CFD Research Corporation (United States)

Refractory high-entropy alloys (RHEAs) are an intriguing class of materials which can possess excellent strength at elevated temperatures, but often at the expense of oxidation resistance and ductility. Herein, we leverage a threefold approach of data-driven modeling, first-principles simulations, and experimental data to study the impact of minor alloying elements in CrMoNbTaV and HfNbTaVW, two RHEAs which possess excellent strength and ductility synergy, but lack sufficient oxidation resistance. An atomistic representation of the RHEA structure is obtained via calculation of the Warren-Cowley short-range order parameters. CrMoNbTaV with a representative coinage metal (Ag), platinum group metal (Pt), and metalloid (Si) are selected to study their elemental surface segregation tendencies and oxygen adsorption properties. The results reveal both strong surface segregation and a substantial impact on oxygen adsorption for each of the minor alloying elements, which is deemed to be beneficial for improving oxidation resistance. The trends in the modified RHEA's oxygen adsorption properties are attributed to the shift in the local atomic density of states at the RHEA surface. Finally, we discover that at atomic concentrations of 5% and above, minor alloying elements can have significant impacts on both the strength and ductility, and clear trends in the directionality of the observed changes in strength and ductility are observed based on the type of minor alloying element present.

Kriging (Gaussian Process Regression) is a powerful technique for creating surrogate models used in engineering design optimization when dealing with computationally expensive simulations. Despite its effectiveness, traditional Kriging implementations face computational challenges when scaling to large datasets, particularly for repeated inference operations required by optimization and uncertainty quantification applications. We present a high-performance PyTorch implementation of Kriging that significantly accelerates both training and inference while maintaining predictive accuracy. Our implementation leverages PyTorch's optimized tensor operations, automatic differentiation capabilities, and efficient memory management to achieve up to 20x speedup during inference with large datasets compared to traditional NumPy implementations. Benchmark results across datasets ranging from 100-2000+ training points demonstrate comparable interpolation accuracy with substantial runtime improvements, particularly for larger prediction set sizes. The results indicate that numerical stability is largely maintained between the NumPy implementation and PyTorch. This implementation is particularly valuable for optimization, uncertainty quantification, adaptive sampling, and other applications requiring repeated surrogate model evaluations. Our work demonstrates that modern deep learning frameworks can significantly benefit classical statistical learning algorithms beyond neural networks.

Refractory high-entropy alloys (RHEAs) are compositionally complex materials which have been demonstrated to have the potential for exceptional strength at high operating temperatures. However, their composition space is vast, and other property requirements, such as acceptable plasticity at room-temperature, must be met. Here, we leverage recently published, state-of-the-art deep learning models to predict compressive yield strength at 1,000 °C and room-temperature plasticity of >100,000 RHEAs. Multiple candidate materials were identified which exhibited exceptional balance between strength and plasticity. Upon experimental synthesis, multiple candidates were proven to outperform any previously reported RHEAs for simultaneous strength and plasticity. Our work demonstrates the power of data-driven approaches for rapid materials design, and enables continued multi-property optimization and materials discovery.

Free-space optical communication (FSOC) systems offer high-bandwidth and secure communication with minimal capital costs. Adaptive optics (AO) are typically added to these systems to decrease atmospheric channel losses; however, the performance of traditional AO wavefront sensors degrades in long-range, deep turbulence conditions. Alternative wavefront sensors using phase diversity can successfully reconstruct wavefronts in deep turbulence, but current implementations require bulky setups with high latency. In this work, we present a compact implementation of phase diversity wavefront sensing using a birefringent metasurface optic in conjunction with a convolutional neural network (CNN) for accurate wavefront reconstruction in deep turbulence. Our metasurface leverages both spatial and polarization multiplexing to simultaneously generate eight PSFs at the detector with various defocus levels. These PSFs are used as inputs to train the CNN in an end-to-end approach where the figure of merit is taken to be the power in bucket obtained at the receiver after propagating the predicted wavefront through the turbulent atmosphere. Our results show high accuracy reconstructions of scintillated wavefronts with Rytov numbers ranging from 0.2 to 0.6, surpassing the performance of traditional wavefront sensors like the Shack-Hartmann in these turbulence regimes. This approach provides new opportunities for compact, reliable wavefront sensing in deep turbulence, enabling extended range and improved accuracy for FSOC systems.

