The SARS-CoV-2 virus causes COVID-19, and several of its gene products have been successfully targeted for antiviral drug development, including the 3C-like protease (3CLpro). The substrate-binding site of 3CLpro exhibits significant conformational plasticity. While available X-ray crystal structures reveal substantial loop variability and molecular dynamics simulations indicate that conformational heterogeneity persists in ligand-bound complexes, the slow-timescale thermodynamic and kinetic landscape of these complexes in solution remains incompletely defined. Using 19F NMR spectroscopy, we characterize a slow-exchange conformational equilibrium in both the covalent nirmatrelvir (NMV) and noncovalent ensitrelvir (ENS) complexes of 3CLpro. For the wild-type 3CLpro–NMV complex at 298 K, joint dual-field line shape and exchange spectroscopy analysis resolves a millisecond-timescale exchange between two distinct states (population ratio ∼ 65:35; kex ≈ 52 s−1; ΔG‡ ≈ 15 kcal mol−1). Co-existing bound states also persist in both the noncovalent NMV–[C145A] mutant (minor state population, pB ≈ 18%) and wild-type ENS (pB ≈ 38%) complexes, demonstrating that this slow-exchange heterogeneity is an intrinsic property of the ligated protease rather than a consequence of covalent attachment. These dynamics are not readily explained by available crystal structures or conventional microsecond molecular dynamics simulations, suggesting that in solution the inhibited enzyme samples alternative, energetically accessible conformations that are not fully represented in the crystal lattice. This study highlights the utility of solution-state 19F NMR for quantifying low-energy conformational states relevant to drug–target energetics and inhibitor design.
For the integration of two-dimensional materials in future devices, a fundamental understanding of their response to external stimuli is needed. Toward this goal, we have investigated the electron and spin dynamics in the metallic van der Waals material Fe3GeTe2 (FGT) in its paramagnetic state after ultrafast optical excitation. To this end, we have employed a zone plate streaking technique with probing energies in the extreme ultraviolet range, tuned to the Fe M2,3 and Te N4,5 absorption edges. This approach provides insights into energy-dependent charge dynamics with a sensitivity to transient absorption changes on the order of ∼ 10 - 4 . We find a slow carrier relaxation time at both elemental edges-up to ( 2.2 ± 0.6 ) ps in Te and exceeding several picoseconds in Fe-which is surprising for a metal. To elucidate the complex time-resolved response, we also employ static x-ray absorption spectroscopy at the corresponding elemental edges, in which we find a double feature at the Fe M2,3 edge. We attribute this to different Fe sites in the pristine material and an oxidized surface layer, and we propose that the time-resolved absorption dynamics show a mixture of signals stemming from the different species. Additionally, we conducted time-resolved x-ray magnetic circular dichroism measurements in FGT at room temperature. We do not find clear evidence of the previously observed light-induced ferromagnetic order above T C . Our study lays the groundwork for a deeper understanding of charge and spin dynamics in FGT after optical excitation as part of a roadmap for 2D spintronics.
Masatsune Kainosho (Kai) was invited to contribute to this volume, but shortly before his death on December 28, 2025, he sent an apologetic email stating that he was too ill to comply. This is the type of contribution Kai might have written had the fates of the subject and narrator been interchanged. Kai and I were mutual scientific admirers, collaborators, and great friends for 49 years. Our wives had common interests in weaving and other crafts. We enjoyed visiting one another, traveling together, staying in each other's homes, and getting to know family members. We shared interests in the good things in life as well as life's milestones of marriage, birth, and death.
