Results for “range prediction” — Page 43

DOE OSTI2022

Vapor–liquid equilibrium estimation of n-alkane/nitrogen mixtures using neural networks

Understanding fluid phase behavior, like VLE, in high P&T conditions is crucial for developing high-fidelity simulations of chemically reacting flows in liquid-fueled combustion systems and also forms an integral part of the design-modeling of the control processes in chemical industries. Two data-driven models have been proposed here in this study, each of which was competent in estimating VLE for the Type III binary systems of C 10 /N 2 and C 12 /N 2 , at pressures ranging up to 50–60 MPa. Both models showed better performance in predicting equilibrium pressure as compared to VLE modeled using PR-EOS. A modified model has also been proposed, capable of estimating the full phase envelope for the binary systems of C 10 /N 2 and C 12 /N 2 across a wide range of temperatures, and thus exhibit the mixture critical pressure at the concerned temperature. The diverse applicability of the proposed network architecture was further exhibited while estimating the VLE of a ternary system of C 1 /C 10 /N 2 .

97 MATHEMATICS AND COMPUTING↗

DOE OSTI2024

Materials Characterization: A Primer for Solid Phase Processing Applications

The Pacific Northwest National Laboratory (PNNL) undertook the Materials Characterization, Prediction, and Control (MCPC) Laboratory Directed Research and Development (LDRD) Project to advance understanding of nuclear material processing and enable multifold acceleration in the development and qualification of new material systems produced via advanced manufacturing methods, such as solid phase processing, for use in national security and advanced energy applications (Smith 2021). As a two-year LDRD investment requiring focused research, the MCPC project applied only a subset of the wide range of available destructive and nondestructive characterization methods to provide data to the predictive modeling and data analytics tasks. The purpose of this report is to review a wide range of destructive and nondestructive characterization methods that are relevant in solid-phase processing (SPP) applications, but not necessarily applied in the MCPC Project as a guide to the planning of characterization activities in future research. Particular attention is given to measured characteristics that can correlate to other material characteristics, with a particular interest in nondestructive evaluation (NDE) that can be applied to samples obtained in the MCPC Project. Destructive examinations include tensile tests, optical and electron microscopy, micro-hardness, and residual stress tests. NDE tests include surface visual inspection, eddy current examination for cracks, 4-point potential drop, ultrasound, x-ray, and computed tomography.

36 MATERIALS SCIENCE↗

DOE OSTI2024

Vortex splitting in two-dimensional fluids and non-neutral electron plasmas with smooth vorticity profiles

Initially, elliptical, quasi-two-dimensional (2D) fluid vortices can split into multiple pieces if the aspect ratio is sufficiently large due to the growth and saturation of perturbations known as Love modes on the vortex edge. Presented here are experiments and numerical simulations, showing that the aspect ratio threshold for vortex splitting is significantly higher for vortices with realistic, smooth edges than that predicted by a simple “vortex patch” model, where the vorticity is treated as piecewise constant inside a deformable boundary. The experiments are conducted by exploiting the E × B drift dynamics of collisionless, pure electron plasmas in a Penning–Malmberg trap, which closely model 2D vortex dynamics due to an isomorphism between the Drift–Poisson equations describing the plasmas and the Euler equations describing ideal fluids. The simulations use a particle-in-cell method to model the evolution of a set of point vortices. The aspect ratio splitting threshold ranges up to about twice as large as the vortex patch prediction and depends on the edge vorticity gradient. This is thought to be due to spatial Landau damping, which decreases the vortex aspect ratio over time and, thus, stabilizes the Love modes. Near the threshold, asymmetric splitting events are observed in which one of the split products contains much less circulation than the other. These results are relevant to a wide range of quasi-2D fluid systems, including geophysical fluids, astrophysical disks, and drift-wave eddies in tokamak plasmas.

