Box Aspect Ratio Research Articles

Statistics of kinetic and thermal energy dissipation rates in two-dimensional thermal vibrational convection

We investigate the statistical properties of kinetic ϵu and thermal ϵθ energy dissipation rates in two-dimensional (2D) thermal vibrational convection (TVC). Direct numerical simulations were conducted in a unit aspect ratio box across a dimensionless angular frequency range of 103≤ω≤107 for amplitudes 0.001≤a≤0.1, with a fixed Prandtl number of 4.38. Our findings indicate ϵu is primarily associated with the characteristics of the vibration force, while ϵθ is more related to the large-scale columnar structures. Both energy dissipation rates exhibit a power-law relationship with the oscillational Reynolds number Reos. ϵu exhibits a scaling relation as ⟨ϵu⟩V,t∼a−1Reos0.93±0.01, while ϵθ exhibits two distinct scaling behaviors, i.e., ⟨ϵθ⟩V,t∼a−1Reos1.97±0.04 for Reos<Reos,cr and ⟨ϵθ⟩V,t∼a−1Reos1.31±0.02 for Reos>Reos,cr, where the fitted critical oscillational Reynolds number is approximately Reos,cr≈80. The different scaling of ⟨ϵθ⟩V,t is determined by the competition between the thermal boundary layer and the oscillating boundary layer. Moreover, the probability density functions (PDFs) of both dissipation rates deviate significantly from the lognormal distribution and exhibit a bimodal shape. By partitioning the contributions from the boundary layer and bulk regions, it is shown that the bulk contributes to the small and moderate dissipation rates, whereas the high dissipation rates are predominantly contributed by the boundary layer. As Reos increases, the heavy tail of the PDFs becomes more pronounced, revealing an enhanced level of small-scale intermittency. This small-scale intermittency is mainly caused by the influence of BL due to vibration. Our study provides insight into the small-scale characteristics of 2D TVC, highlighting the non-trivial scaling laws and intermittent behavior of energy dissipation rates with respect to vibration intensity.

Highly Efficient Anchor-Free Oriented Small Object Detection for Remote Sensing Images via Periodic Pseudo-Domain

With the continuous progress of remote sensing image object detection tasks in recent years, researchers in this field have gradually shifted the focus of their research from horizontal object detection to the study of object detection in arbitrary directions. It is worth noting that some properties are different from horizontal object detection during oriented object detection that researchers have yet to notice much. This article presents the design of a straightforward and efficient arbitrary-oriented detection system, leveraging the inherent properties of the orientation task, including the rotation angle and box aspect ratio. In the detection of low aspect ratio objects, the angle is of little importance to the orientation bounding box, and it is even difficult to define the angle information in extreme categories. Conversely, in the detection of objects with high aspect ratios, the angle information plays a crucial role and can have a decisive impact on the quality of the detection results. By exploiting the aspect ratio of different targets, this letter proposes a ratio-balanced angle loss that allows the model to make a better trade-off between low-aspect ratio objects and high-aspect ratio objects. The rotation angle of each oriented object, which we naturally embed into a two-dimensional Euclidean space for regression, thus avoids an overly redundant design and preserving the topological properties of the circular space. The performance of the UCAS-AOD, HRSC2016, and DLR-3K datasets show that the proposed model in this paper achieves a leading level in terms of both accuracy and speed.

Predicting the effect of pallet overhang on the box compression strength

AbstractA corrugated box's compression strength can easily be affected by how the box is used, including pallet overhang which reduces a box's effective compression strength (BCT). The specific impact of a given amount of overhang on BCT remains poorly defined. In the current study, a range of box sizes and constructions were examined in over a dozen single‐side overhang configurations and five adjacent‐side overhang scenarios to identify key factors contributing to loss of strength, producing an average reduction in BCT of up to 40%. These results indicated that common safety factors fit well with adjacent overhang scenarios but overestimate single‐side overhang scenarios. We developed a range of multiple linear and nonlinear regression models, estimating the change in a box's compression strength due to overhang compared with a no‐overhang scenario. The impact of overhang on the short and long sides, whether overhang exists on a single side or adjacent side, box size, and board type were all statistically significant. This work also indicates the need for further research refining the first‐order model and extending it to other materials, box sizes, and box aspect ratios.