Initial drainage in coalbed methane (CBM) horizontal wells is critical, as adhesive micron-sized coal fines mix with water to form slurry, altering flow and bed formation. These effects impose long-term constraints on drainage strategy regulation during subsequent production. Existing studies overlook rheology changes in the mixed slurry caused by micron-sized coal fines. This limits guidance on bed evolution during the initial drainage phase. To this end, we experimentally determined the viscosity of the mixed slurry with 0%-6% micron-sized coal fines by mass. We then established a fitted relationship between viscosity and micron-sized coal fines mass fraction. On this basis, we used the Hertz Mindlin-Johnson Kendall Roberts contact model to describe the coal particles adhesion behavior. A Computational Fluid Dynamics-Discrete Element Method coupled model was developed, integrating slurry rheological characteristics with particle adhesion. This framework links experimental characterization to the flow patterns of graded coal fines in horizontal CBM wellbores. The results show that the viscosity of the mixed slurry is one of the key factors controlling the evolution of bed thickness and morphology. Its influence becomes more pronounced at higher drainage velocities. At a drainage velocity of 0.3 m/s, increasing the viscosity from 1.03 to 2.00 mPa s reduces the bed thickness by approximately 21.6%. The proposed model provides theoretical support for regulating drainage strategies during the initial drainage stage of CBM horizontal wells.
Double-diffusive convection of low-Prandtl-number liquid metals under magnetic fields is numerically investigated in a laterally heated rectangular cavity, with particular emphasis on the coupled effects of aspect ratio and magnetic-field orientation ( beta = 0 degrees , 90 degrees ). High-order numerical schemes are employed to resolve flow topology and transport characteristics over a wide range of aspect ratios ( A is an element of [ 1 , 12 ] ) at Ra = 10(4) and 10(5) . At the moderate Rayleigh number ( Ra = 10(4) ), the flow remains in steady or periodically oscillatory states for all aspect ratios considered, and the heat and mass transfer rates exhibit pronounced non-monotonic dependence on aspect ratio. These extrema are shown to be closely associated with transitions in the number of convection rolls, highlighting the role of flow structural reorganization in transport enhancement. At the higher Rayleigh number ( R a = 10(5) ), the flow dynamics become significantly more complex, with increasing aspect ratio leading to successive transitions from steady convection to periodic oscillations and eventually chaotic behavior. The onset of chaos is attributed to the combined effects of intensified buoyancy forcing and geometric stretching of the flow domain, which promote nonlinear interactions that cannot be fully suppressed by magnetic damping. The magnetic field orientation is found to play a crucial role in modulating flow stability and transport efficiency, with vertical magnetic fields sustaining more coherent convective motion and yielding higher heat and mass transfer rates than horizontal fields due to the anisotropic action of the Lorentz force.
The application of converging shock waves to compress solid materials can generate high-temperature and high-pressure environments, achieving a high compression ratio and high-density state, which is of significant importance for industrial applications. However, due to loading methods and material defects, converging shock waves often exhibit spatiotemporal non-uniform characteristics, leading to complex wave structures that reduce the compression performance of the converging shock waves. Therefore, it is crucial to study the propagation and evolution characteristics of non-uniform converging shock waves in solid media for achieving high-density and high-compression states in materials. Based on magnetically driven loading device, this study uses uniform cylindrical flyers to impact preabricated cylindrical targets with different amplitudes, generating non-uniform converging shock waves propagating inward. By employing 24-channel integrated laser velocimeter probes arranged in an offset pattern, the free-surface jump time and velocity are measured. Combined with numerical simulations, the study investigates the propagation characteristics of non-uniform converging shock waves and reveals the dynamic response characteristics of the interaction between non-uniform converging shock waves and materials. The research findings indicate that the intersection region of non-uniform converging shock waves exhibits an intersection enhancement effect, leading to a reverse growth characteristic of the free-surface spike amplitude. Unlike converging shock waves in gas, under the action of non-uniform converging shock waves and their derivative wave systems (free-surface reflection waves), the free surface exhibits spallation. Additionally, under the combined effects of convergence and shock wave intersection, the spallation and ejection of the working layer are more dispersed, the velocity of ejection are higher. Furthermore, shock wave intersections cause the free surface to accelerate, resulting in high-velocity points. Under the action of the derivative wave system of non-uniform converging shock waves, the free surface also exhibits secondary loading phenomena. These findings provide a certain foundation for the dynamic propagation study of non-uniform converging shock waves and hold theoretical significance for the development of compression science.