Physics↗

DOE OSTI2020

Towards improved speed and accuracy of laser powder bed fusion simulations via multiscale spatial representations

Due to the growing popularity of laser powder bed fusion (LPBF) as a metal additive manufacturing technique, there is a strong need to be able to accurately predict build outcomes. Full fidelity simulations of this process are not feasible due to the vast range of length and time scales inherent to it. While part-scale codes for simulating residual stress and distortion have shown reasonable predictive capability, they often neglect many aspects of the process occurring over smaller length/time scales, and thus are unable to capture effects of process parameter adjustments or the behavior of fine features. One way of capturing aspects at more refined length scales is through the use of adaptive mesh refinement (AMR). AMR allows for the process to be simulated at scales approaching the physical spatial dimensions without drastically increasing the total degrees of freedom in the simulation. This manuscript describes the implementation of an AMR algorithm within a multiphysics, parallelized finite element code, and its application to the LPBF problem. In this work, part-scale examples are provided where the use of AMR has allowed for higher fidelity thermal and thermomechanical simulations, as compared to experimental measurements. Results from these higher resolution simulations show that while AMR is a necessary component for increased accuracy in a computationally efficient manner, other improvements are also necessary, including handling of the multiple time scales inherent to the problem and the need for improved AM-specific material models.

42 ENGINEERING↗

DOE OSTI2024

Experimental and Computational Studies of Crystal Nucleation in Composition Gradients (Final Report)

The major goals of this project were to develop and validate a predictive nucleation model that incorporates composition gradients and is applicable to a large range of metallic systems which form both stochiometric and non-stochiometric compounds. Computationally we planned to expand the extent of thermodynamics-based theories and lay the groundwork to improve the general understanding and predictive capabilities of the role that gradients play in phase formation, glass formability and stability, and nucleation and growth events. To address these goals, we used Molecular Dynamics (MD) simulations and both isothermal and isochronal nanocalorimetric experiments on amorphous phases with a controlled composition gradient with the hopes of validating and improving the classical nucleation model and its use in solid solutions. While the computations were successful and identified ways to improve the classical nucleation model, the in situ nanocalorimetry studies proved very challenging due to unexpected difficulties in fabricating effective calorimeters and samples. Thus, we could not experimentally validate our predicted influence of composition gradients on nucleation. Nonetheless, important insights were gained and modifications to the classical nucleation theory were suggested.

36 MATERIALS SCIENCE↗

DOE OSTI2020

Proof-of-concept Talbot–Lau x-ray interferometry with a high-intensity, high-repetition-rate, laser-driven K-alpha source

Talbot–Lau x-ray interferometry is a grating-based phase-contrast technique, which enables measurement of refractive index changes in matter with micrometric spatial resolution. The technique has been established using a variety of hard x-ray sources, including synchrotron, free-electron lasers, and x-ray tubes, and could be used in the optical range for low-density plasmas. The tremendous development of table-top high-power lasers makes the use of high-intensity, laser-driven K-alpha sources appealing for Talbot–Lau interferometer applications in both high-energy-density plasma experiments and biological imaging. To this end, we present the first, to the best of our knowledge, feasibility study of Talbot–Lau phase-contrast imaging using a high-repetition-rate laser of moderate energy (100 mJ at a repetition rate of 10 Hz) to irradiate a copper backlighter foil. The results from up to 900 laser pulses were integrated to form interferometric images. A constant fringe contrast of 20% is demonstrated over 100 accumulations, while the signal-to-noise ratio continued to increase with the number of shots. Phase retrieval is demonstrated without prior ex-situ phase stepping. Instead, correlation matrices are used to compensate for the displacement between reference acquisition and the probing of a PMMA target rod. The steps for improved measurements with more energetic laser systems are discussed. The final results are in good agreement with the theoretically predicted outcomes, demonstrating the applicability of this diagnostic to a range of laser facilities for use across several disciplines.

