GIST Scholar
Not a member yet
30271 research outputs found
Sort by
Multiscale design of microwave absorbing sandwich structures via surrogate assisted topology optimization
No full textThis paper proposes a multiscale design framework for microwave absorbing sandwich structures (MASSs) that simultaneously optimizes macroscopic structural variables and microscopic unit cell geometry. The framework incorporates topology optimization to generate free-form microstructures within the sandwich core. To reduce computational cost and enable rapid exploration of diverse unit cell designs, a surrogate model using an artificial neural network (ANN) is employed. The multiscale optimization problem is formulated with the objective functions of minimizing the total mass and beam deflection of the MASSs. A constraint is imposed on the reflection loss to ensure sufficient microwave absorption at the target frequency. The formulated optimization problem is then solved using three metaheuristic optimization algorithms. To validate the performance of the optimal designs, high-fidelity finite element re-analyses are conducted. Structural validation is performed through 3-point bending and shear deformation analyses, while electromagnetic performance is evaluated using full-wave simulations. The re-analysis results confirm the effectiveness of the proposed multiscale design approach in achieving high-performance MASSs with tailored mechanical and electromagnetic characteristics. © 2025 Elsevier LtdFALSEsciescopu
Multi-User Semantic Communications with Interference-Mitigation Learning
No full textWhile semantic communications has shown great potential in single-user settings, interference over a shared multiple access channel (MAC) remains a key challenge in multi-user scenarios. In this letter, we propose a novel joint source-channel coding scheme with interference-mitigation learning (JSCC-IM) for task-oriented multi-user semantic communications. The proposed JSCC-IM employs a single-layer decoder to separate desired semantic features from multi-user interference in the aggregated signals over MAC. Then, we design a loss function that explicitly suppresses multi-user interference while preserving desired semantic features. Simulation results show that the JSCC-IM improves inference accuracy over conventional JSCC schemes by more than 5% and 3% in multi-user semantic inference and multi-view semantic fusion, respectively. © 2012 IEEE.FALSEsciescopu
Polarization-dependent terahertz emission from ART-assisted InP nanostructures monolithically grown on Si (001) substrates
No full textWe report polarization-dependent terahertz (THz) emission from aspect-ratio-trapping (ART)-assisted InP nanostructures monolithically grown on Si (001) using metal-organic chemical vapor deposition (MOCVD). Periodic SiO2 slabs laterally confine InP nanostructures over centimeter-scale areas. Transmission electron microscopy (TEM) and high-resolution X-ray diffraction (HRXRD) confirm high crystallinity and a fully relaxed lattice. Fs-laser-pumped THz time-domain spectroscopy reveals pronounced anisotropy. Excitation with the pump polarization perpendicular to the nanostructure axis yields more than twice the emission amplitude of the parallel geometry, with the integrated-field ratio saturating at 2.4. In the frequency domain, the perpendicular configuration also exhibits a higher peak frequency, indicating a shorter effective carrier lifetime. We attribute the enhanced emission to abrupt, repeated velocity modulations of photo-carriers that encounter nanoscale SiO2 boundaries transverse to transport, which amplifies the temporal derivative of the photocurrent via a lateral photo-Dember-type mechanism. These results establish ART-InP nanostructures on Si as a Si-compatible platform for compact, polarization-controllable THz emitters and provide clear geometric handles to engineer anisotropic THz responses. © 2025 Elsevier LtdFALSEsciescopu
Comprehensive characterization of lignin depolymerization products from woody biomass after thermoalkaliphilic laccase treatment
No full textIdentifying lignin-derived products is challenging due to their inherent structural complexity from diverse linkage types within lignin. A significant analytical hurdle is the matrix interference from oligosaccharides and lipids in lignin depolymerization products. Therefore, novel identification methods are needed to distinguish lignin-derived compounds from these interferences. In this study, we explored comprehensive strategies to identify lignin-derived products by thermoalkaliphilic laccase (CtLac, V243D) treatment of woody biomass. Our approach combined suspect screening with in silico MS/MS-based identification, integrating targeted and untargeted methods. By applying relative mass defect filtering specific to lignin and elucidating lignin-derived