Vertical Flow Component Research Articles

S3DCN-OLSR: A shallow 3D CNN method for online learning state recognition

The repeated recurrence of COVID-19 has significantly disrupted learning for students in face-to-face instructional settings. While moving from offline to online instruction has proven to be one of the best solutions, classroom management and capturing students' learning states have emerged as important challenges with the increasing popularity of online instruction. To address these challenges, in this paper we propose an online learning status recognition method based on shallow 3D convolution (S3DC-OLSR) for online students, to identify students' online learning states by analysing their micro-expressions. Specifically, we first use the data augmentation method proposed in this paper to decompose the students' online video file into three features: horizontal component of optical flow, vertical component of optical flow and optical amplitude. Next, the students' online learning status is recognised by feeding the processed data into a shallow 3D convolution neural network. To test the performance of our method, we conduct extensive experiments on the CASME II and SMIC datasets, and the results indicate that our method outperforms the other state-of-the-art methods considered in terms of recognition accuracy, UF1 and UAR, which demonstrates the superiority of our method in identifying students’ online learning states.

A Two‐Dimensional Closed‐Form Analytical Solution for Heat Transport With Nonvertical Flow in Riparian Zones

AbstractHeat as a tracer has received increasing attention in recent decades and has become a popular and widely used tool for quantifying streambed fluxes, among many other applications. Nonvertical flow component is an important parameter for heat transport in riparian zones. However, previous investigations using heat as a tracer in the quantification of streambed fluxes frequently overlooked or simplified nonvertical flow component with erroneous presumptions. In this study, a novelty two‐dimensional closed‐form analytical solution for heat transport with vertical and nonvertical flow components and arbitrary initial and boundary conditions under losing conditions is presented for the first time using Green's function method. The new model is tested by finite‐element solution and stochastic modeling. The capabilities of the new analytical solution are demonstrated using field data. Results show that the nonvertical flow component has a significant impact on heat transport in streambed. The temperature difference between the streambed and the surface water increases as the nonvertical flow component increases, and the amplitude of the temperature curves decreases with increasing horizontal distance from the center of stream at the same depth due to the influences of nonvertical flow. Stochastic modeling shows that an increasing nonuniform flow can result in overestimation of vertical and nonvertical flow components near the bank of stream, and the new analytical model works well in a heterogeneous streambed when the variance of natural logarithm of the streambed hydraulic conductivity is less than 0.10 (). The vertical and nonvertical flow components are more sensitive to temperature‐time data than the other parameters. The new analytical model is a significant advancement of the previous one‐dimensional analytical models which only considered vertical flow.

Schmidt method to study the disturbance of steady-state heat flows by an arbitrary oriented crack in bonded functionally graded strips

This article aims to investigate the impact of a partially insulated crack on vertical and horizontal components of heat flow. The crack is arbitrary oriented and its center is located at the interface of bonded finitely thick functionally graded strips. Employing the integral transformation, the thermoelastic equations are reduced to a set of singular integral equations. The unknown variables of these equations are the jump in temperature and displacements across the crack surfaces. Representing the unknowns as a series of Jacobi polynomials, the solutions of the singular integral equations are obtained by employing the Schmidt method which further helps to determine the analytical forms of crack tip heat flux intensity factors (HFIFs) and thermal stress intensity factors (TSIFs). The results of the present study are also validated for a particular case.To quantify the strength of heat flow components, the crack tip HFIFs and TSIFs as a function of crack angle, crack insulation parameter, strips’ width, and non-homogeneity parameters’ ratio are demonstrated pictorially.

Upper Limit of Light-Levitated Droplet Motion.

The light-enabled droplet levitation shows promising potential in applications in biotechnology, clinical medicine, and nanomaterials. In particular, light-levitated droplets have good followability with a moving laser beam, resulting in flexibility in manipulating their motion. However, it is still unclear whether there exists an upper limit to the light-levitated droplet motion with a moving laser beam. Therefore, the motion of light-levitated droplets above the free interface is studied to determine the upper limit of motions of the droplets with a moving laser beam. We demonstrate that an inefficient interface temperature response because of a very high moving speed of the laser beam and the resultant small upward vertical component of vapor flow are responsible for the existence of an upper-limit velocity. Above the upper limit, the light-levitated droplets are unable to stably move with the laser beam and finally disappear. By contrast, the droplets can stably move with the laser beam in a wide range at or below this upper limit. In addition, an almost linear relationship between the upper-limit velocity of the light-levitated droplets and the input laser power is presented. The findings of the present study are informative for the implementation of this light levitation technology.

