Adaptive Spatial Resolution Research Articles

Numerical study of liquid sloshing using smoothed particle hydrodynamics with adaptive spatial resolution

Liquid sloshing is a typical problem of free surface flow in a number of engineering applications. The efficient simulation of liquid sloshing, particularly the large deformation and breakup of free surface, is a challenge for numerical methods. This paper presents a numerical method based on smoothed particle hydrodynamics with adaptive spatial resolution (SPH-ASR) for liquid sloshing. The SPH-ASR method can adaptively adjust the spatial resolution based on the distance to the free surface and the moving boundary. The numerical method was validated against experimental results from literature. Then the SPH-ASR method was applied to study the liquid sloshing problem in a prismatic tank and a rectangular tank. The effects of liquid fill level, external excitation frequency and amplitude on liquid sloshing were investigated. The pressure at different points on the tank wall and the total moment on the rotating shaft of the tank were compared and analyzed.

Simulation of binary collision of liquid drops using smoothed particle hydrodynamics with adaptive spatial resolution.

The binary collision of water drops in the air is studied by two-dimensional numerical simulation utilizing smoothed particle hydrodynamics with adaptive spatial resolution. The numerical method is validated by comparing the simulation with experiment. Three basic modes of equal-size drop collision are observed in numerical simulations at Weber number 3≤We≤120 and impact parameter 0≤x≤0.8, namely, reflexive separation, stretching separation, and coalescence collision. Based on the numerical results of different collision modes, the specific phenomena, evolution patterns, and physical principles are discussed. In particular, the detailed processes of the necking phenomenon and the propagation of surface wave in separation collision are obtained, corroborating the "end-pinching" theory proposed in the literature. At higher Weber numbers, the recoalescence of satellite drops is observed. The collision of unequal-size drops is also investigated. The effects of three dimensionless parameters, namely, drop diameter ratio, Weber number, and impact parameter are discussed. The physical mechanisms of some special phenomena are expressed in detail.

Adaptive Resolution Enhancement for Visual Attention Regions Based on Spatial Interpolation.

Resolution enhancement is crucial for human vision. However, it can be resource-consuming in the display pipeline. Therefore, there is a need to develop a lightweight resolution improvement algorithm specifically targeting visual attention regions. This paper presents a spatial-interpolation-based algorithm to improve the resolution of the visual attention area. The eye-tracking system consists of a near-infrared camera and an event camera is proposed to obtain the 3D gaze vector and eye moving trajectory. Secondly, the observation coordinates are obtained by gaze vectors, and the visual attention region is defined by the sensitive field-of-view angle. Then, interpolation-based adaptive spatial resolution enhancement and contrast enhancement adjustment are performed in the visual attention area. Finally, the feasibility of the proposed method is tested on both qualitative and quantitative dimensions. The experimental results demonstrate that the proposed method can significantly improve the visual effects.

Modal spectral element method with modified Legendre polynomials to analyze binary crossed gratings.

In a previous paper, a modal spectral element method (SEM), the originality of which comes from the use of a hierarchical basis built with modified Legendre polynomials, was shown to be very powerful for the analysis of lamellar gratings. In this work, keeping the same ingredients, the method has been extended to the general case of binary crossed gratings. The geometric versatility of the SEM is illustrated with gratings whose patterns are not aligned with the boundaries of the elementary cell. The method is validated by a comparison to the Fourier modal method (FMM) in the case of anisotropic crossed gratings and with the FMM with adaptive spatial resolution in the case of a square-hole array in a silver film.

Learning Spatiotemporal-Selected Representations in Videos for Action Recognition

Action recognition is a challenging task of modeling both spatial and temporal context. Numerous works focus on architectures modality and successfully make worthy progress on this task. While due to the redundancy in time and the limit of computation resources, several works focus on the efficiency study like frame sampling, some for untrimmed videos, and some for trimmed videos. With the intent of improving the effectiveness of action recognition, we propose a novel Computational Spatiotemporal Selector (CSS) to refine and reinforce the key frames with discriminative information in video. Specifically, CSS includes two modules: Temporal Adaptive Sampling (TAS) module and Spatial Frame Resolution (SFR) module. The former can refine the key frames in the temporal space for capturing the key motion information, while the latter can further zoom out some refined frames in the spatial space for eliminating the discrimination-irrelevant structural information. The proposed CSS is flexible to be embedded into most representative action recognition models. Experiments on two challenging action recognition benchmarks, i.e., ActivityNet1.3 and UCF101, show that the proposed CSS improves the performance over most existing models, not only on trimmed videos but also untrimmed videos.