fragment ions, we isolated and characterized chemical features containing phenylpropanoid moieties from plant biomass matrix interferents. Suspect screening using LC-ESI-MS/MS analysis of CtLac-treated woody biomass identified 14 phenolic compounds. An in silico MS/MS database search using MS-FINDER yielded 27 and 18 tentative lignin depolymerization products from six woody biomass and cypress-OSL, respectively. Notably, V243D demonstrated superior performance, producing 5 lignin oligomers at significantly higher levels compared to CtLac. By providing high-coverage identification of degradation products, our approach lays a solid analytical foundation for lignin valorization. Taken together, this methodology advances the characterization of enzymatic lignin depolymerization products and holds potential for broader applications including chemical degradation processes. © 2026 The AuthorsTRUEsciescopu
Computational and Experimental Study on Hypersonic Turbulent Transition with Porous Surface
No full textTurbulent transition in the hypersonic boundary layer is computationally and experimentally investigated. The goal of the current study is to demonstrate the transition delay capability of porous surface. The current study explores a systematic approach comprising of computational stability analysis, precision fabrication of a porous test model, and experiment in a hypersonic shock tunnel. The stability analysis predicts transition delay on a well-chosen porous surface. The most effective porous surface is chosen from off-the-shelves perforated plates via the stability analysis. The chosen porous plate is cut and welded precisely to a sharp cone model with laser. Flow visualization in hypersonic experiments shows that the hypersonic boundary layer on the porous surface is laminar at least in the current visualization zone where the turbulent transition occurs on smooth surface.FALSEsciescopuskc
Communication Efficient Over-the-Air Federated Learning With Random FLARE Algorithm
No full textIn this letter, we propose a communication efficient federated learning algorithm, coined random FLARE (R-FLARE), using a novel error compensation method within a framework of random sparsification. In the R-FLARE, all devices sparsify the local gradients using a common set of randomly selected indices to improve communication efficiency with over-the-air computation. To upload local gradients, only the selected gradient elements are compensated by the local errors accumulated due to sparsification, which prevents redundant error compensation. We conduct a theoretical analysis on the convergence of R-FLARE using the l2 norm-based error compensation, which shows that it achieves the same convergence rate as the state-of-the-art algorithms. Numerical results show that the R-FLARE using l1- and l2-norm based error compensations outperform conventional algorithms in test accuracy and training speed. © 2025 IEEE.FALSEsciescopu
Increasing resolution and accuracy in sub-seasonal forecasting through 3D U-Net: the western US
No full textSub-seasonal weather forecasting is a major challenge, particularly when high spatial resolution is needed to capture complex patterns and extreme events. Traditional Numerical Weather Prediction (NWP) models struggle with accurate forecasting at finer scales, especially for precipitation. In this study, we investigate the use of 3D U-Net architecture for post-processing sub-seasonal forecasts to enhance both predictability and spatial resolution, focusing on the western U.S. Using the ECMWF ensemble forecasting system (input) and high-resolution PRISM data (target), we tested different combinations of ensemble members and meteorological variables. Our results demonstrate that the 3D U-Net model significantly improves temperature predictability and consistently outperforms NWP models across multiple metrics. However, challenges remain in accurately forecasting extreme precipitation events, as the model tends to underestimate precipitation in coastal and mountainous regions. While ensemble members contribute to forecast accuracy, their impact is modest compared to the improvements achieved through downscaling. The model using the ensemble mean and only the target variables was most efficient. This model improved the pattern correlation coefficient for temperature and precipitation by 0.12 and 0.19, respectively, over a 32 d lead time. This study lays the groundwork for further development of neural network-based post-processing methods, showing their potential to enhance weather forecasts at sub-seasonal timescales.TRUEsciescopu
Gravity-driven removal of tetracycline from water using a hierarchically porous adsorptive nanofibrous membrane system functionalized with metal–organic framework