Flexural behavior of wood in the transverse direction investigated using novel computer vision and machine learning approach

Abstract A deep-learning-based semantic segmentation approach (U-Net) was used to partition the anatomical features in the cross-section of hinoki (Chamaecyparis obtusa) wood during a micro three-point bending test. Using the Crocker–Grier linking algorithm, thousands of cells were successfully extracted, and several parameters (area, eccentricity, fitted ellipse aspect ratio, bounding box aspect ratio) were used to evaluate the intensity of the cells’ deformation. Thus, the 2D map of the deformation intensity distribution was constructed. By analyzing flat-sawn, quarter-sawn, and rift-sawn specimens, it was confirmed that the annual ring orientation affects the flexural behavior of wood in the transverse direction. The quarter-sawn specimens exhibited the largest modulus of elasticity (MOE) and modulus of rupture (MOR). The ray tissue aligned against the load may have contributed to the restriction of cell deformation. The rift-sawn specimens exhibited the smallest MOE and MOR, possibly owing to the loading of the specimen in the in-plane off-axial direction, which induced the shear deformation of the cell wall. For all three specimen types, the fracture had high occurrence probability in the tension part of the specimen, which exhibited large cell deformation. Therefore, the proposed method can be adapted to the prediction of wood specimen fractures. With different test wood species, this approach can be of great help in elucidating the relationship between the anatomical features and the mechanical behavior of wood to improve the effective utilization of wood resources.

Weed Detection by Faster RCNN Model: An Enhanced Anchor Box Approach

To apply weed control treatments effectively, the weeds must be accurately detected. Deep learning (DL) has been quite successful in performing the weed identification task. However, various aspects of the DL have not been explored in previous studies. This research aimed to achieve a high average precision (AP) of eight classes of weeds and a negative (non-weed) class, using the DeepWeeds dataset. In this regard, a DL-based two-step methodology has been proposed. This article is the second stage of the research, while the first stage has already been published. The former phase presented a weed detection pipeline and consisted of the evaluation of various neural networks, image resizers, and weight optimization techniques. Although a significant improvement in the mean average precision (mAP) was attained. However, the Chinee apple weed did not reach a high average precision. This result provided a solid ground for the next stage of the study. Hence, this paper presents an in-depth analysis of the Faster Region-based Convolutional Neural Network (RCNN) with ResNet-101, the best-obtained model in the past step. The architectural details of the Faster RCNN model have been thoroughly studied to investigate each class of weeds. It was empirically found that the generation of anchor boxes affects the training and testing performance of the Faster RCNN model. An enhancement to the anchor box scales and aspect ratios has been attempted by various combinations. The final results, with the addition of 64 × 64 scale size, and aspect ratio of 1:3 and 3:1, produced the best classification and localization of all classes of weeds and a negative class. An enhancement of 24.95% AP was obtained in Chinee apple weed. Furthermore, the mAP was improved by 2.58%. The robustness of the approach has been shown by the stratified k-fold cross-validation technique and testing on an external dataset.

Investigation of flow dynamics and heat transfer mechanism in turbulent Rayleigh–Bénard convection over multi-scale rough surfaces