Ultrasonic wave propagation in multiphase fluid-saturated porous media involves complex interactions among fluid rheology, interfacial dynamics, and dispersed gas phases, particularly in the ultrasonic frequency range, where relaxation processes become significant. Although these phenomena are relevant to a range of engineering applications, including ultrasonic-assisted enhanced oil recovery, a consistent theoretical framework for fundamentally understanding and integrating these coupled mechanisms remains limited. In this study, we develop a continuum-mechanical framework for ultrasonic wave propagation in multiphase porous media by extending multiphase poroelastic wave theory. The proposed model incorporates heavy-oil viscoelasticity, dynamic capillary pressure effects, and the oscillatory response of dispersed gas bubbles within a unified linearized formulation. Based on the derived dispersion relations, three longitudinal wave modes are identified and analyzed in terms of their phase velocities, attenuation characteristics, and underlying physical mechanisms. The results reveal that propagation behavior is strongly frequency-dependent and controlled by coupled processes involving viscous dissipation, viscoelastic relaxation, interfacial dynamics, and bubble oscillations, with distinct modal responses governed by different physical mechanisms.
The horseshoe vortex (HV) generated around a bed-mounted cylinder is the dominant flow structure responsible for initiating local scour at pile foundations of bridges, offshore platforms, port terminals, and other coastal and hydraulic engineering structures. Despite existing knowledge on HV characteristics and their influencing factors, several key knowledge gaps remain. First, the Reynolds number can vary over a wide range from laminar to fully turbulent conditions, yet systematic investigations covering the entire spectrum are lacking. Second, existing studies have primarily focused on velocity and vortex strength, while other critical dynamic parameters—such as bed shear stress and pressure gradient—remain underexplored. Third, a universal mesh size criterion for high-fidelity simulations of the HV is lacking. To address these issues, this study employs the detached-eddy simulation method to perform high-resolution three-dimensional numerical simulations over Reynolds numbers (ReD) ranging from 100 to 106. The results reveal the multi-stage dynamics of the HV, characterized by four distinct regimes: laminar, transitional, weakly turbulent, and fully turbulent. The mechanical characteristics of the HV in each regime—including morphology, velocity and pressure fields, vortex strength, and bed shear stress—are identified, and engineering-oriented dimensionless empirical correlations are proposed to describe the variations of key HV parameters within the investigated Reynolds-number range. In addition, a mesh criterion is proposed based on grid independence tests.
Understanding the hydrodynamic forces on exposed particles above a particulate bed is crucial for predicting particle resuspension and sediment transport. This study uses the lattice Boltzmann method to quantify the forces and torque acting on particles located above structured porous beds exposed to a linear shear flow, as a function of bed structure and number density of exposed particles. Increasing gap distance between bed particles (increasing bed porosity) causes the exposed particles to sink downward in the bed, which decreases associated forces and torque due to enhanced shielding from upstream bed particles. Multiple exposed particles reduce the near-bed flow, so when multiple particles sink lower in a bed with large gap distances, the flow interference is reduced. This has the effect of increasing the local fluid velocity incident on the particles and increasing the particle forces and torque. These two opposing mechanisms largely offset one another, resulting in drag and lift forces on exposed particles that are nearly insensitive to bed porosity at sufficiently high exposed-particle number densities. A torque balance analysis of the particle rolling pivot axis demonstrates that the drag force is primarily responsible for incipient particle rolling, although the torque induced by the drag force decreases with increasing particle gap distance. A computation of the critical Shields parameter yields results within the range of experimentally observed values. The findings highlight the local gap distance between bed particles as a key parameter influencing the incipient motion of exposed particles, through its effects on both torque balance and the critical Shields parameter.
