Search NASA⌕ Search

Engineering topics

Liu, Yanfang

Publications and source records attributed to Liu, Yanfang.

12 records
All sourcesNASA NTRSDOE OSTI
DOE OSTI2024

A novel conditional generative model for efficient ensemble forecasts of state variables in large-scale geological carbon storage

Integrating monitoring data to efficiently update reservoir pressure and CO 2 plume distribution forecasts presents a significant challenge in geological carbon storage (GCS) applications. Inverse modeling techniques are commonly used to fuse observational data and refine reservoir model parameters, thereby improving state variable forecasts. However, these techniques often rely on linear or Gaussian assumptions, which can limit their effectiveness in accurately predicting state variables. Moreover, simulating large-scale three-dimensional (3D) GCS problems is computationally expensive, making iterative runs in inverse problems prohibitive. To address these challenges, we propose a conditional generative model utilizing the score-based diffusion method for real-time 3D pressure and saturation field distribution predictions. Our approach involves solving the score function with a mini-batch-based Monte Carlo estimator to generate labeled data. This data is subsequently employed to train a fully connected neural network, enabling it to learn the conditional sample generator within a supervised learning framework. This method enables the rapid generation of a large ensemble of predictions, facilitating comprehensive uncertainty quantification of state variables. Here we applied our method to forecast the dynamic 3D distributions of pressure and saturation fields over a 30-year injection period. The statistical assessment with low root mean square error (RMSE) values demonstrates that our method can accurately predict the spatiotemporal distributions of both pressure and saturation fields. Moreover, the developed conditional generative model shows high computational efficiency by generating 100 ensemble forecasts of 3D state variables in less than 10 min. The consistency between ensemble averages and ground truth values further illustrates the model’s capability to capture state variable dynamics during the CO 2 plume injection process. Notably, the ground truth values fall within the ensemble forecasts, indicating that our uncertainty quantification effectively captures variability and potential noise in the observations. Thus, the developed conditional generative model proves to be a more efficient, accurate, and practical tool for GCS applications, facilitating timely risk analysis and informed decision-making.

58 GEOSCIENCES↗
DOE OSTI2024

A Diffusion‐Based Uncertainty Quantification Method to Advance E3SM Land Model Calibration

Abstract Calibrating land surface models and accurately quantifying their uncertainty are crucial for improving the reliability of simulations of complex environmental processes. This, in turn, advances our predictive understanding of ecosystems and supports climate‐resilient decision‐making. Traditional calibration methods, however, face challenges of high computational costs and difficulties in accurately quantifying parameter uncertainties. To address these issues, we develop a diffusion‐based uncertainty quantification (DBUQ) method. Unlike conventional generative diffusion methods, which are computationally expensive and memory‐intensive, DBUQ innovates by formulating a parameterized generative model and approximates this model through supervised learning, which enables quick generation of parameter posterior samples to quantify its uncertainty. DBUQ is effective, efficient, and general‐purpose, making it suitable for site‐specific ecosystem model calibration and broadly applicable for parameter uncertainty quantification across various earth system models. In this study, we applied DBUQ to calibrate the Energy Exascale Earth System Model land model at the Missouri Ozark AmeriFlux forest site. Results indicated that DBUQ produced accurate parameter posterior distributions similar to those from Markov Chain Monte Carlo sampling but with 30 times less computing time. This significant improvement in efficiency suggests that DBUQ can enable rapid, site‐level model calibration at a global scale, enhancing our predictive understanding of climate impacts on terrestrial ecosystems.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2024

Advancing earth system model calibration: a diffusion-based method

Understanding of climate impact on ecosystems globally requires site-specific model calibration. Here we introduce a novel diffusion-based uncertainty quantification (DBUQ) method for efficient model calibration. DBUQ is a score-based diffusion model that leverages Monte Carlo simulation to estimate the score function and evaluates a simple neural network to quickly generate samples for approximating parameter posterior distributions. DBUQ is stable, efficient, and can effectively calibrate the model given diverse observations, thereby enabling rapid and site-specific model calibration on a global scale. This capability significantly advances Earth system modeling and our understanding of climate impacts on Earth systems. We demonstrate DBUQ's capability in E3SM land model calibration at the Missouri Ozark AmeriFlux forest site. Both synthetic and real-data applications indicate that DBUQ produces accurate parameter posterior distributions similar to those generated by Markov Chain Monte Carlo sampling but with 30X less computing time. This efficiency marks a significant stride in model calibration, paving the way for more effective and timely climate impact analyses.