We present a two-dimensional numerical study of turbulent Rayleigh–Bénard convection with air as the working fluid over a multi-scale randomly roughened surface in a rectangular box of aspect ratio 2 over three decades of Rayleigh number ( $10^8 \leq {\textit {Ra}} \leq 10^{11}$ ). With varied response of the roughness elements at different ${\textit {Ra}}$ , enhanced heat transfer scaling ( $\gamma =0.41$ ) is retained throughout the explored ${\textit {Ra}}$ range. The plume emission process is triggered from the taller roughness elements at a lower ${\textit {Ra}}$ , while smaller elements contribute significantly at higher ${\textit {Ra}}$ . Increased plume emission frequency compared to the smooth case is reflected from the enhanced volume fraction and thermal dissipation rate of plumes. The tip of the roughness elements exhibits the highest temperature and vertical velocity fluctuations, while washing out of the trapped fluid in the throat region holds the key to enhanced heat flux at higher ${\textit {Ra}}$ . Increased localized pockets of fluid at higher ${\textit {Ra}}$ indicate better inter-scale energy transfer, which is reflected in higher energy content at all scales in the computed temperature spectra. The decomposition of the flow field into orthogonal modes reveals that heat transfer enhancement at higher ${\textit {Ra}}$ is associated with multiple small-scale structures. Owing to better energy transfer and intense localized fluctuations, modal distribution of energy is less severe for higher ${\textit {Ra}}$ , and stable twin large-scale rolls do not favour an efficient heat transport process.

Scalable processing of granular crystals by high-frequency oscillation

High-frequency vibration proves a fundamental step towards scalable growth in colloidal crystals. Various experimentally-replicable vibration mechanisms are hereby devised and exerted on randomly generated silicoid particles to explore geometric and dynamic effects on the induced crystal at equilibrium state. The Discrete Element Method is employed, with inter-particle contact idealized by a linear mass–spring–dashpot system with Coulomb sliding friction. With the vertical (Z) direction determined by gravity, the mechanisms comprise vertical, circular, and angular oscillation, with the latter realized in the plane normal to gravity. Characterization of the generated crystal is quantified by a hybrid machine-learning algorithm developed upon direct correlation between the coordinate data, crystallinity and grain-surface parameters. Parametric analysis is carried out in terms of vibration parameters (frequency, amplitude and duration) and the square-prismatic box aspect ratio. Results manifest optimum values of box aspect ratio, vibration amplitude and frequency, with crystallinity parameters mostly saturating after 2 minutes of vibration.

Use of convolutional networks in the conceptual structural design of shear wall buildings layout

In the structural design of shear wall buildings, the initial process requires the interaction between the architecture and engineering teams to define the appropriate distribution of the walls, a stage typically carried out through a trial-and-error procedure, without any consideration of previous similar projects. In previous work, a database of 165 Chilean residential projects of reinforced shear wall concrete buildings was built, which fed a regressive neural network model to predict the wall’s engineering thickness and length values from an architectural 30-feature input vector, which accounts for geometric and topological properties, archiving remarkable results regarding the coefficient of determination (R2). However, a regressive model of this nature does not incorporate a spatial detail or contextual information of each wall’s perimeter, and also, the prediction of other parameters such as the wall translation has a poor performance. For this reason, the present research proposes a framework based on convolutional neural network (CNN) models to generate the final engineering floor plan by combining two independent floor plan predictions, considering the architectural data as input. The first plan prediction is assembled using two regressive models that predict the wall engineering values of the thickness, the length, the wall translation on both axes from the architectural plan, and the floor bounding box width and aspect ratio. The second plan prediction is assembled using a model that generates a likely image of each wall’s engineering floor plan. Both independently predicted plans are combined to lead the final engineering floor plan, which allows predicting the wall’s rectangles design parameters and propose new structural elements not present in architecture, making the methodology an excellent candidate to accelerate the building wall layout’s early conceptual design.

Flow organisation in laterally unconfined Rayleigh–Bénard turbulence

Abstract

Direct Numerical Simulation of a Low Prandtl Number Rayleigh–Bénard Convection in a Square Box

Direct numerical simulations for low Prandtl number fluid (Pr = 0.0216) are used to study the steady-state Rayleigh–Bénard convection (RB) in a two-dimensional unit aspect ratio box. The steady-state RB convection is characterized by analyzing the time-averaged temperature-field, and flow field for a wide range of Rayleigh number (2.1 × 105 ⩽ Ra ⩽ 2.1 × 108). It is seen that the time-averaged and space-averaged Nusselt number (Nuh¯) at the hot-wall monotonically increases with the increase in Rayleigh number (Ra) and the results show a power law scaling Nuh¯∝Ra0.2593. The current Nusselt number results are compared with the results available in the literature. The complex flow is analyzed by studying the frequency power spectra of the steady-state signal of the vertical velocity at the midpoint of the box for different Ra and probability density function of dimensionless temperature at various locations along the midline of the box.