Bouffetier, V. (ORCID:0000000160791260)↗

DOE OSTI2021

Data from: "Ecophysiological variation in two provenances of Pinus flexilis seedlings across an elevation gradient from forest to alpine"

This archive contains data used to support conclusions drawn in “Ecophysiological variation in two provenances of Pinus flexilis seedlings across an elevation gradient from forest to alpine”, by Reinhardt et al., 2011. Data were collected over one summer season in plots within the Alpine Treeline Warming Experiment (ATWE), before climate manipulations began. The experiment was located on Niwot Ridge, in the Front Range of the Colorado Rocky Mountains. This data package includes five comma-separated-values (.csv) files, five Microsoft Excel (.xlsx) files, one .pdf file, and two types of geospatial files: keyhole markup language (.kml), and ESRI shapefiles (.shp). .csv files can be opened using any simple text-editing software (such as Notepad and TextEdit), R, and Microsoft Excel. .xlsx files can only be opened using Microsoft Excel. The .pdf file can be opened using Adobe Acrobat Reader or any other compatible file viewing software. The .kml file can be opened using Google Earth and Google Maps, and shapefiles can be opened using any software compatible with the file type, such as ESRI’s ArcGIS suite and QGIS.Data archived contain gas exchange and plant physiology measurements, non-structural carbohydrate data, among others. Geospatial files are also provided for additional locational context. The files and their contents in this data package are summarized under "Data Summary" in the included Data User's Guide. All files (excluding geospatial) are available in both Microsoft Excel and in .csv format, and are indicated in the Data Summary list as well.-----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------Climate change is predicted to cause upward shifts in forest tree distributions, which will require seedling recruitment beyond current forest boundaries. However, predicting the likelihood of successful plant establishment beyond current species’ ranges under changing climate is complicated by the interaction of genetic and environmental controls on seedling establishment. To determine how genetics and climate may interact to affect seedling establishment, we transplanted recently germinated seedlings from high- and low-elevation provenances (HI and LO, respectively) of Pinus flexilis in common gardens arrayed along an elevation and canopy gradient from subalpine forest into the alpine zone and examined differences in physiology and morphology between provenances and among sites. Plant dry mass, projected leaf area and shoot:root ratios were 12–40% greater in LO compared with HI seedlings at each elevation. There were no significant changes in these variables among sites except for decreased dry mass of LO seedlings in the alpine site. Photosynthesis, carbon balance (photosynthesis/respiration) and conductance increased >2× with elevation for both provenances, and were 35–77% greater in LO seedlings compared with HI seedlings. There were no differences in dark-adapted chlorophyll fluorescence (Fv/Fm) among sites or between provenances. Our results suggest that for P. flexilis seedlings, provenances selected for above-ground growth may outperform those selected for stress resistance in the absence of harsh climatic conditions, even well above the species’ range limits in the alpine zone. This indicates that forest genetics may be important to understanding and managing species’ range adjustments due to climate change.

54 ENVIRONMENTAL SCIENCES↗

DOE OSTI2024

A physics-informed deep learning description of Knudsen layer reactivity reduction

A physics-informed neural network (PINN) is used to evaluate the fast ion distribution in the hot spot of an inertial confinement fusion target. The use of tailored input and output layers to the neural network is shown to enable a PINN to learn the parametric solution to the Vlasov–Fokker–Planck equation in the absence of any synthetic or experimental data. As an explicit demonstration of the approach, the specific problem of Knudsen layer fusion yield reduction is treated. Here, the predictions from the Vlasov–Fokker–Planck PINN are used to provide a non-perturbative solution of the fast ion tail in the vicinity of the hot spot, thus allowing the spatial profile of the fusion reactivity to be evaluated for a range of collisionalities and hot spot conditions. Excellent agreement is found between the predictions of the Vlasov–Fokker–Planck PINN and the results from traditional numerical solvers with respect to both the energy and spatial distribution of fast ions and the fusion reactivity profile, demonstrating that the Vlasov–Fokker–Planck PINN provides an accurate and efficient means of determining the impact of Knudsen layer yield reduction across a broad range of plasma conditions.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