Liu, Yanfang↗
DOE OSTI2024

Diffusion-Model-Assisted Supervised Learning of Generative Models for Density Estimation

Here, we present a supervised learning framework of training generative models for density estimation. Generative models, including generative adversarial networks (GANs), normalizing flows, and variational auto-encoders (VAEs), are usually considered as unsupervised learning models, because labeled data are usually unavailable for training. Despite the success of the generative models, there are several issues with the unsupervised training, e.g., requirement of reversible architectures, vanishing gradients, and training instability. To enable supervised learning in generative models, we utilize the score-based diffusion model to generate labeled data. Unlike existing diffusion models that train neural networks to learn the score function, we develop a training-free score estimation method. This approach uses mini-batch-based Monte Carlo estimators to directly approximate the score function at any spatial-temporal location in solving an ordinary differential equation (ODE), corresponding to the reverse-time stochastic differential equation (SDE). This approach can offer both high accuracy and substantial time savings in neural network training. Once the labeled data are generated, we can train a simple, fully connected neural network to learn the generative model in the supervised manner. Compared with existing normalizing flow models, our method does not require the use of reversible neural networks and avoids the computation of the Jacobian matrix. Compared with existing diffusion models, our method does not need to solve the reverse-time SDE to generate new samples. As a result, the sampling efficiency is significantly improved. We demonstrate the performance of our method by applying it to a set of 2D datasets as well as real data from the University of California Irvine (UCI) repository.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2023

Exceptional thermal stability of additively manufactured CoCrFeMnNi high-entropy alloy with cellular dislocation structures

CoCrFeMnNi high-entropy alloy (HEA) was additively manufactured (AM) by laser powder-bed fusion (L-PBF). The AM CoCrFeMnNi has prominent cellular dislocation structures with a small number of Mn-rich oxides. The thermal stability of the AM CoCrFeMnNi was investigated by isochronal annealing treatment at various temperatures from 400 to 1300°C for 1h. Microstructural analysis shows slow dislocation recovery, retarded recrystallization process, and precipitation of additional Cr-Mn based oxides during thermal annealing, resulting in exceptional thermal stability and retained high hardness at elevated temperatures. Further, by thermodynamic calculations, a low stored energy of 1.31 MJ/m 3 and a high activation energy of 353 kJ/mol for recrystallization were estimated for the AM CoCrFeMnNi. The exceptional thermal stability of the AM CoCrFeMnNi HEA is mechanistically attributed to the low crystallographic misorientations across the dislocation cell walls, sluggish atomic diffusion, and the pinning effects of the oxide nanoprecipitates.

36 MATERIALS SCIENCE↗
DOE OSTI2022

Strong yet ductile nanolamellar high-entropy alloys by additive manufacturing

Additive manufacturing produces net-shaped components layer by layer for engineering applications. The additive manufacture of metal alloys by laser powder bed fusion (L-PBF) involves large temperature gradients and rapid cooling, which enables microstructural refinement at the nanoscale to achieve high strength. However, high-strength nanostructured alloys produced by laser additive manufacturing often have limited ductility. In this work, we use L-PBF to print dual-phase nanolamellar high-entropy alloys (HEAs) of AlCoCrFeNi 2.1 that exhibit a combination of a high yield strength of about 1.3 gigapascals and a large uniform elongation of about 14 percent, which surpasses those of other state-of-the-art additively manufactured metal alloys. The high yield strength stems from the strong strengthening effects of the dual-phase structures that consist of alternating face-centred cubic and body-centred cubic nanolamellae; the body-centred cubic nanolamellae exhibit higher strengths and higher hardening rates than the face-centred cubic nanolamellae. The large tensile ductility arises owing to the high work-hardening capability of the as-printed hierarchical microstructures in the form of dual-phase nanolamellae embedded in microscale eutectic colonies, which have nearly random orientations to promote isotropic mechanical properties. The mechanistic insights into the deformation behaviour of additively manufactured HEAs have broad implications for the development of hierarchical, dual- and multi-phase, nanostructured alloys with exceptional mechanical properties.

36 MATERIALS SCIENCE↗
DOE OSTI2022

Removal of phosphorus using biochar derived from Fenton sludge: Mechanism and performance insights

Abstract A phosphorus removal biochar adsorbent was prepared from Fenton sludge. The adsorption process was optimized, and its phosphorus adsorption mechanism was discussed. It was found that the phosphorus adsorption performance of biochar prepared from single Fenton sludge (FBC‐400) was better than that of co‐pyrolysis of Fenton sludge and bamboo powder. The optimum condition was that Fenton sludge pyrolyzed at 400°C (FBC‐400). FBC‐400 had a larger specific surface area than that prepared by co‐pyrolysis with bamboo powder. And the high content of iron element could provide a higher surface charge of the biochar, thereby increasing the electrostatic adsorption of phosphorus onto FBC‐400. The phosphorus adsorption was highly pH dependent by FBC‐400, which can enhance electrostatic adsorption and increase adsorption capacity in acidic conditions. The effect of coexisting anion on adsorption performance was mainly affected by CO 3 2− , reducing the adsorption capacity by at least 49%, whereas other anions had no obvious interference. The adsorption process of FBC‐400 accorded with the pseudo‐second‐order kinetic model and the Langmuir model, which indicated that the adsorption process was monolayer adsorption and mainly chemical adsorption, and the maximum saturated phosphorus adsorption capacity was 8.77 mg g −1 . The mechanisms for phosphorus adsorption were electrostatic adsorption and inner‐sphere complexing. 1 M NaOH was used for desorption, and the adsorption capacity remained at 81% in the fifth cycle. Practitioner Points The Fenton sludge biochar usage as an adsorbent could be a win‐win strategy to convert waste biomass to valuable ‐ product. The adsorption process accorded with the Langmuir model, the maximum phosphorus adsorption capacity was 8.77 mg/g at 25°C. The adsorption mechanisms were electrostatic adsorption and inner‐sphere complexing. 1M NaOH was used for desorption, and the adsorption capacity remained at 81% in the fifth cycle.