The effect of boundary conditions on the single-well push-pull test in a partially penetrated well

Boundary conditions of well screen are critical formodeling reactive transport around a partially penetrated well in a single-well push–pull (SWPP) test. Previous SWPP test models often assumed that water flux was uniformly distributed along the well screen of the partially penetrated well, named the uniformly distributed flux boundary condition (UFBC). To investigate the influence of such an assumption on reactive transport in the SWPP test, we develop a new finite-difference numerical model under a flow field with a finite hydraulic diffusivity in this study, considering the wellbore storage. The uniformly distributed head boundary condition (UHBC) is employed for both groundwater flow and reactive transport across the well screen. The dispersion coefficients are in a tensor form, depending on both horizontal and vertical flow components. The results show that the UFBC model may cause non-negligible errors when modeling reactive transport in the SWPP test with a partially penetrated well. Such errors are obvious at the early stage of the extraction phase, but become negligible at late stage of the extraction phase. The UHBC solution for reactive transport is smaller than the UFBC solution at the middle of the well screen, but greater than the UFBC solution at the top or bottom of the well screen. Using the SWPP test to investigate the reactive transport in subsurface, in addition to the breakthrough curves (BTCs), the information of the flow field (like well geometry information and the observed time-drawdown data) is necessary and could be used to calibrate the model. Field application demonstrates that the UHBC model performs better than the UFBC model when interpreting BTCs of the SWPP test in the extraction phase. The findings of this study can help for better understanding of reactive transport in the SWPP test.

Micro-Expression Recognition Method Based on Spatial Attention Mechanism and Optical Flow Features

<p indent="0mm">In order to solve the locality problem of micro-expression movement, an automatic micro-expression recognition method based on spatial attention mechanism of deep learning and optical flow features of micro-expressions is proposed. First, the frame difference method is used to identify apex frame for some micro-expression samples lack of apex frame labels. Then, TV<italic>-L</italic>₁ optical flow method is used to extract the horizontal and vertical components map of optical flow between the onset frame and the apex frame of micro-expression, and the corresponding optical flow strain pattern map is derived according to the horizontal and vertical components of optical flow. The three optical flow maps are connected in the way of channel superposition to form an optical flow characteristic map of micro-expression. Finally, a kind of spatial attention unit with learnable parameters is designed in the convolutional neural network built by the Inception module, which makes the model pay more attention to the regions with micro-expression motion in the feature extraction process. In the spatial attention unit, two convolution kernels of 3×3 and 7×7 are used for spatial attention inference, so that the model can comprehensively consider the attention inference results of different scale convolution kernels. Experiments on the MEGC2019 comprehensive micro-expression datasets show that the accuracy of the method is 0.788, which is better than the existing automatic micro-expression recognition method.

Single trunk multi-scale network for micro-expression recognition

Micro-expressions are the external manifestations of human psychological activities. Therefore, micro-expression recognition has important research and application value in many fields such as public services, criminal investigations, and clinical diagnosis. However, the particular characteristics (e.g., short duration and subtle changes) of micro-expressions bring great challenges to micro-expression recognition. In this paper, we explore the differences in the direction of facial muscle movement when people make different expressions to recognize micro-expressions. We first use optical flow to capture the subtle changes in the facial movement when a micro-expression occurs. Next, we extract facial movement information to an aniso-weighted optical flow image based on anisotropically weighting the horizontal and vertical components of the optical flow. Finally, we feed the aniso-weighted optical flow image into the proposed Single Trunk Multi-scale Network for micro-expression recognition. In particular, the designed multi-scale feature catcher in the network can capture features of micro-expressions with different intensities. We conduct extensive experiments on four spontaneous micro-expression datasets, and the experiment results show that our proposed method is competitive and effective.