Simulation of two successive drops impacting a heated wall by smoothed particle hydrodynamics with adaptivespatial resolution

<p indent="0mm">A numerical method for evaporating multiphase flow was developed by introducing a diffusion equation of species and a phase change model to smoothed particle hydrodynamics with adaptive spatial resolution (SPH-ASR). The method was applied to study the impact problem of two successive drops on a heated wall. The numerical method was validated against experimental results. Different factors were considered in the simulations, including drop diameter, impact velocity, wall temperature and wall inclined angle. The typical impact regimes were obtained, including spread, breakup, contact splash, rebound, and film splash. The numerical results indicated that the outcome of two drops impacting a solid wall mainly depends on the temperature and inclined angle of the wall, Weber number, and Reynolds number. Different phase diagrams of impact regimes were obtained using different parameters. A linear relationship between the critical Weber number of splash and the inclined angle of wall was obtained by data fitting.

Rigorous coupled wave analysis upgraded with combined boundary conditions method for 2D metamaterials in microwave.

We combine rigorous coupled wave analysis (RCWA) and combined boundary conditions method (CBCM) to analyze optical properties of two-dimensional periodic metamaterial grating with arbitrary shapes in microwave domain, and the adaptive spatial resolution (ASR) function is introduced to increase convergence especially for the crossed metamaterial gratings. The numerical results show that RCWA with CBCM can get the correct convergence result and distinguish the contribution of the grating lobes to the total transmission and reflection. The ASR technique improves dramatically the convergence speed of crossed grating. This method provides a more effective computational tool for the design of metamaterial grating in microwave domain.

Unknown terrain imaging with adaptive spatial resolution using UAV

Recently, Unmanned Aerial Vehicles (UAVs), or drones equipped with advanced cameras, have introduced a major revolution in surveying wide landscapes to collect ground data very fast. Capturing images of an unknown area with the desired resolution is of utmost importance for many real-life applications. For a UAV fitted with a given camera, its altitude from the ground surface primarily determines the spatial resolution of mapping. Hence, for an uneven terrain, it is a great challenge to maneuver the UAV dynamically to capture snapshots maintaining the desired spatial resolution throughout the area. This paper proposes a novel fully-automated technique for simple nadir drones to capture images of an unknown 3D terrain satisfying a predefined spatial resolution criterion. Also, an adaptive real-time flight path planning scheme is presented to minimize the distance traversed by the drone. A detailed performance evaluation study on real terrain data shows that the proposed technique can cover almost 97−99% of the area, maintaining the specified resolution with an almost 46−59% reduction in the path length compared to the conventional techniques.

SPH-ASR study of drop impact on a heated surface with consideration of inclined angle and evaporation

This paper studies the problem of drop impact on a heated surface by using a numerical method based on smoothed particle hydrodynamics with adaptive spatial resolution (SPH-ASR). The physical model considers the interaction and heat transfer between the liquid drop, solid wall and ambient gas, and the phase change between liquid and gas phases. The phase change process will change the particle mass and resolution. The SPH-ASR method is able to adaptively adjust the spatial resolution according to the distance to the interface between different phases. A series of numerical simulations were performed to investigate the effects of different factors on drop-wall interaction, including the drop diameter, impact velocity, liquid evaporation, wall temperature, and the inclined angle of wall. The comparison of impact regimes shows good agreement between the numerical and experimental results. The characteristics of drop impact on a heated surface were studied, particularly the drop spreading factor and mass evaporation ratio. The numerical results indicate that the maximum spreading factor and the mass evaporation ratio can be predicted as a function of dimensionless numbers.

2D rigorous coupled wave analysis with adaptive spatial resolution for a multilayer periodic structure.

We utilize the 2D modal solution of rigorous coupled wave analysis with adaptive spatial resolution to enhance the reliability of standard RCWA for multilayer gratings with metal-dielectric structures. The conversion relation in modal solutions between the Cartesian system and the transformed system is for the first time established. Based on the proposed conversion relation, the modal solution of the metal-dielectric structure obtained in the transformed system can match the boundary conditions with modal solutions of other different grating layers in the Cartesian system. Numerical results show that even for metal patches in microwave band, the above method can still achieve good convergence and is perfectly suitable for multi-layer structure calculation.