IntroductionSepsis is characterized by life-threatening organ dysfunction caused by dysregulated host response to infection. A key contributor is the disruption of neutrophil-endothelial interactions. Despite extensive research, there are no FDA-approved therapies that directly target altered neutrophil function in sepsis.MethodsWe previously identified three functionally distinct neutrophil phenotypes in sepsis patients: Hyperimmune, Hypoimmune, and Hybrid, using clinical profiling, organ-on-chip models, and proteomics. In this study, we applied bioinformatics tools to elucidate the molecular pathways and druggable targets associated with each phenotype. Differentially expressed proteins were identified using ExpressAnalyst, while pathway enrichment and modeling were performed via Metascape and KEGG-based analyses. DrugBank and the Broad Institute Drug Repurposing Hub were queried to identify FDA-approved therapeutics. STRING and Cytoscape were used to build protein–protein interaction networks and prioritize hub targets.ResultsIn our study, the Hyperimmune and Hybrid neutrophil phenotypes had similar numbers of upregulated proteins, while the Hypoimmune and Hybrid neutrophil phenotypes had approximately the same numbers of downregulated proteins. Functional enrichment analysis highlighted several biological processes and pathways that impacted adhesion/migration patterns, such as calcium transport and neutrophil degranulation. Neutrophil pathway analysis highlighted nine differentially expressed proteins that were directly implicated in known neutrophil processes related to sepsis, such as leukocyte transendothelial migration. These findings were leveraged to identify FDA-approved therapeutics that could be repurposed to target proteins within each phenotype highlighting the impact in normalizing altered neutrophil-related responses such as adhesion, migration and pro-inflammatory mediator release. Finally, a protein-protein interaction network was employed to prioritize these target proteins within each phenotype using network analysis and identified three distinct drug targets across phenotypes that could modulate the neutrophil response in sepsis: VTN in the Hybrid phenotype, TRPV2 in the Hypoimmune phenotype and H2AC21 in the Hyperimmune phenotype.DiscussionOur integrative approach highlights phenotype-specific drug targets and FDA-approved candidates to modulate dysfunctional neutrophil responses in sepsis. This strategy supports a precision medicine framework for repurposing existing drugs based on neutrophil functional phenotyping.

One of the most significant future needs for space technology development and infrastructure is for robust, low cost and mature systems that can provide on-demand power, “multi-mode” propulsion systems for missions, and thermal management. Increasingly lengthy missions coupled with variable “response time to orbit” needs challenge tandem flexibility requirements. This places a premium on development of storable propellants with high-performance, and if possible, dual use applications to produce electrical power and balance thermal system applications. Based on previous efforts for DARPA and working with NASA MSFC, it was discovered combining nitrogen tetroxide and nitrous oxide would yield an innovative self-pressurizing, non-cryogenic liquid oxidizer, with propulsion performance characteristics approaching that of liquid oxygen. Called Marshall Enriched Storable Oxidizer (MESO), it can also be used to generate electricity in a fuel cell. And based on current efforts for the U.S. Army, the ability to provide for breathable, synthetic air from the decomposition of nitrous oxide can also utilize the enthalpy generated to drive a thermoelectric generator. Combinations of propulsion, power, and even life-support system ramifications, coupled with thermal capacity, could be beneficial for deep space exploration. This paper touches on lunar / deep space manned or robotic mission aspects for the seemingly simplistic needs of keeping systems warm, while simultaneously enabling operations needing propulsion or power. A goal is that by describing current efforts, it will initiate dialog at ASCEND and beyond for future synergistic and tandem collaborations, eventually yielding successful off-world activities for humanity.