Simplified equations for lower crustal flow driven by lateral pressure gradients

SUMMARY Evidence from seismology, geology and geodynamic studies suggests that regional-scale lower crustal flow occurs in many tectonic settings. Pressure gradients caused by mantle processes and crustal density heterogeneity can provide driving force for lower crustal flow. Numerically modelling such flow can be computationally expensive. However, by exploiting symmetry in the physical system, it is possible to represent the vertical component of flow in terms of its lateral components, thereby reducing the problem’s spatial dimension by one. Here, we present a mathematical formulation for flow in a viscous channel below an elastic upper plate, which is optimized for solution by common numerical methods. Our formulation drastically reduces the computational load required to simulate lower crustal flow over large areas and long timescales. We apply this model to two example problems, with and without an elastic upper plate, identifying combinations of parameters that are capable of generating measurable geologic uplift.

Three-dimensional hydrostratigraphy and groundwater flow models in complex Quaternary deposits and weathered/fractured bedrock: evaluating increasing model complexity

Greater complexity in three-dimensional (3D) model structures yields more plausible groundwater recharge/discharge patterns, especially in groundwater/surface-water interactions. The construction of a 3D hydrostratigraphic model prior to flow modelling is beneficial when the hydraulic conductivity of geological units varies considerably. A workflow for 3D hydrostratigraphic modelling with Leapfrog Geo and flow modelling with MODFLOW-NWT was developed. It was used to evaluate how the modelling results for groundwater flow and recharge/discharge patterns differ when using simple or more complex hydrostratigraphic models. The workflow was applied to a study site consisting of complex Quaternary sediments underlain by fractured and weathered crystalline bedrock. Increasing the hydrostratigraphic detail appeared to improve the fit between the observed and simulated water table, and created more plausible groundwater flow patterns. Interlayered zones of low and high conductivity disperse the recharge/discharge patterns, increasing the vertical flow component. Groundwater flow was predominantly horizontal in models in which Quaternary sediments and bedrock were simplified as one layer per unit. It appears to be important to define the interlayered low-conductivity units, which can limit groundwater infiltration and also affect groundwater discharge patterns. Explicit modelling with Leapfrog Geo was found to be effective but time-consuming in the generation of scattered and thin-layered strata.

Wind Tunnel Study of a “Floating” Wind Turbine’s Wake in an Atmospheric Boundary Layer with Imposed Characteristic Surge Motion

Floating offshore wind turbines (FOWTs) are a potential source for increased offshore energy production. As the technology is still in a pre-industrial state several questions remain to be addressed where little field data is available. This study uses physical modelling at a reduced scale to investigate the unsteady behaviour and the development of the wake in a simplified FOWT model. A porous disc model is placed in an atmospheric wind tunnel and subjected to a range of different motions. The effects of induced sinusoidal surge motion on the characteristics of the model’s wake at a fixed downstream distance of 4.6D are studied. First results show unchanged mean velocity values but modified turbulence intensity and turbulent kinetic energy profiles. Spectra taken in the wake show shifts to higher frequencies in the longitudinal flow component while the opposite is true for the vertical flow component. Further research into spatial characteristics of the wake are necessary to confirm these results.

Three-dimensional response to a partially penetration well pumping in a general anisotropic three-layer aquifer system

Three-dimensional response to well pumping in a multilayer aquifer system with rigorous treatment of interfacial flow attracts great attention among the subsurface hydrologists because of its theoretical importance and practical merits. In this study, a general theory of three-dimensional groundwater flow to a well of infinitesimal radius in an anisotropic three-layer aquifer system is developed with a constant-rate pumping in the lower layer, and associated semi-analytical solutions of drawdowns in individual layers are obtained. This theory discards many previous assumptions for studying a multilayer aquifer system such as simplified flow configuration in the aquitard, treatment of cross-formation flow as a volumetric averaged term adding to the governing equations of aquifer flows, and sometimes negligence of vertical flow components in the aquifers. Additionally, this theory considers three types of commonly used top boundary conditions: A water table with delayed gravity drainage boundary condition (Case 1), a constant-head top boundary condition (Case 2), and a no-flow top boundary condition (Case 3). The dimensionless drawdown solutions in Laplace domain are obtained and inverted numerically to calculate the time-domain solutions. The solution encompasses existing solutions for a two-layer or a single-layer (e.g. unconfined or confined) aquifer system as subsets. The developed solutions are tested extensively with a finite-element numerical solution using COMSOL and other existing solutions, and a sensitivity analysis is performed to prioritize the influences of different parameter groups. The results of this study can be used to predict the pumping induced drawdown at any position with an observation well, and to estimate the hydraulic parameters of the lower pumped layer and upper layer, as well as the middle layer. The developed theory provides a basis for other potential applications such as geotechnical engineering and groundwater management.