DOE OSTI2024

A Physics-Informed Deep Learning Description of Knudsen Layer Reactivity Reduction

A physics-informed neural network (PINN) is used to evaluate the fast ion distribution in the hot spot of an inertial confinement fusion target. The use of tailored input and output layers to the neural network is shown to enable a PINN to learn the parametric solution to the Vlasov–Fokker–Planck equation in the absence of any synthetic or experimental data. As an explicit demonstration of the approach, the specific problem of Knudsen layer fusion yield reduction is treated. Here, the predictions from the Vlasov–Fokker–Planck PINN are used to provide a non-perturbative solution of the fast ion tail in the vicinity of the hot spot, thus allowing the spatial profile of the fusion reactivity to be evaluated for a range of collisionalities and hot spot conditions. Excellent agreement is found between the predictions of the Vlasov–Fokker–Planck PINN and the results from traditional numerical solvers with respect to both the energy and spatial distribution of fast ions and the fusion reactivity profile, demonstrating that the Vlasov–Fokker–Planck PINN provides an accurate and efficient means of determining the impact of Knudsen layer yield reduction across a broad range of plasma conditions.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

DOE OSTI2022

Application of the metabolic modeling pipeline in KBase to categorize reactions, predict essential genes, and predict pathways in an isolate genome

The DOE Systems Biology Knowledgebase (KBase) platform offers a range of powerful tools for the reconstruction, refinement, and analysis of genome-scale metabolic models built from microbial isolate genomes. In this chapter, we describe and demonstrate these tools in action with an analysis of isoprene production in the Bacillus subtilis DSM genome. Two different methods are applied to build initial metabolic models for the DSM genome, then the models are gapfilled in three different growth conditions. Next, flux balance analysis (FBA) and flux variability analysis (FVA) techniques are applied to both study the growth of these models in minimal media and classify reactions within each model based on essentiality and functionality. The models are applied with the FBA method to predict essential genes, which are then compared to an updated list of essential genes obtained for B. subtilis 168, a very similar strain to the DSM isolate. The models are also applied to simulate Biolog growth conditions, and these results are compared with Biolog data collected for B. subtilis 168. Finally, the DSM metabolic models are applied to explore the pathways and genes responsible for producing isoprene in this strain. These studies demonstrate the accuracy and utility of models generated from the KBase pipelines, as well as exploring the tools available for analyzing these models.

DOE knowledgebase↗

DOE OSTI2021

Application of the Metabolic Modeling Pipeline in KBase to Categorize Reactions, Predict Essential Genes, and Predict Pathways in an Isolate Genome

The DOE Systems Biology Knowledgebase (KBase) platform offers a range of powerful tools for the reconstruction, refinement, and analysis of genome-scale metabolic models built from microbial isolate genomes. In this chapter, we describe and demonstrate these tools in action with an analysis of isoprene production in the Bacillus subtilis DSM genome. Two different methods are applied to build initial metabolic models for the DSM genome, then the models are gapfilled in three different growth conditions. Next, flux balance analysis (FBA) and flux variability analysis (FVA) techniques are applied to both study the growth of these models in minimal media and classify reactions within each model based on essentiality and functionality. The models are applied with the FBA method to predict essential genes, which are then compared to an updated list of essential genes obtained for B. subtilis 168, a very similar strain to the DSM isolate. The models are also applied to simulate Biolog growth conditions, and these results are compared with Biolog data collected for B. subtilis 168. Finally, the DSM metabolic models are applied to explore the pathways and genes responsible for producing isoprene in this strain. These studies demonstrate the accuracy and utility of models generated from the KBase pipelines, as well as exploring the tools available for analyzing these models.