No full textAntibiotics such as tetracycline (TC) persist in aquatic environments, posing serious risks to ecosystems and public health. Additionally, the excessive and incorrect use of TC has contributed to the development of antimicrobial resistance. Adsorption-based technologies are potential solutions for removing pollutants such as TC from water owing to their operational simplicity, energy efficiency, and compatibility with actual water systems. We fabricated a porous adsorptive nanofibrous membrane via growing ZIF-67 crystals in situ on amidoxime-functionalized polymer of intrinsic microporosity (AO-PIM-1) for removing TC from aquatic environments. The membrane has a hierarchically porous structure with many active sites for effectively capturing TC. The performance of the developed membrane was thoroughly evaluated in a real aquatic environment under static and gravity-driven dynamic conditions. The gravity-driven membrane (GDM) filtration with ethanol regeneration system required no additional energy input, offering an environmentally friendly and sustainable solution for TC removal. The developed membrane showed high TC removal efficiency (>99 %), strong anti-interference performance across various real water conditions, and exhibited minimal performance loss after eight GDM cycles and a total processing volume of 4000 L m−2. Its advanced material design incorporates practical water purification technologies to reduce the antibiotic contamination in engineered water systems. © 2025FALSEsciescopu
Comparative metabolomic and transcriptomic analysis of 2D and 3D mesenchymal stem cell cultures for improved therapeutic applications
No full textBackground and Purpose Mesenchymal stem cells (MSCs) are used widely in regenerative medicine due to their multipotency and immunomodulatory properties. Compared to conventional two-dimensional (2D) monolayer cultures, three-dimensional (3D) spheroid cultures better mimic the in vivo microenvironment, influencing the metabolic activity and therapeutic efficacy of MSCs. This study aimed to evaluate how 2D and 3D culture conditions affect the behaviour, proliferation, and functional properties of MSCs.Experimental Approach Metabolomic and transcriptomic analyses were conducted on MSCs cultured under 2D and 3D conditions. To assess the metabolic differences, polar metabolites were extracted and analysed using 1H-NMR spectroscopy. The data was processed with Chenomx and subjected to multivariate statistical analysis. For transcriptomic analysis, RNA sequencing was performed, followed by differential gene expression and gene set enrichment analysis.Key Results The findings reveal that MSCs in 3D spheroids exhibit reduced proliferation, enhanced stemness, and distinct metabolic adaptations, including increased glycolysis and altered nutrient metabolism. Additionally, genes associated with ribosome biogenesis and cell cycle progression were downregulated in 3D MSCs. These changes promote a more quiescent state, favouring its applications on tissue repair and immune modulation.Conclusions and Implications Understanding these metabolic adaptations offers valuable insights for optimising culture conditions, improving MSC-based therapies, and identifying novel therapeutic targets and biomarkers.TRUEsciescopu
Uncovering Nitro Compounds in Water: Photolysis-Based Analytical Methods and Insights into Their Formation during Ozonation
No full textNitro compounds, including aliphatic, aromatic, and
halogenated structures, are potentially toxic oxidation products formed
during water treatment, yet remain poorly characterized beyond a few
known species. This study presents two novel analytical methods for
nitro compound quantification without the need for compound-specific
standards: a batch UV254 photolysis coupled with colorimetry for total
nitro, and liquid chromatography with postcolumn photolysis (LC-
PCUV) for individual compounds. Both methods rely on photolytic
nitrite formation as a proxy for nitro compounds, yielding average molar
nitrite yields of 68 ± 17% for aliphatic compounds and 11 ± 11% for
aromatic compounds. The methods demonstrated high sensitivity in
complex matrices, achieving detection limits down to a few nM nitro-
equivalent when combined with solid phase extraction. A proof-of-concept ozonation experiment with model amines and amino acids confirmed the applicability of the methods, providing new insights into γ-aminobutyric acid ozonation. Notably, it is reported here for the first time that over 20% of amine-containing moieties in extracted natural organic matter can be converted to nitro during ozonation. Significant nitro concentrations (140−180 nM) were also detected in wastewater effluents and found to further increase upon ozonation. Together, these methods offer valuable tools to investigate the fate of reactive nitrogen moieties and nitro formation, including (halo)nitroalkanes during oxidative water treatment.FALSEsciescopu