predicted regions of heterogeneity in the solidification conditions corresponded with differences in the observed microstructure. In conclusion, this work shows a viable path for estimating the microstructural heterogeneity for additively manufactured parts to either limit microstructural variation throughout a part or to enable functionality-based variation of the microstructure.

36 MATERIALS SCIENCE↗
DOE OSTI2024

Mesoscale modeling and semi-analytical approach for the microstructure-aware effective thermal conductivity of porous polygranular materials

Here we established a comprehensive modeling approach for investigating the microstructure-aware effective thermal conductivity ($κ_{eff}$) for porous microstructures containing solid particles and gaseous pores. Our approach combines the mesoscale computational modeling framework and the semi-analytical method, allowing for efficient prediction of $κ_{eff}$ for realistic porous microstructures, while considering complicated microstructural thermal conduction pathways effectively in the prediction. We used the diffuse-interface mesoscale computational model to generate extensive simulated $κ_{eff}$ data for realistic digital representations of microstructures with wide ranges of porosity ($f_p$), thermal conductivity of the gas phase ($κ_g$), and thermal conductivity of the solid phase ($κ_s$). From the simulated data, we identified two property variation regimes for $κ_{eff}$: (1) a slow $κ_{eff}$ increase for $κ_s ~ κ_g$; and (2) a faster $κ_{eff}$ increase for $κ_s \gg κ_g$. To capture the key features of the relationship between the microstructure and $κ_{eff}$, we derived a semi-analytical model by introducing structure and intensification factors. The two new factors incorporate the calibrated effective contribution of the solid volume with $κ_s$ and additional interfacial effects into the prediction of $κ_{eff}$, respectively, allowing for consideration of parallel, serial, and interfacial conduction mechanisms effectively. Using the selected simulation data, we quantified key model parameters within the semi-analytical model and verified that the parameterized model exhibits excellent agreement with simulated $κ_{eff}$ for the entire range of the parameter space.

36 MATERIALS SCIENCE↗
DOE OSTI2026

Inverse design of hypoeutectoid pearlite steel microstructures using a deep learning and genetic algorithm optimization framework

Goal-oriented microstructure design in metallic materials is a challenging task due to complex structure-property relationships. Traditional experimental and computational approaches are time-intensive and economically inefficient, limiting their applicability for large-scale design space exploration. Here, in this work, we propose an end-to-end framework that integrates deep learning models with genetic optimization to design microstructures with targeted mechanical properties. Deep learning models enable accurate forward design, while their integration with genetic optimization enables efficient inverse design within a few hours, compared to days or weeks using conventional finite element simulations. The framework combines experimental characterization and finite element modeling to analyze the influence of microstructural features on the mechanical behavior of hypoeutectoid steels. Data from both experiments and simulations are used to train the deep learning models. To demonstrate its effectiveness, we apply the framework to 0.63% carbon steel with proeutectoid ferrite and pearlite phases, commonly used in industrial applications. In this study, 2D microstructures were used for modeling, selected primarily for computational efficiency and to establish proof of concept. The framework successfully optimizes microstructures for targeted yield strength, ultimate strength, and stress concentration factors while significantly reducing computational time. Beyond hypoeutectoid steels, this scalable framework can be extended to other material systems and integrated with additive manufacturing, offering an efficient approach for accelerating microstructure design for specific engineering applications.

ConvLSTM↗
DOE OSTI2023

Microstructural diversity and digestion yields of select bituminous and subbituminous coals as raw material candidates for carbon fiber precursor production

Here, this work investigates the use of coals as raw materials for carbon fiber precursor production as a new alternative to coal utilization. Coal being a highly complex and heterogeneous material with different macerals and minerals complicates this task. Extensive microstructural characterization as well as preliminary digestion studies are performed on three bituminous coals (Herrin and Springfield from the Illinois Basin and the Blue Gem from the Central Appalachians) and one subbituminous coal from the Powder River Basin (Monarch), as raw material candidates. The subbituminous coal was richer in oxygen, lower in sulfur, and had a higher volatile matter yield. Microstructural investigations were performed using X-ray diffraction, X-ray/neutron computed tomography, scanning electron microscopy with energy dispersive spectroscopy, and petrology. Mild solvent extraction studies were conducted using creosote and decant oil as solvents in microreactors. The extraction yield was sensitive to temperature and time (350 to 450 °C between 30 and 120 mins.) for both creosote and decant oil digestions. While the Blue Gem coal had more desirable microstructural properties with less mineral content and cleaner macerals, it had the lowest coal conversion to quinoline soluble ‘liquid’ (of the bituminous coals). The lower coal conversion yield was hypothesized to be connected to the lack of FeS2 which could act as a catalyst when Fe is liberated from the structure under solvent extraction conditions. The Herrin and Springfield coals revealed similar microstructures and coal conversion efficiencies (higher than Blue Gem). These coals were the most promising candidates for further examination from this first approximations study. The subbituminous Monarch coal, however, was deemed less suitable due to poor coal conversion and less desirable microstructures. Additionally, the presented results with combined microstructural data from multiple length scales established a framework necessary for a first approximations study in this new coal utilization approach.

01 COAL, LIGNITE, AND PEAT↗
DOE OSTI2022

Microstructure-sensitive modeling of high temperature creep in grade-91 alloy

Predicting the effects of microstructure on high-temperature creep responses of steel components is critical to minimize the risks of failure and maximize economic viability in the energy sector. In this research, a recently developed advanced mechanistic constitutive model is employed to study the effect of microstructure on the creep responses and to rationalize experimentally reported variability in the performances of grade-91 alloy. Hundreds of experimental creep tests from the literature are used to assess the predictability of the model. The model proposed is shown to accurately predict the steady-state creep rates and also the less commonly considered, yet important, primary-to-secondary transients for a wide range of creep conditions. Further, the model is capable of extrapolating the creep response of grade 91 alloy even outside of the calibration regime. Using this model, the roles of initial microstructure described in terms of grain size, dislocation density, and precipitate content, on the creep responses are investigated. The effect of initial microstructure on the steady-state creep rates and creep-strain is found to vary with stress and temperature. Grain size plays a significant role in the low-stress creep, where the diffusional creep is dominant. On the other hand, the effects of dislocation densities and precipitate content are significant in the dislocation-plasticity-dominated high-stress and temperature regimes. Furthermore, by comparing model predictions against a large experimental creep database, it is found that variability in the creep responses can be rationalized on the basis of differences in the initial microstructure. Overall, this work provides a robust pathway for microstructure-aware engineering design, verification, and validation of metallic components for energy applications.

36 MATERIALS SCIENCE↗