A guideline for appropriate application of vertically-integrated modeling approaches for geologic carbon storage modeling

Mathematical modeling is an essential tool for answering questions related to geologic carbon storage (GCS). The choice of modeling approach depends on the type of questions being asked. In this paper we discuss a series of approaches with a hierarchical complexity including vertically-integrated single-phase flow approaches, vertically-integrated multi-phase flow approaches (with and without vertical equilibrium assumption), three-dimensional multi-phase flow approaches, and fully-coupled multi-phase flow approaches that couple flow with geochemistry and/or geomechanics. Three spatial scales are used to categorize the questions to be addressed by modeling: regional scale (encompasses CO2 plume extent and majority of area of pressure impact of one or more injection operations), site scale (includes the CO2 plume extent and some of the area impacted by the pressure increase of a single injection site), and well scale (the immediate vicinity of an injection well). A set of guidelines is developed to help modelers choose the most appropriate modeling approach, and show when simpler modeling approaches may be the better choice. Vertically-integrated single-phase flow models are the most appropriate choice at both the site and regional scales, if the pressure impact outside of the CO2 plume is of interest. Vertically-integrated multi-phase flow models should be chosen at the regional scale, if the locations of CO2 plumes are of interest, and at the site scale if vertical segregation of CO2 and brine is fast or vertical heterogeneity in properties can be presented by distinct, continuous layers. Three-dimensional multi-phase flow models are the appropriate choice at the well and site scales for cases with significant vertical flow components of CO2 and brine. Fully-coupled multi-phase flow models should only be chosen if pore-space alteration through geochemistry or geomechanics feeds back to fluid flow.

Magnitude-Orientation Stream network and depth information applied to activity recognition

The temporal component of videos provides an important clue for activity recognition, as a number of activities can be reliably recognized based on the motion information. In view of that, this work proposes a novel temporal stream for two-stream convolutional networks based on images computed from the optical flow magnitude and orientation, named Magnitude-Orientation Stream (MOS), to learn the motion in a better and richer manner. Our method applies simple non-linear transformations on the vertical and horizontal components of the optical flow to generate input images for the temporal stream. Moreover, we also employ depth information to use as a weighting scheme on the magnitude information to compensate the distance of the subjects performing the activity to the camera. Experimental results, carried on two well-known datasets (UCF101 and NTU), demonstrate that using our proposed temporal stream as input to existing neural network architectures can improve their performance for activity recognition. Results demonstrate that our temporal stream provides complementary information able to improve the classical two-stream methods, indicating the suitability of our approach to be used as a temporal video representation.

Combined helioseismic inversions for 3D vector flows and sound-speed perturbations

Context. Time–distance helioseismology is the method of the study of the propagation of waves through the solar interior via the travel times of those waves. The travel times of wave packets contain information about the conditions in the interior integrated along the propagation path of the wave. The travel times are sensitive to perturbations of a variety of quantities. The usual task is to invert for the vector of plasma flows or the sound–speed perturbations separately. The separate inversions may be polluted by systematic bias, for instance, originating in the leakage of vector flows into the sound–speed perturbations and vice versa (called a cross-talk). Information about the cross-talk is necessary for a proper interpretation of results. Aims. We introduce an improved methodology of the time-distance helioseismology which allows us to invert for a full 3D vector of plasma flows and the sound–speed perturbations at once. Using this methodology one can also derive the mean value of the vertical component of plasma flows and the cross-talk between the plasma flows and the sound–speed perturbations. Methods. We used the Subtractive Optimally Localised Averaging method with a minimisation of the cross-talk as a tool for inverse modelling. In the forward model, we use Born approximation travel-time sensitivity kernels with the Model S as a background. The methodology was validated using forward-modelled travel times with both mean and difference point-to-annulus averaging geometries applied to a snapshot of fully self-consistent simulation of the convection. Results. We tested the methodology on synthetic data. We demonstrate that we are able to recover flows and sound–speed perturbations in the near-surface layers. We have taken the advantage of the sensitivity of our methodology to entire vertical velocity, and not only to its variations as in other available methodologies. The cross-talk from both the vertical flow component and the sound–speed perturbation has only a negligible effect for inversions for the horizontal flow components. Furthermore, this cross-talk can be minimised if needed. The inversions for the vertical component of the vector flows or for the sound–speed perturbations are affected by the cross-talk from the horizontal components, which needs to be minimised in order to provide valid results. It seems that there is a nearly constant cross-talk between the vertical component of the vector flows and the sound–speed perturbations.