Existing formulas for predicting the equilibrium scour depth below pipelines are primarily derived from empirical fittings of laboratory data. Consequently, they suffer from significant scale effects and poor generalizability. To address this fundamental limitation, this study develops a physics-based theoretical model for predicting equilibrium scour depth below a pipeline under steady unidirectional currents and non-cohesive sandy seabed conditions. A generalized framework is first established via dimensional analysis and similarity principles. Subsequently, a theoretical formula is developed by integrating turbulence phenomenology within a first principles based scaling framework, applicable to both clear-water and live-bed scour conditions. The proposed model, grounded in first principles, integrates all key governing parameters, thereby effectively mitigating the inherent scale effects. High predictive accuracy was confirmed by validating the model against 310 experimental datasets. The model achieves excellent agreement in clear-water conditions, with its predictions for live-bed scour also remaining within acceptable error margins. A comprehensive parametric analysis for two representative pipeline diameters (D = 0.1 and 1.0 m) reveals the distinct influences of relative roughness, gap ratio, relative water depth, and Froude number on scour depth, demonstrating scale-dependent behaviors and providing process-based insights into the underlying physical mechanisms. This work provides a paradigm shift from purely empirical correlations to physics-based prediction, yielding valuable insights for pipeline scour predictions.
Accurate reconstruction and prediction of complex flow fields from sparse measurements remain a critical challenge in chemical engineering, particularly for nonlinear and multiphase systems. Classical linear reduced-order models, such as dynamic mode decomposition (DMD), often fail to capture strongly unsteady and nonlinear flow dynamics, while high-fidelity computational fluid dynamics (CFD) simulations are computationally prohibitive for real-time analysis and large-scale parametric studies. In this work, we evaluate a time-evolving neural operator (TENO) framework that learns the continuous spatiotemporal evolution of flow fields directly from practically limited and irregularly distributed observations. The performance of TENO is systematically evaluated on benchmark problems, from single-phase flows to an industrially relevant bubble column, where it consistently outperforms DMD in reconstruction and prediction accuracy. Compared with CFD simulations, TENO achieves relative errors on the order of
This research examines the effects of plunging motion on the formation of the dynamic stall vortex (DSV) over an airfoil and its mitigation using an optimized plasma actuator configuration. A closed-loop genetic algorithm (GA) optimization framework was employed to systematically determine the optimal location (x/c = 0.0-1.0) and duty-cycle frequency (f = 10-1000 Hz) of the plasma actuator for plunging airfoil with amplitudes of A degrees/c = 0.25-1.00 at Reynolds number of Re = 6.6 & times; 10(4). Numerical validation against particle image velocimetry confirms the accuracy of the plasma model in capturing induced-flow physics, revealing three distinct flow patterns at frequencies of 10, 100, and 1000 Hz. When the induced flow couples with the incoming velocity, the DSV is transformed, resulting in an attached shear layer. The GA identifies two amplitude-dependent flow control patterns: (1) At A degrees/c = 0.25 and 0.50, the reduced duty-cycle frequency (f(+)) is 37.2 and 38.8, which generates a stabilized vortex pattern that reattaches the separated flow and suppresses shear-layer fluctuations. (2) For A degrees/c = 0.75 and 1.00, the GA converges the f(+) to 4.4 and 5.6, which produces an intermixing vortex pattern that dissipates massive detached vortices into smaller, surface-attached structures. Concurrently, the optimal actuator location shifts systematically upstream from x/c = 0.045 to 0.015 with increasing A degrees/c, targeting progressively earlier separation onset. This upstream placement ensures the induced flow interacts with the incipient shear layer, suppressing DSV formation through enhanced mixing and momentum entrainment. These controlled flow patterns recover the boundary layer, reduce vortex circulation and shear-layer angle, and enhance the lift-to-drag ratio of the plunging airfoil.