Allen, Benjamin↗

DOE OSTI2025

Bayesian model mixing with multireference energy density functional

Reliably predicting nuclear properties across the entire chart of isotopes is important for applications ranging from nuclear astrophysics to superheavy science to nuclear technology. To this day, however, all the theoretical models that can scale at the level of the chart of isotopes remain semiphenomenological. Because they are fitted locally, their predictive power can vary significantly; different versions of the same theory provide different predictions. Bayesian model mixing takes advantage of such imperfect models to build a local mixture of a set of models to make improved predictions. Earlier attempts to use Bayesian model mixing for mass table calculations relied on models treated at single-reference energy density functional level, which fail to capture some of the correlations caused by configuration mixing or the restoration of broken symmetries. In this study we have applied Bayesian model mixing techniques within a multireference energy density functional (MR-EDF) framework. We considered predictions of two-particle separation energies from particle number projection or angular momentum projection with four different energy density functionals—a total of eight different MR-EDF models. We used a hierarchical Bayesian stacking framework with a Dirichlet prior distribution over weights together with an inverse log-ratio transform to enable positive correlations between different models. We found that Bayesian model mixing provides significantly improved predictions compared to the participating models. Published by the American Physical Society 2025

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗

DOE OSTI2025

G2PDeep-v2: A Web-Based Deep-Learning Framework for Phenotype Prediction and Biomarker Discovery for All Organisms Using Multi-Omics Data

Multi-omics data offers rich insights into complex traits across organisms, yet integrating and analyzing these datasets for phenotype prediction and marker discovery remains challenging. Researchers need accessible tools that combine deep learning, hyperparameter optimization, visualization, and downstream analysis in a unified web platform. To address this, we developed G2PDeep-v2, a web-based platform powered by deep learning for phenotype prediction and marker discovery from multi-omics data across a wide range of organisms, including humans and plants. The server provides multiple services for researchers to create deep-learning models through an interactive interface and train these models using an automated hyperparameter tuning algorithm on high-performance computing resources. Users can visualize the results of phenotype and markers predictions and perform Gene Set Enrichment Analysis for the significant markers to provide insights into the molecular mechanisms underlying complex diseases, conditions and other biological phenotypes being studied.

59 BASIC BIOLOGICAL SCIENCES↗

DOE OSTI2021

Navigating Transition-Metal Chemical Space: Artificial Intelligence for First-Principles Design

Conspectus The variability of chemical bonding in open-shell transition-metal complexes not only motivates their study as functional materials and catalysts but also challenges conventional computational modeling tools. Here, tailoring ligand chemistry can alter preferred spin or oxidation states as well as electronic structure properties and reactivity, creating vast regions of chemical space to explore when designing new materials atom by atom. Although first-principles density functional theory (DFT) remains the workhorse of computational chemistry in mechanism deduction and property prediction, it is of limited use here. DFT is both far too computationally costly for widespread exploration of transition-metal chemical space and also prone to inaccuracies that limit its predictive performance for localized d electrons in transition-metal complexes. These challenges starkly contrast with the well-trodden regions of small-organic-molecule chemical space, where the analytical forms of molecular mechanics force fields and semiempirical theories have for decades accelerated the discovery of new molecules, accurate DFT functional performance has been demonstrated, and gold-standard methods from correlated wavefunction theory can predict experimental results to chemical accuracy. The combined promise of transition-metal chemical space exploration and lack of established tools has mandated a distinct approach. In this Account, we outline the path we charted in exploration of transition-metal chemical space starting from the first machine learning (ML) models (i.e., artificial neural network and kernel ridge regression) and representations for the prediction of open-shell transition-metal complex properties. The distinct importance of the immediate coordination environment of the metal center as well as the lack of low-level methods to accurately predict structural properties in this coordination environment first motivated and then benefited from these ML models and representations. Once developed, the recipe for prediction of geometric, spin state, and redox potential properties was straightforwardly extended to a diverse range of other properties, including in catalysis, computational “feasibility”, and the gas separation properties of periodic metal–organic frameworks. Interpretation of selected features most important for model prediction revealed new ways to encapsulate design rules and confirmed that models were robustly mapping essential structure–property relationships. Encountering the special challenge of ensuring that good model performance could generalize to new discovery targets motivated investigation of how to best carry out model uncertainty quantification. Distance-based approaches, whether in model latent space or in carefully engineered feature space, provided intuitive measures of the domain of applicability. With all of these pieces together, ML can be harnessed as an engine to tackle the large-scale exploration of transition-metal chemical space needed to satisfy multiple objectives using efficient global optimization methods. In practical terms, bringing these artificial intelligence tools to bear on the problems of transition-metal chemical space exploration has resulted in ML-model assessments of large, multimillion compound spaces in minutes and validated new design leads in weeks instead of decades.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