Performance analysis of the deflector integrated cross axis wind turbine

Wind turbines in the context of integration in an urban environment are considered as one of the most promising technologies for the efficient diffusion of renewable energy sources. However, situations such as low wind velocity, highly turbulent and skewed wind flow conditions are commonly encountered in this environment. This misaligned flow could affect the horizontal axis wind turbine performance. The vertical axis wind turbine (VAWT) is more suitable to operate under these unfavorable conditions. Nevertheless, the conventional VAWT has low efficiency and poor self-starting characteristics. Therefore, a novel design of deflector integrated cross axis wind turbine (CAWT) is proposed. Tests were carried out to harness wind energy from both the horizontal and vertical components of the skewed flow induced by the deflectors. A semi-empirical approach was also developed to analyse the performance of the CAWT. The aerodynamic behaviour of the CAWT was determined by combining the blade element momentum (BEM) model and double multiple streamtube theory (DMST). The approach showed notable agreement between the experimental and theoretical results (9.8%–14.2% difference). In conclusion, the findings from the experiment and the model showed that the concept has very good potential in the renewable energy industry.

Spatial and temporal distributions of turbulence under bichromatic breaking waves

The present study aims to extend insights of surf zone turbulence dynamics to wave groups. In a large-scale wave flume, bichromatic wave groups were produced with 31.5 s group period, 4.2 s mean wave period, and a 0.58 m maximum wave height near the paddle. This condition resulted in plunging-type wave breaking over a fixed, gravel-bed, barred profile. Optic, acoustic and electromagnetic instruments were used to measure the flow and the spatial and temporal distributions of turbulent kinetic energy (TKE). The measurements showed that turbulence in the shoaling region is primarily bed-generated and decays almost fully within one wave cycle, leading to TKE variations at the short wave frequency. The wave breaking-generated turbulence, in contrast, decays over multiple wave cycles, leading to a gradual increase and decay of TKE during a wave group cycle. In the wave breaking region, TKE dynamics are driven by the production and subsequent downward transport of turbulence under the successive breaking waves in the group. Consequently, the maximum near-bed TKE in the breaking region can lag the highest breaking wave by up to 2.5 wave cycles. The net cross-shore transport of TKE is in the shoaling region primarily driven by short-wave velocities and is shoreward-directed; in the wave breaking region, the TKE transport is seaward-directed by the undertow and the long-wave velocities. Downward transport of TKE is driven by the vertical component of the time-averaged flow. The cross-shore and vertical diffusive transport rates are small relative to the advective transport rates.

Blind test comparison on the wake behind a yawed wind turbine

Abstract. This article summarizes the results of the “Blind test 5” workshop, which was held in Visby, Sweden, in May 2017. This study compares the numerical predictions of the wake flow behind a model wind turbine operated in yaw to experimental wind tunnel results. Prior to the workshop, research groups were invited to predict the turbine performance and wake flow properties using computational fluid dynamics (CFD) methods. For this purpose, the power, thrust, and yaw moments for a 30∘ yawed model turbine, as well as the wake's mean and turbulent streamwise and vertical flow components, were measured in the wind tunnel at the Norwegian University of Science and Technology (NTNU). In order to increase the complexity, a non-yawed downstream turbine was added in a second test case, while a third test case challenged the modelers with a new rotor and turbine geometry. Four participants submitted predictions using different flow solvers, three of which were based on large eddy simulations (LES) while another one used an improved delayed detached eddy simulation (IDDES) model. The performance of a single yawed turbine was fairly well predicted by all simulations, both in the first and third test cases. The scatter in the downstream turbine performance predictions in the second test case, however, was found to be significantly larger. The complex asymmetric shape of the mean streamwise and vertical velocities was generally well predicted by all the simulations for all test cases. The largest improvement with respect to previous blind tests is the good prediction of the levels of TKE in the wake, even for the complex case of yaw misalignment. These very promising results confirm the mature development stage of LES/DES simulations for wind turbine wake modeling, while competitive advantages might be obtained by faster computational methods.