The preliminary design of compressors depends upon a large number of parameters. Initially, the main characteristics and velocity triangles are defined by specifying the flow topology, i.e., whether the compressor is axial, mixed-flow, or centrifugal, the flow coefficient, and the loading coefficient, among other parameters. This paper presents a data-driven approach toward the creation of Smith charts for axial, mixed-flow, and centrifugal compressors on the basis of targeted three-dimensional computational fluid dynamics. The influence of the inlet hub-to-tip ratio, Mach number, the inlet Reynolds number, and the inlet boundary layer thickness on the Smith charts are analyzed. The results are applicable to the design of both single-stage and multi-stage compressors, such as the Mach number, Reynolds number, hub-to-tip ratio, and blockage vary from stage to stage. The compressors designed are then compared to the Cordier diagram. It is shown that the Mach number, Reynolds number, and hub-to-tip ratio significantly influence the Smith charts of compressors and limit the available design space in the Cordier diagram. In addition, changes in the suitable range of specific speeds of the topologies depending on the inlet conditions are shown. The diagrams presented in this paper serve as a guideline for the topology selection based on the flow coefficient and loading coefficient, considering the upstream flow conditions.
Helical flow is critical to cardiovascular hemodynamics, yet its transition to turbulence remains poorly defined, particularly under physiological pulsatile conditions. This experimental study systematically investigates the turbulent flow characteristics in helical tubes by varying curvature radius and torsion pitch using steady and high-pulsatility flows. Laser Doppler velocimetry and pressure measurements were used to quantify turbulence and flow resistance. Results suggest that under steady flow, the upstream flow exhibits the classical subcritical transition. Downstream, the helical geometry alters the mechanism, transforming intermittent turbulent “puffs” into a continuous, high-intensity turbulent field. Crucially, the curvature radius was the dominant stabilizing factor: large radii significantly suppressed downstream turbulence, while pitch effect was negligible. Friction factor results confirmed a higher energy cost to sustain secondary flows in the laminar regime, with a gradual increase during transition that supports the continuous turbulence development observed locally. Under pulsatile flow, flow stability is highly sensitive to the Pulsatility Index (PI). For PI > 3 and near-transition mean Re, intense bursts of turbulence were triggered. These instability bursts aligned specifically with the flow deceleration phase downstream, indicating that the helical geometry exacerbates deceleration-induced instability. These findings provide insights into biomedical device design and benchmarks for computational models studying the complex fluid dynamics of tortuous vascular structures.
While electric fields and ionic additives are traditionally considered the primary governing factors of electrohydrodynamic (EHD) cone-jets, this work demonstrates that dissolved CO2 acts as an effective molecular-scale regulator of jet formation and fragmentation. Through systematic molecular dynamics simulations, the EHD cone-jet process of ethanol is investigated under varying electric field strengths, NaCl concentrations, and dissolved CO2 conditions, which have been insufficiently explored at the molecular scale. The results show that increasing electric field strength accelerates Taylor cone formation and jet emission by enhancing electrostatic stresses. The addition of NaCl facilitates jet formation by improving charge transport and weakening hydrogen-bond interactions, thereby affecting jet stability and droplet dispersion. In contrast, dissolved CO2 plays a distinct role compared with electric field and ionic effects. Increasing CO2 concentration accelerates the onset of cone-jet formation and produces thinner jets with more frequent nanoscale fragmentation. Molecular-scale analyses reveal that dissolved CO2 reduces liquid structural compactness and disrupts hydrogen-bond interactions by altering local molecular arrangement and intermolecular spacing. These changes make the liquid interface more susceptible to deformation under an applied electric field, leading to earlier jet formation and enhanced droplet breakup. The coupled evolution of hydrogen-bond number, solvent accessible surface area, cluster number, and interaction energies consistently supports these mechanisms. These findings provide new molecular-level insights into how dissolved CO2 modifies the liquid structure and regulates EHD cone-jet dynamics, offering guidance for EHD cone-jet-based technologies.