Efficient and accurate adaptive resolution for weakly-compressible SPH

In this paper we propose an accurate, and computationally efficient method for incorporating adaptive spatial resolution into weakly-compressible Smoothed Particle Hydrodynamics (SPH) schemes. Particles are adaptively split and merged in an accurate manner. Critically, the method ensures that the number of neighbors of each particle is optimal, leading to an efficient algorithm. A set of background particles is used to specify either geometry-based spatial resolution, where the resolution is a function of distance to a solid body, or solution-based adaptive resolution, where the resolution is a function of the computed solution. This allows us to simulate problems using particles having length variations of the order of 1:250 with much fewer particles than currently reported with other techniques. The method is designed to automatically adapt when any solid bodies move. The algorithms employed are fully parallel. We consider a suite of benchmark problems to demonstrate the accuracy of the approach. We then consider the classic problem of the flow past a circular cylinder at a range of Reynolds numbers and show that the proposed method produces accurate results with a significantly reduced number of particles. We provide an open source implementation and a fully reproducible manuscript.

Parallel adaptive weakly-compressible SPH for complex moving geometries

The use of adaptive spatial resolution to simulate flows of practical interest using Smoothed Particle Hydrodynamics (SPH) is of considerable importance. Recently, Muta and Ramachandran [1] have proposed an efficient adaptive SPH method which is capable of handling large changes in particle resolution. This allows the authors to simulate problems with much fewer particles than was possible earlier. The method was not demonstrated or tested with moving bodies or multiple bodies. In addition, the original method employed a large number of background particles to determine the spatial resolution of the fluid particles. In the present work we establish the formulation's effectiveness for simulating flow around stationary and moving geometries. We eliminate the need for the background particles in order to specify the geometry-based or solution-based adaptivity and we discuss the algorithms employed in detail. We consider a variety of benchmark problems, including the flow past two stationary cylinders, flow past different NACA airfoils at a range of Reynolds numbers, a moving square at various Reynolds numbers, and the flow past an oscillating cylinder. We also demonstrate different types of motions using single and multiple bodies. The source code is made available under an open source license, and our results are reproducible. Program summaryProgram Title: Parallel adaptive EDAC-SPHCPC Library link to program files:https://doi.org/10.17632/62rhbb9p4t.1Developer's repository link:https://gitlab.com/pypr/asph_motion.Code Ocean capsule:https://codeocean.com/capsule/4785489Licensing provisions: BSD 3-clauseProgramming language: PythonExternal routines/libraries:: PySPH (https://github.com/pypr/pysph), matplotlib (https://pypi.org/project/matplotlib/), NumPy (https://pypi.org/project/numpy/), automan (https://pypi.org/project/automan/), compyle (https://pypi.org/project/compyle/).Nature of problem: Simulating fluid flow around multiple moving bodies requires the local resolution to be automatically adapted in order to capture all the necessary flow features. If a single resolution is used throughout the domain, the number of particles would be excessively large.Solution method: We validate and demonstrate the accuracy of the adaptive particle refinement algorithm in simulating flow past moving multiple geometries. Our algorithm is fully parallel, and we provide an open-source implementation.

Towards convergence of turbulent dynamo amplification in cosmological simulations of galaxies

ABSTRACT Our understanding of the process through which magnetic fields reached their observed strengths in present-day galaxies remains incomplete. One of the advocated solutions is a turbulent dynamo mechanism that rapidly amplifies weak magnetic field seeds to the order of ∼$\mu$G. However, simulating the turbulent dynamo is a very challenging computational task due to the demanding span of spatial scales and the complexity of the required numerical methods. In particular, turbulent velocity and magnetic fields are extremely sensitive to the spatial discretization of simulated domains. To explore how refinement schemes affect galactic turbulence and amplification of magnetic fields in cosmological simulations, we compare two refinement strategies. A traditional quasi-Lagrangian adaptive mesh refinement approach focusing spatial resolution on dense regions, and a new refinement method that resolves the entire galaxy with a high resolution quasi-uniform grid. Our new refinement strategy yields much faster magnetic energy amplification than the quasi-Lagrangian method, which is also significantly greater than the adiabatic compressional estimate indicating that the extra amplification is produced through stretching of magnetic field lines. Furthermore, with our new refinement the magnetic energy growth factor scales with resolution following $\propto {\Delta x}_\text{max}^{-1/2}$, in much better agreement with small-scale turbulent box simulations. Finally, we find evidence suggesting most magnetic amplification in our simulated galaxies occurs in the warm phase of their interstellar medium, which has a better developed turbulent field with our new refinement strategy.