Liu, Yanfang↗
DOE OSTI2021

Inactivated STAT5 pathway underlies a novel inhibitory role of EBF1 in chronic lymphocytic leukemia

B-cell chronic lymphocytic leukemia (CLL) is a disease caused by gradual accumulation of functionally incompetent lymphocytes. The majority of CLL cases are accompanied by chemoresistance. Early B cell factor 1 (EBF1) is a crucial contributor to B-cell lymphopoiesis. This study is to explore the effect of EBF1 on CLL cell progression and its involvement in regulating the signal transducers and activators of transcription 5 (STAT5) pathway. We conducted a correlation analysis between EBF1 and the clinical characteristics of CLL patients. Subsequently, EBF1 was overexpressed by transfection with EBF1 overexpression plasmid and the STAT5 pathway was also blocked by treatment with SH-4-54 in isolated CD20{sup +} B lymphocytes to investigate their roles in the regulation of cellular functions. STAT5, Janus kinase 2 (JAK2) expression and their phosphorylation levels were determined by quantitative PCR and Western blot analyses. The in vivo effects of EBF1 on tumor growth were evaluated using a xenotransplant model. Downregulation of EBF1 was observed in CD20{sup +} B lymphocytes of CLL patients. EBF1 overexpression disrupted the activation of STAT5 pathway, as evidenced by decreased expression and phosphorylation levels of STAT5 and JAK2. Furthermore, overexpression of EBF1 repressed viability and cell cycle entry, and increased apoptosis of CD20{sup +} B lymphocytes by inhibiting the STAT5 pathway. Finally, EBF1 exerted antitumor effects in nude mice. Overall, our study elucidates the inhibitory role of EBF1 in CLL through inactivation of the STAT5 pathway, which may provide new targets for CLL treatment.

60 APPLIED LIFE SCIENCES↗
DOE OSTI2020

Recent Measurements of Heavy Quarkonium Production in Au+Au, p+Au and p+p Collisions at STAR

Heavy quarkonia are useful probes of the quark-gluon plasma, where quarkonium dissociation is expected at high enough temperature. Indeed, such a suppression is clearly present in the latest STAR measurements of J/ψ production in Au+Au collisions. The suppression is observed to have little dependence on p T , although cold-nuclear matter (CNM) effects, such as shadowing, anti-shadowing and nuclear absorption can play a significant role at low p T . Measurements of the J/ψ production in p+A collisions can help us to disentangle the CNM effects from the hot medium effects in A+A collisions. In these proceedings, we will present the measurement of J/ψ production in p+Au collisions at √ s NN = 200 GeV by the STAR experiment. In addition, we will present recent STAR measurements of J/ψ and Υ production in p+p collisions including their dependence on the charged-particle multiplicity at mid-rapidity. Furthermore, these measurements are important for understanding the mechanism of the heavy quarkonium production in elementary nucleon-nucleon collisions and the interplay of soft and hard processes.

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗
DOE OSTI2020

3D printing of metal-based materials for renewable energy applications

Large-scale renewable energy must overcome conversion and storage challenges before it can replace fossil fuels due to its intermittent nature. However, current sustainable energy devices still suffer from high cost, low efficiency, and poor service life problems. Recently, porous metal-based materials have been widely used as desirable cross-functional platforms for electrochemical and photochemical energy systems for their unique electrical conductivity, catalytic activity, and chemical stability. To tailor the porosity length scale, ordering, and compositions, 3D printing has been applied as a disruptive manufacturing revolution to create complex architected components by directly joining sequential layers into designed structures. In conclusion, this article intends to summarize cutting-edge advances of metal-based materials for renewable energy devices (e.g., fuel cells, solar cells, supercapacitors, and batteries) over the past decade.

36 MATERIALS SCIENCE↗
DOE OSTI2020

Additive manufacturing of three-dimensional (3D)-architected CoCrFeNiMn high- entropy alloy with great energy absorption

In this study, we developed an approach to additive manufacturing of three-dimensional (3D)-architected CoCrFeNiMn high-entropy alloys by direct ink writing combined with thermal sintering. The 3D-architected CoCrFeNiMn lattices exhibit the outstanding energy absorption capacity that expands the envelope of existing architected materials. Such a high-energy absorption ability originates from the bend-dominated deformation mode of the 3D-architecture as well as the fully-annealed homogeneous microstructure of equiaxed grains, which collectively lead to remarkable strain hardening upon deformation. Our study provides a new avenue for 3D-printing advanced architected materials with extreme mechanical energy absorption for a myriad of structural applications.

mechanical property↗