Generation of turbulence through frontogenesis in sheared stratified flows

The large-scale structures in the ocean and the atmosphere are in geostrophic balance, and a conduit must be found to channel the energy to the small scales where it can be dissipated. In turbulence, this takes the form of an energy cascade, whereas a possible mechanism in a balanced flow is through the formation of fronts, a common occurrence in geophysics. We show that an iconic configuration in laboratory and numerical experiments for the study of turbulence, the so-called Taylor-Green or von Kármán swirling flow, can be suitably adapted to domains with large aspect ratios, leading to the creation of an imposed large-scale vertical shear. To this effect, we use direct numerical simulations of the Boussinesq equations without net rotation and with no small-scale modeling. Various grid spacings are used, up to 20482 × 256 spatial points. The grids are always isotropic, with box aspect ratios of either 1:4 or 1:8. We find that when shear and stratification are comparable, the imposed shear layer resulting from the forcing leads to the formation of fronts and filaments which destabilize and evolve into a turbulent flow in the bulk, with a sizable amount of dissipation and mixing, following a cycle of front creation, instability, and development of turbulence. The results depend on the vertical length scales of shear and stratification.

An object-based approach for tree species extraction from digital orthophoto maps

Abstract. Tree segmentation is an active and ongoing research area in the field of photogrammetry and remote sensing. It is more challenging due to both intra-class and inter-class similarities among various tree species. In this study, we exploited various statistical features for extraction of hazelnut trees from 1 : 5000 scaled digital orthophoto maps. Initially, the non-vegetation areas were eliminated using traditional normalized difference vegetation index (NDVI) followed by application of mean shift segmentation for transforming the pixels into meaningful homogeneous objects. In order to eliminate false positives, morphological opening and closing was employed on candidate objects. A number of heuristics were also derived to eliminate unwanted effects such as shadow and bounding box aspect ratios, before passing them into the classification stage. Finally, a knowledge based decision tree was constructed to distinguish the hazelnut trees from rest of objects which include manmade objects and other type of vegetation. We evaluated the proposed methodology on 10 sample orthophoto maps obtained from Giresun province in Turkey. The manually digitized hazelnut tree boundaries were taken as reference data for accuracy assessment. Both manually digitized and segmented tree borders were converted into binary images and the differences were calculated. According to the obtained results, the proposed methodology obtained an overall accuracy of more than 85 % for all sample images.

Scalings of heat transport and energy spectra of turbulent Rayleigh-Bénard convection in a large-aspect-ratio box

Direct Numerical Simulations of turbulent convection in a large aspect-ratio box are carried out in the range of Rayleigh number 7 × 104 ≤ Ra ≤ 2 × 106 at Prandtl number Pr=0.71. A strong correlation between the vertical velocity and temperature is observed in the turbulent regime at almost all the length scales. Frequency spectra of all the velocities and temperature show a −5/3 law for a wide band of frequencies. The variances of horizontal velocities at different points in the flow yield a single power-law. Probability density functions of velocities and temperature are close to Gaussian only at higher Rayleigh numbers. The mean and variance of temperature clearly show boundary layers, surface layers and a near-homogeneous bulk region. The boundary layer thickness decreases and bulk-homogeneity is enhanced on increasing the Rayleigh numbers. The wave number spectra of the turbulent kinetic energy exhibit Kolmogorov like (E(k)∼k−5/3) and Bolginao-Obukhov like (E(k)∼k−11/5) behaviour respectively in the central and near-wall regions of the container. An approximate balance between the production due to buoyancy and the dissipation is found in the turbulent kinetic energy budget. Taylor’s approximate equation of the production due to turbulent stretching and the dissipation of turbulent enstrophy is modified by the inclusion of buoyancy production in the enstrophy budget. The present results support the previously proposed 2/7 power-law dependence of the average Nusselt number on the Rayleigh number by yielding an exponent of 0.272, but do not necessarily support the proposed classification of “soft” and “hard” turbulence on the basis of this exponent.