This paper is a contribution to a special session focused on the analysis and optimization of a vehicle inspired by the Stratolaunch Talon–A vehicle, the baseline case of which we will refer to as Talon-P. Mach 6.0 flow over a generic boost glide vehicle was investigated through laminar CFD simulations. Comparisons were made between simulated results and experimental data collected in Purdue University’s BAM6QT hypersonic tunnel. Similar surface heating patterns, such as streaks on the lower and upper surfaces of the vehicle, were observed both in CFD simulations and experimental infrared imaging. Six different nose geometries were investigated, and overall good agreement between the experimental and simulated results was observed. Two configurations of the Talon-P experienced substantial disagreement between simulated predictions and experimental results; possible reasons for these discrepancies are discussed. The effect of nose geometry on downstream flow behavior is also investigated, with special emphasis on the vortical structures which form near the nose on the bottom surface of the geometry. The CFD results presented in this paper are part of a larger design, downselection, and analysis process.

─The National Aeronautics and Space Administration (NASA) is currently working to return humans to the moon. multiple space vehicles will be utilized. In the execution of this, a plethora of in-space, dynamic events will occur, many of which require on-board crew to be awake and alert. During system and mission design, vehicle flight profiles are calculated and disseminated with a crew awake-sleep profile developed as an end product of analysis cycles. This singular crew awake-sleep profile is then integrated into the following design analysis cycle for assessment by the medical operations community. To enable faster roundtrip design and analysis of flight trajectory with crew awake-sleep implications, a novel approach and tool, HLS Yawn Prevention 'N' Outer Space (HYPNOS), was created for the Human Landing System (HLS) program out of Marshall Space Flight Center (MSFC) and has been used recently to great effect. Instead of manually crafting the crew sleep timeline based on pre-determined sleep requirements and other flight rules, these constraints are ingested into the MATLAB-based tool. The tool then automatically generates an initial assessment of a sleep schedule that violates the fewest crew requirements and / or maximizes crew sleep and performance. This is performed for the hundreds of flight trajectories that are available to ingest. The quality of these sleep shift profiles is also aggregated into an overall ‘ “heatmap ”, presenting a distribution of the best options and suggesting what the stressing requirements are. Analysis assessments are delivered to mission planning Subject Matter Experts, the medical operations community, and the Flight Operations Directorate (FOD) periodically to aid in the Integrated Analysis Cycles. This integrated approach to design and analysis of crew sleep schedules with the HYPNOS tool is reminiscent of assessments performed by the NASA Jet Propulsion Lab's (JPL) Team X and the European Space Agency's (ESA) Concurrent Design Facility (CDF). The HYPNOS tool is reducing the risk and cost of complicated manned space mission planning for NASA in the HLS program.

Cryogenic fluids, while offering higher energy density than storable propellants, present several handling challenges, including gradual boil-off and tank self-pressurization. Transferring these propellants between tanks can be problematic, as the receiving tank's temperature and pressure often lead to flashing and rapid boil-off upon injection. The No-Vent Fill (NVF) technique mitigates these risks by preventing liquid fuel from being expelled overboard. By controlling tank pressure and ensuring appropriate condensation in the ullage region, the NVF process avoids exceeding vent pressure limits. In this study, Loci/STREAM-VOF simulations were compared with data from NASA K-site test runs on a 2.1 cubic meter tank with a bottom orifice under various no-vent fill conditions, including different wall temperatures, flow rates, and inlet liquid temperatures. The effort demonstrated that Loci/STREAM-VOF is capable of quantitatively capturing the important physics of NVF with submerged jets: (1) validation of the liquid-vapor interface turbulence velocity using previous experimental data showed good agreement at different fill levels; (2) validation of the CFD Stanton number for mass transfer provided an adequate mass transfer coefficient range; (3) consistent pressure agreements were found in the pressure history for both cold and warm conditions; and (4) the variations of final pressure with flow rate and inlet liquid temperature were well predicted by the CFD simulation. The application of NVF for low-gravity conditions will be studied in future work.