This study experimentally investigates two key fluid mechanical behaviors in gas-liquid two-phase flow within a square-section serpentine microchannel: the curvature-induced satellite bubble formation mechanism and the axial decay of local mass transfer. Experiments were conducted using carbon dioxide (CO2) and an aqueous solution containing 0.4 wt. % sodium dodecyl sulfate. Various flow patterns were observed, including long slug-long slug flow (LS-LS), long slug-slug flow (LS-S), slug-slug flow (S-S), slug-droplet flow (S-D), droplet-droplet flow (D-D) and slug-satellite flow (S-SL). Notably, when the gas-iquid total volumetric flow rate exceeded 500 ml/h, a distinctive satellite bubble generation phenomenon was observed at the channel bend. High-speed imaging revealed the underlying mechanism: centrifugal force causes the bubble tail to detach from the inner wall, followed by neck thinning and rupture under the combined action of liquid-phase inertial stretching and capillary pinch-off. Analysis of flow parameters indicates that bubble volume and generation frequency are governed by gas-liquid shear competition, while the pressure drop is primarily controlled by the gas phase velocity. The liquid-side mass transfer coefficient is dominated by the liquid-phase velocity, reflecting the crucial role of liquid-phase internal circulation in interface renewal. To quantify the spatial evolution of mass transfer, the channel was divided into ten consecutive segments. The results show that the liquid-side local mass transfer coefficient first decays rapidly along the flow direction and then levels off, indicating that the effective mass transfer process is highly concentrated in the front section of the channel. These findings provide a physical basis for understanding interfacial instability and mass transfer localization in curved microchannels.
Mixed wettability surfaces strongly influence droplet departure in surface engineering applications. Although homogeneous superhydrophobic surfaces provide optimal removal performance, real surfaces typically exhibit heterogeneous wettability. Here, we investigate droplet departure mechanisms on smooth surfaces with alternating wettability patches using a multicomponent pseudopotential lattice-Boltzmann method. Gravity-driven pendant detachment and gravity-assisted coalescence-induced departure are examined over a wide range of heterogeneity length scales relative to droplet wetted base length ( a/w) and size ( a/R-eq). Results show that homogeneous surfaces generally exhibit lower critical Eotvos numbers ( Eo(crit)) than heterogeneous surfaces, with heterogeneity increasing the detachment threshold by approximately 5% and 16% for pendant and coalescence-induced departure, respectively. For pendant droplets, increasing a/w shifts detachment from continuous contact-line recession to pinch-off at a/w approximate to 1; in this regime, Eo(crit) increases monotonically up to 15% on net hydrophobic surfaces ( theta(eff)=141 degrees) but decreases by approximately 8% at larger patch length scales under near-neutral wetting ( theta(eff)=105 degrees). For coalescence-induced departure, a competition between droplet coalescence and contact line jumping governs detachment behavior when a/R-eq approximate to 0.06. In this regime, frequent stick-slip dynamics promote premature contact line jumping, which increases Eocrit by 20%. Under the Cassie framework of chemically inhomogeneous surfaces, strong heterogeneity can instead enhance droplet removal through controlled contact-line pinning, increasing droplet detachment performance up to 28%. A hysteresis analysis accounts for these effects of patch wettability contrasts, while a new proposed viscous-dissipation-based framework offers a preliminary mechanistic explanation for the influence of relative patch length on droplet departure.