Homogeneous shear turbulence – bypass concept via interplay of linear transient growth and nonlinear transverse cascade

We performed direct numerical simulations of homogeneous shear turbulence to study the mechanism of the self-sustenance of subcritical turbulence in spectrally stable (constant) shear flows. For this purpose, we analyzed the turbulence dynamics in Fourier/wavenumber/spectral space based on the simulation data for the domain aspect ratio 1 : 1 : 1. Specifically, we examined the interplay of linear transient growth of Fourier harmonics and nonlinear processes. The transient growth of harmonics is strongly anisotropic in spectral space. This, in turn, leads to anisotropy of nonlinear processes in spectral space and, as a result, the main nonlinear process appears to be not a direct/inverse, but rather a transverse/angular redistribution of harmonics in Fourier space referred to as the nonlinear transverse cascade. It is demonstrated that the turbulence is sustained by the interplay of the linear transient, or nonmodal growth and the transverse cascade. This course of events reliably exemplifies the wellknown bypass scenario of subcritical turbulence in spectrally stable shear flows. These processes mainly operate at large length scales, comparable to the box size. Consequently, the central, small wavenumber area of Fourier space (the size of which is determined below) is crucial in the self-sustenance and is labeled the vital area. Outside the vital area, the transient growth and the transverse cascade are of secondary importance - Fourier harmonics are transferred to dissipative scales by the nonlinear direct cascade. The number of harmonics actively participating in the self-sustaining process (i.e., the harmonics whose energies grow more than 10% of the maximum spectral energy at least once during evolution) is quite large - it is equal to 36 for the considered box aspect ratio - and obviously cannot be described by low-order models.

Natural convection in thermal plumes emerging from a single heat source

We report numerical simulations of confined natural convection from a single heat source, leading to the evolution of thermal plumes in two and three dimensions. Thermal plumes are driven through a single heat source mounted flush at the bottom of a slender cavity where vertical and top walls are isothermal heat sinks. Velocity and temperature fields were obtained for two Prandtl numbers, P=0.025,0.71 at three different values of the Rayleigh number, R=104,5×104,105 and for different box aspect ratios. Two kind of flow solutions were found: (i) Steady states corresponding to stable thermal plumes characterized by a well defined flow circulation inside the cavity and (ii) periodic states where both the flow and thermal fields oscillate in time. Unsteadiness of fluid and thermal flows is favored by choosing low Prandtl number fluids, working at high Rayleigh numbers inside high aspect ratio cavities. Instabilities are characterized by a periodic and propagative motion of the thermal plume in both transverse and vertical direction. It can be attributed to destabilizing shear stresses between ascending and descending fluid layers.

Diffusion in bulk liquids: finite-size effects in anisotropic systems

We investigate systematically the effect of the cell size and shape on the diffusion properties in molecular dynamics simulations. Specifically, we consider a bulk Lennard–Jones fluid in orthorhombic cells with one length differing from the other two. The components of the diffusion tensor display complex variations as a function of the two independent lengths and may even become in some cases larger than the macroscopic limit for a cubic cell. These results can be perfectly explained by a purely hydrodynamic theory, which extends results obtained previously for the isotropic case. We provide the explicit expression of the diffusion tensor, including the effect of the finite size of the diffusing particle. The simulation results follow a simple scaling as a function of box size and aspect ratio and the corresponding scaling functions are determined numerically. These findings should have implications for the practically more relevant case of confined fluids.