DOE OSTI2021

Modelling of a Flow-Induced Oscillation, Two-Cylinder, Hydrokinetic Energy Converter Based on Experimental Data

The VIVACE Converter consists of cylindrical oscillators in tandem subjected to transverse flow-induced oscillations (FIOs) that can be improved by varying the system parameters for a given in-flow velocity: damping, stiffness, and in-flow center-to-center spacing. Compared to a single isolated cylinder, tandem cylinders can harness more hydrokinetic energy due to synergy in FIO. Experimental and numerical methods have been utilized to analyze the FIO and energy harnessing of VIVACE. A surrogate-based model of two tandem cylinders is developed to predict the power harvesting and corresponding efficiency by introducing a backpropagation neural network. It is then utilized to reduce excessive experimental or computational testing. The effects of spacing, damping, and stiffness on harvested power and efficiency of the established prediction-model are analyzed. At each selected flow velocity, optimization results of power harvesting using the prediction-model are calculated under different combinations of damping and stiffness. The main conclusions are: (1) The surrogate model, built on extensive experimental data for tandem cylinders, can predict the cylinder oscillatory response accurately. (2) Increasing the damping ratio range from 0–0.24 to 0–0.30 is beneficial for improving power efficiency, but has no significant effect on power harvesting. (3) In galloping, a spacing ratio of 1.57 has the highest optimal harnessed power and efficiency compared with other spacing values. (4) Two tandem cylinders can harness 2.01–4.67 times the optimal power of an isolated cylinder. In addition, the former can achieve 1.46–4.01 times the efficiency of the latter. (5) The surrogate model is an efficient predictive tool defining parameters of the Converter for improved energy acquisition.

backpropagation neural network↗

DOE OSTI2023

Explainable AI (XAI)-driven vibration sensing scheme for surface quality monitoring in a smart surface grinding process

Local Interpretable and Model-agnostic Explanation (LIME), an explainable artificial intelligence (XAI) approach is adapted to identify the globally important time-frequency bands for predicting average surface roughness (Ra) in a smart grinding process. The smart grinding setup consisted of a Supertech CNC precision surface grinding machine, instrumented with a Dytran piezoelectric accelerometer attached to the tailstock quill along the tangential direction (Y-axis). For every grinding pass, vibration signatures were captured, and the ground truth surface roughness values were recorded using a Mahr Marsurf M300C portable surface roughness profilometer. The roughness values ranged from 0.06 to 0.14 microns over the complete set of experiments. Time-frequency domain spectrogram frames were extracted for each of the vibration signals collected during the grinding process. Convolutional Neural Networks (CNNs) were modeled to predict the surface roughness based on these spectrogram frames and their image augmentations. The best CNN model was able to predict the roughness values with an overall R2-score of 0.95, training R2-score of 0.99, and testing R2-score of 0.81 with only 80 sets of vibration signals corresponding to 4 experiments with 20 trials each. Although the data size is not large enough to guarantee such performance metrics in real-world scenarios, one can extract statistically consistent explanations underlying the relationships these complex deep learning models capture. Further, the LIME methodology was implemented on the developed surface roughness CNN model to identify the important time-frequency bands (i.e., the superpixels of a spectrogram) influencing the predictions. Based on the identified important regions on the spectrogram frames, the corresponding frequency characteristics were determined that influence the surface roughness predictions. The important frequency range based on LIME results was approximately 11.7 to 19.1 kHz. The power of XAI was demonstrated by cutting down the sampling rate from 160 kHz to 30, 20, 10, and 5 kHz based on the important frequency range and considering Nyquist criteria. Separate CNN models were developed for these ranges by only extracting time-frequency contents below their corresponding Nyquist cut-offs. A proper data acquisition strategy is proposed by comparing the model performances to argue the selection of a sufficient sampling rate to capture the grinding process successfully and robustly.