General and custom deep learning autosegmentation models for organs in head and neck, abdomen, and male pelvis.

To reduce workload and inconsistencies in organ segmentation for radiation treatment planning, we developed and evaluated general and custom autosegmentation models on computed tomography (CT) for three major tumor sites using a well-established deep convolutional neural network (DCNN). Five CT-based autosegmentation models for 42 organs at risk (OARs) in head and neck (HN), abdomen (ABD), and male pelvis (MP) were developed using a full three-dimensional (3D) DCNN architecture. Two types of deep learning (DL) models were separately trained using either general diversified multi-institutional datasets or custom well-controlled single-institution datasets. To improve segmentation accuracy, an adaptive spatial resolution approach for small and/or narrow OARs and a pseudo scan extension approach, when CT scan length is too short to cover entire organs, were implemented. The performance of the obtained models was evaluated based on accuracy and clinical applicability of the autosegmented contours using qualitative visual inspection and quantitative calculation of dice similarity coefficient (DSC), mean distance to agreement (MDA), and time efficiency. The five DL autosegmentation models developed for the three anatomical sites were found to have high accuracy (DSC ranging from 0.8 to 0.98) for 74% OARs and marginally acceptable for 26% OARs. The custom models performed slightly better than the general models, even with smaller custom datasets used for the custom model training. The organ-based approaches improved autosegmentation accuracy for small or complex organs (e.g., eye lens, optic nerves, inner ears, and bowels). Compared with traditional manual contouring times, the autosegmentation times, including subsequent manual editing, if necessary, were substantially reduced by 88% for MP, 80% for HN, and 65% for ABD models. The obtained autosegmentation models, incorporating organ-based approaches, were found to be effective and accurate for most OARs in the male pelvis, head and neck, and abdomen. We have demonstrated that our multianatomical DL autosegmentation models are clinically useful for radiation treatment planning.

FLatt Pack: A research-focussed lattice design program

Lattice structures are an important aspect of design for additive manufacturing (DfAM). They enable significant component light-weighting and the tailoring of a wide range of physical responses; mechanical, thermal, acoustic, etc. In turn, lattice design relies on fundamental research to uncover useful structure-property relationships, such as the influence of cell geometry and volume fraction. A number of commercial computer-aided-design (CAD) programs exist that offer lattice generation, but these tend to prioritise product design. This paper describes the FLatt Pack program (or Functional Lattice Package), which was created to address the paucity of research-focussed lattice design software. It possesses a number of features with this in mind, including; (i) it is free to use for research, (ii) it is standalone software with minimal, and also free, dependencies, and (iii) it undergoes frequent and rapid development based on state of the art lattice information and modelling methods. FLatt Pack includes twenty-three lattice cell types covering a broad range of pore connectivity, structural anisotropy, and surface area; a clear GUI presenting the lattice design stages in a sequential manner; and the option to export designs in appropriate formats for AM and finite element (FE) simulation. The program also features conformal lattice generation in arbitrary shapes, arbitrary volume fraction grading, and resource-efficient computation through an adaptive spatial resolution based on the user’s design choices. The most recent version of FLatt Pack is freely available at: www.github.com/ian27ax/FLatt_Pack_dist .

Smoothed particle hydrodynamics with adaptive spatial resolution for multiphase flows with large density ratio.

The smoothed particle hydrodynamics (SPH) method has been increasingly applied to simulate multiphase flows with large density ratios, but the method is computationally expensive when a high resolution is needed. Traditionally a uniform spatial resolution (USR) is used in an SPH simulation. In our previous work, an adaptive spatial resolution (ASR) method was developed for use in the SPH simulations of multiphase flows; the spatial resolution changes adaptively according to the distance to the interface between different phases. In this paper, the SPH-ASR method for multiphase flows is improved by introducing a particle shifting technique to improve the distribution of particles. This particle shifting technique considers the variable smoothing length. The present SPH-ASR method is further enhanced by optimizing the algorithm for adaptive resolution. In addition, the current SPH-ASR method is extended from two- to three-dimensional applications. The improved SPH-ASR method is validated by simulating the dam-break flows, liquid drop formation, and drop impact on a solid surface. Results show that the present SPH-ASR method can greatly reduce the computational cost while maintain the same accuracy to the uniform resolutions.