Shelterbelts can effectively mitigate the aerodynamic deterioration of high-speed trains traversing tunnel-flat transition sections under crosswinds. An improved delayed detached-eddy simulation method combined with overset grid technology is employed to establish a numerical shelterbelt model incorporating realistic geometric features of tree trunks and canopies. The influence of tree spacing on temporal aerodynamic loads (AELs), surface pressure distributions, and flow field evolution is investigated, and a nonlinear mapping relationship between tree spacing and aerodynamic load amplitudes is established. The results demonstrate that shelterbelts significantly suppress abrupt fluctuations in aerodynamic loads. Compared with the case without a shelterbelt, the aerodynamic load coefficient amplitudes for the leading car are reduced by 13.61%-51.94%. The relationship between the amplitudes and tree spacing follows a logistic regression model, with the protective effect becoming negligible when the spacing increases to 8 m. The shelterbelt reduces the mean pressure coefficient difference across the leading car by 25.05% and inhibits the shedding of large-scale vortex structures on the leeward side, resulting in more uniform turbulent kinetic energy. It also promotes the dissipation of large-scale vortex structures into smaller scales, suppressing the broad low-pressure zone and enabling faster flow stabilization after the train leaves the transition section. The findings provide a theoretical basis and practical guidelines for the layout design of shelterbelts at high-speed railway tunnel portals.
Surrogate modeling has become a powerful approach for accelerating computational fluid dynamics (CFD) simulations, yet existing three-dimensional (3D) geometric learning models-based on meshes, point clouds, or voxels-remain fundamentally limited by resolution, memory cost, and dependence on explicit geometric discretizations. We introduce TripNet, a triplane-based neural framework that encodes 3D geometry into a compact, continuous, resolution-independent representation. Unlike traditional mesh-dependent models, TripNet supports query-based predictions at arbitrary spatial locations, enabling high-fidelity aerodynamic field estimation without relying on mesh connectivity or downsampled geometry. TripNet achieves state-of-the-art performance on the DrivAerNet and DrivAerNet++ datasets across three major aerodynamic tasks: drag coefficient prediction, surface pressure and wall-shear stress regression, and full 3D volumetric flow-field prediction. TripNet attains an R- 2 of 0.972 for drag prediction and a relative L2 error of 10.39% for velocity magnitude. TripNet reduces inference time by several orders of magnitude-predicting drag in 0.01 s and full 3D flow fields in 2 s on a single graphics processing unit-while improving accuracy and reducing memory cost compared to strong baselines such as FigConvNet, Transolver, and MeshGraphNet. Beyond in-distribution prediction, we demonstrate that triplane features encode robust, transferable geometric-aerodynamic priors. TripNet exhibits strong cross-shape generalization, outperforming prior works when trained on two car-body classes and tested on an unseen third shape. Furthermore, TripNet achieves data-efficient cross-dataset transfer: when pretrained on DrivAerNet++ and finetuned on the high-fidelity DrivAerML dataset, TripNet surpasses models trained solely on DrivAerML across all metrics, even with as few as 10-50 gradient steps. Collectively, these results highlight triplanes as a scalable, generalizable representation for partial differential equation surrogate modeling and establish TripNet as an accurate, efficient alternative to traditional CFD solvers and existing geometric deep learning models.
A series of underwater explosion experiments were conducted, where the morphological characteristics of explosion-induced water mounds were recorded, and the relationships between underwater explosion-induced wave loading (UEWL) and key explosion test parameters were analyzed. This study aimed to examine the damage characteristics of reinforced concrete (RC) piles subjected to underwater explosion-induced waves (UEIWs). First, a numerical model of UEIWs was developed using the Structured Arbitrary Lagrangian-Eulerian algorithm in LS-DYNA finite element software and validated by experimental data. Additionally, numerical models of underwater explosions were developed for both rigid cylindrical piles and RC cylindrical piles to analyze the distribution of UEWL around these structures. This study further investigates the dynamic response and failure characteristics of RC piles subjected to UEWL. The results indicate that wave impact on cylindrical piles is most significant at a dimensionless distance of 0.77. Compared with displacements induced by shock waves and bubble pulsation, those caused by UEWL accounted for only 5% and 1.8% at the mid-span, and 10.8% and 5.3% at the air-water interface, respectively. These results demonstrate that UEWL can still inflict noticeable secondary damage on RC piles, which is critical for the structural safety of high-pile wharves.