42 ENGINEERING↗

DOE OSTI2023

Reprint of: Explainable AI (XAI)-driven vibration sensing scheme for surface quality monitoring in a smart surface grinding process

Local Interpretable and Model-agnostic Explanation (LIME), an explainable artificial intelligence (XAI) approach is adapted to identify the globally important time-frequency bands for predicting average surface roughness (Ra) in a smart grinding process. The smart grinding setup consisted of a Supertech CNC precision surface grinding machine, instrumented with a Dytran piezoelectric accelerometer attached to the tailstock quill along the tangential direction (Y-axis). For every grinding pass, vibration signatures were captured, and the ground truth surface roughness values were recorded using a Mahr Marsurf M300C portable surface roughness profilometer. The roughness values ranged from 0.06 to 0.14 microns over the complete set of experiments. Time-frequency domain spectrogram frames were extracted for each of the vibration signals collected during the grinding process. Convolutional Neural Networks (CNNs) were modeled to predict the surface roughness based on these spectrogram frames and their image augmentations. The best CNN model was able to predict the roughness values with an overall R2-score of 0.95, training R2-score of 0.99, and testing R2-score of 0.81 with only 80 sets of vibration signals corresponding to 4 experiments with 20 trials each. Although the data size is not large enough to guarantee such performance metrics in real-world scenarios, one can extract statistically consistent explanations underlying the relationships these complex deep learning models capture. Further, the LIME methodology was implemented on the developed surface roughness CNN model to identify the important time-frequency bands (i.e., the superpixels of a spectrogram) influencing the predictions. Based on the identified important regions on the spectrogram frames, the corresponding frequency characteristics were determined that influence the surface roughness predictions. The important frequency range based on LIME results was approximately 11.7 to 19.1 kHz. The power of XAI was demonstrated by cutting down the sampling rate from 160 kHz to 30, 20, 10, and 5 kHz based on the important frequency range and considering Nyquist criteria. Separate CNN models were developed for these ranges by only extracting time-frequency contents below their corresponding Nyquist cut-offs. A proper data acquisition strategy is proposed by comparing the model performances to argue the selection of a sufficient sampling rate to capture the grinding process successfully and robustly.

47 OTHER INSTRUMENTATION↗

DOE OSTI2025

Next Generation Heat Transfer Fluids for Two-Phase Immersion Cooling of Data Centers

The purpose of this study is to evaluate the performance of next generation dielectric fluids in a Two-Phase Immersion Cooling (2PIC) system, which was designed for use in data centers. Hence, this report contains the performance evaluations of a new developmental dielectric fluid, Opteon™ 2P50, in a commercially available small-scale 2PIC system under typical and off-design range of operating conditions. Accordingly, ambient temperature and thermal loads were varied to simulate different ambient conditions. Additionally, this research report describes the development of a semi-empirical lumped model to predict the energy efficiency of the 2PIC system using Opteon™ 2P50 across a wide range of conditions. The model aims to offer a comprehensive understanding of the system’s efficiency and potential improvements. The outcomes of this study are expected to contribute to the adoption of sustainable 2PIC cooling technologies in data centers.

32 ENERGY CONSERVATION, CONSUMPTION, AND UTILIZATI↗