Period-Doubling Cascade to Chaos of the Rayleigh–Bénard Convection in a Rectangular Box

A dynamical system governing the Rayleigh–Bénard convection of an electrically conducting fluid contained in a rectangular box heated from below with application of a uniform horizontal magnetic field \(\mathcal{B}_{0}\) is constructed with the aid of the Galerkin procedures. Let the externally imposed temperatures be denoted by T1 and T2 on the box bottom and top walls respectively. The box dimensions are 2Lx in length, 2Ly in width, and 2Lz in depth with the gravitational acceleration directed downward along the z-axis. The system is characterized by a set of the non-dimensional external parameters: the Rayleigh number Ra proportional to the temperature difference T1 − T2; the Chandrasekhar number Q proportional to \(\mathcal{B}_{0}^{2}\); the Prandtl number Pr of the fluid; and the box aspect ratios Γx = Lx/Lz and Γy = Ly/Lz. Computer simulations of time evolution of the convection give the results that for a fluid of Pr = 0.025 in a box of Γx = 4 and Γy = 1 at a suitably chosen value of Q = 22 with increase of Ra the fluid motion undergoes a successive period-doubling bifurcations up to 1/8 the fundamental frequency eventually leading to non-periodicity.

Scaling and direct stability analyses of natural convection induced by absorption of solar radiation in a parallelepiped cavity

The present study considers the early stage of the flow development in a water-filled, three-dimensional parallelepiped cavity subject to solar radiation. The thermal structure consists of an upper stable stratification due to internal heating, provided by the direct absorption of radiation, and a bottom potentially unstable boundary layer due to the absorption and re-emission of the residual radiation by the bottom boundary. The growth of the competing thermal layers and the stability properties of the bottom thermal boundary layer are investigated by means of a scaling analysis and a direct stability analysis based on three-dimensional numerical simulation. The scaling relations are benchmarked against the numerical simulation, and the effects of the normalised water depth, the Rayleigh number and the box aspect ratio on the thermal instability are studied.

Influences of the depth-dependence of thermal conductivity and expansivity on thermal convection with temperature-dependent viscosity

In this paper we carried out numerical experiments of a time-dependent thermal convection in a two-dimensional Cartesian box of aspect ratio (width/height) of 6, in order to study the influences on the convecting flow patterns of the spatial variations in physical properties (viscosity η, thermal conductivity k and expansivity α). A series of calculations by systematically varying the magnitude of spatial variations in these properties showed that the strongly temperature-dependent η induces the change in flow pattern into the “stagnant lid” (ST) regime, regardless of the increase in k and/or decrease in α with depth, where the convection occurs only beneath a stagnant lid of cold fluid at the top surface. In particular, we found that the increase in thermal conductivity k with depth significantly affects the convecting flow patterns in the presence of strong temperature-dependence in η. Compared with those with uniform k, the patterns of ST convection with non-uniform k are characterized by (i) thinner top thermal boundary layers or lids and (ii) larger horizontal length scales of convection cells beneath the stagnant lid. In addition, the variation in k with depth decreases the threshold values of the temperature-dependence in η above which the ST-mode of convection takes place, which may also help stabilize the convection cells of large horizontal length scales beneath stagnant lids. Our results may highlight the potential importance of the increase in thermal conductivity with depth (or pressure) in controlling the planforms of thermal convection in the mantle of terrestrial planets.

Long-term nonlinear behaviour of the magnetorotational instability in a localized model of an accretion disc

For more than a decade, the so-called shearing box model has been used to study the fundamental local dynamics of accretion discs. This approach has proved to be very useful because it allows high resolution and long term studies to be carried out, studies that would not be possible for a global disc. Localised disc studies have largely focused on examining the rate of enhanced transport of angular momentum, essentially a sum of the Reynolds and Maxwell stresses. The dominant radial-azimuthal component of this stress tensor is, in the classic Shakura-Sunayaev model, expressed as a constant alpha times the pressure. Previous studies have estimated alpha based on a modest number of orbital times. Here we use much longer baselines, and perform a cumulative average for alpha. Great care must be exercised when trying to extract numerical alpha values from simulations: dissipation scales, computational box aspect ratio, and even numerical algorithms all affect the result. This study suggests that estimating alpha becomes more, not less, difficult as computational power increases.