Smoothed particle hydrodynamics with adaptive spatial resolution (SPH-ASR) for free surface flows

A numerical method based on smoothed particle hydrodynamics with adaptive spatial resolution (SPH-ASR) was developed for simulating free surface flows. This method can reduce the computational demands while maintaining the numerical accuracy. In this method, the spatial resolution changes adaptively according to the distance to the free surface by numerical particle splitting and merging. The particles are split for refinement when they are near the free surface, while they are merged for coarsening when they are away from the free surface. A search algorithm was implemented for identifying the particles at the free surface. A particle shifting technique, considering variable smoothing length, was introduced to improve the particle distribution. The presented SPH-ASR method was validated by simulating various free surface flows, and the results were compared to those obtained using SPH with uniform spatial resolution (USR) and experimental data.

An Adaptive Spatial Resolution Method Based on the ST-ResNet Model for Hourly Property Crime Prediction

Effective predictive policing can guide police patrols and deter crime. Hourly crime prediction is expected to save police time. The selection of spatial resolution is important due to its strong relationship with the accuracy of crime prediction. In this paper, we propose an adaptive spatial resolution method to select the best spatial resolution for hourly crime prediction. The ST-ResNet model is applied to predict crime risk, with historical crime data and weather data as predictive variables. A prediction accuracy index (PAI) is used to evaluate the accuracy of the results. Data on property crimes committed in Suzhou, a big city in China, were selected as the research data. The experiment results indicate that a 2.4 km spatial resolution leads to the best performance for crime prediction. The adaptive spatial resolution method can be used to guide police deployment.

Coordinate transformation method for modeling three-dimensional photonic structures with curved boundaries

The coordinate transformation method (C method) is a powerful tool for modeling photonic structures with curved boundaries of discontinuities. As a modal method upon the Fourier basis, the C method has superior computational efficiency and rich physical intuitiveness compared to other full-wave numerical methods. But presently the C method is limited to two-dimensional (2D) structures if the boundaries between adjacent z-invariant layers are of generally different profiles [with (x,y,z) being the Cartesian coordinate]. Here we report a nontrivial extension of the C method to the general case of three-dimensional (3D) structures with curved boundaries of different profiles between adjacent layers. This extension drastically enlarges the applicability of the C method to the various interesting structures in nanophotonics and plasmonics. The extended 3D-C method adopts a hybrid coordinate transformation which includes not only the z-direction coordinate transformation in the classical C method but also the x- and y-direction matched coordinates adopted in the Fourier modal method (FMM), so as to exactly model the curved boundaries in all the three directions. The method also incorporates the perfectly matched layers (PMLs) for aperiodic structures and the adaptive spatial resolution (ASR) for enhancing the convergence. A modified numerically-stable scattering-matrix algorithm is proposed for solving the equations of boundary condition between adjacent z-invariant layers, which are derived via a transformation of the full 3D covariant field-components between the different curvilinear coordinate systems defined by the different-profile top and bottom boundaries of each layer. The validity of the extended 3D-C method is tested with several numerical examples.

Sniffing out fugitive methane emissions: autonomous remote gas inspection with a mobile robot

Air pollution causes millions of premature deaths every year, and fugitive emissions of, e.g., methane are major causes of global warming. Correspondingly, air pollution monitoring systems are urgently needed. Mobile, autonomous monitoring can provide adaptive and higher spatial resolution compared with traditional monitoring stations and allows fast deployment and operation in adverse environments. We present a mobile robot solution for autonomous gas detection and gas distribution mapping using remote gas sensing. Our “Autonomous Remote Methane Explorer” () is equipped with an actuated spectroscopy-based remote gas sensor, which collects integral gas measurements along up to 30 m long optical beams. State-of-the-art 3D mapping and robot localization allow the precise location of the optical beams to be determined, which then facilitates gas tomography (tomographic reconstruction of local gas distributions from sets of integral gas measurements). To autonomously obtain informative sampling strategies for gas tomography, we reduce the search space for gas inspection missions by defining a sweep of the remote gas sensor over a selectable field of view as a sensing configuration. We describe two different ways to find sequences of sensing configurations that optimize the criteria for gas detection and gas distribution mapping while minimizing the number of measurements and distance traveled. We evaluated an prototype deployed in a large, challenging indoor environment with eight gas sources. In comparison with human experts teleoperating the platform from a distant building, the autonomous strategy produced better gas maps with a lower number of sensing configurations and a slightly longer route.