Validation of the real-time-response ProCap measurement system for full field wake scans behind a yawed model-scale wind turbine

For an accurate prediction of the complex flow conditions in wind farms, model scale wake flow measurements represent important references in well-defined boundary conditions. State-of-the art flow measurement techniques are often time-expensive and require elaborate post-processing to assess the data quality. In this wind tunnel study we demonstrate the advantages of the real-time-response system Probe Capture (ProCap) for measurements of a complex three-dimensional wind turbine wake flow. The complex wake flow behind a yawed model wind turbine is measured with both a Laser-Doppler Anemometer (LDA) and the ProCap system. Both the streamwise and vertical flow component show an accurate agreement with the LDA reference experiment for various measured downstream distances and turbine yaw angles. Areas of strong rotation in the wake flow are accurately resolved by the ProCap measurement system confirming its applicability for wind turbine wake measurement. The system appears to greatly facilitate and speed up larger wake surveys by a factor of 30 and thus has the potential to enhance the understanding of the complex flow topology in wind turbine wakes.

Learning deep facial expression features from image and optical flow sequences using 3D CNN

Facial expression is highly correlated with the facial motion. According to whether the temporal information of facial motion is used or not, the facial expression features can be classified as static and dynamic features. The former, which mainly includes the geometric features and appearance features, can be extracted by convolution or other learning filters; the latter, which are aimed to model the dynamic properties of facial motion, can be calculated through optical flow or other methods, respectively. When 3D convolutional neural networks (CNNs) are introduced, the extraction of two different types of features mentioned above becomes easy. In this paper, one 3D CNN architecture is presented to learn the static and dynamic features from facial image sequences and extract high-level dynamic features from optical flow sequences. Two types of dense optical flow, which contain the tracking information of facial muscle movement, are calculated according to different image pair construction methods. One is the common optical flow, and the other is an enhanced optical flow which is called accumulative optical flow. Four components of each type of optical flow are used in experiments. Three databases, two acted databases and one nearly realistic database, are selected to conduct the experiments. The experiments on the two acted databases achieve state-of-the-art accuracy, and indicate that the vertical component of optical flow has an advantage over other components in recognizing facial expression. The experimental results on the three selected databases show that more discriminative features can be learned from image sequences than from optical flow or accumulative optical flow sequences, and the accumulative optical flow contains more motion information than optical flow if the frame distance of the image pairs used to calculate them is not too large.

Some statistical properties of three-dimensional zonostrophic turbulence

We conduct in-depth analysis of statistical flow properties from direct numerical simulations that reproduce gas giants macroturbulence, namely large-scale zonal winds. Our numerical model has been specifically designed to simulate a recent laboratory device that reports zonal jets in the configuration of deep turbulent planetary layers. In this framework, the so-called zonostrophic regime is achieved when large topographical variations of the fluid layer combine with rapid rotation in a well developed three-dimensional (3D) turbulent flow. At steady state, strongly energetic, zonally dominated, large-scale axisymmetric structures emerge scaling with Rhines’ theoretical scale. This model differs from the shallow-layer scenario where the flow is confined to a quasi-two-dimensional (2D) fluid shell and the anisotropic -effect arises from latitudinal variation of the Coriolis force. Thus, we aim to reveal, in the specific framework of the deep-layer scenario, signatures of the zonostrophic regime and of a -topography in statistical flow properties. To do so, we run two large-scale 3D direct numerical simulations in a cylindrical geometry of a highly turbulent and rapidly rotating flow. These two simulations use similar set of parameters but with and without topographical -effect. We propose a comparative phenomenological description of the temporal and spatial statistics of the three components of the velocity field. Interestingly, we report that peculiar correlations occur between the vertical and radial flow components when a -topography is imposed and show a feature possibly due to a zonostrophic dynamics in 3D frequency spectra. These results suggest the development of new tools to remotely investigate gas giants zonal winds by extracting statistical flow properties from direct observations. Ultimately our analysis may support the relevance of the deep models in the study of prevalent features of planetary dynamics.