Explore the vast range of titles available.

\"Numerical simulation models are used in all engineering disciplines for modeling physical phenomena to learn how the phenomena work, and to identify problems and optimize behavior. Smart proxy models provide an opportunity to replicate numerical simulations with very high accuracy and can be run on a laptop within a few minutes, thereby simplifying the use of complex numerical simulations which can otherwise take tens of hours. This book focuses on smart proxy modeling and provides readers with all the essential details on how to develop smart proxy models using artificial intelligence and machine learning, as well as how it may be used in real-world cases. Covers replication of highly accurate numerical simulations using artificial intelligence and machine learning. Details application in reservoir simulation and modeling, and computational fluid dynamics. Includes real case studies based on commercially available simulators. Smart Proxy Modeling is ideal for petroleum, chemical, environmental, and mechanical engineers, as well as statisticians and others working with applications of data-driven analytics\"-- Provided by publisher.

We present a proof-of-concept for the adaptive mesh refinement method applied to atmospheric boundary-layer simulations. Such a method may form an attractive alternative to static grids for studies on atmospheric flows that have a high degree of scale separation in space and/or time. Examples include the diurnal cycle and a convective boundary layer capped by a strong inversion. For such cases, large-eddy simulations using regular grids often have to rely on a subgrid-scale closure for the most challenging regions in the spatial and/or temporal domain. Here we analyze a flow configuration that describes the growth and subsequent decay of a convective boundary layer using direct numerical simulation (DNS). We validate the obtained results and benchmark the performance of the adaptive solver against two runs using fixed regular grids. It appears that the adaptive-mesh algorithm is able to coarsen and refine the grid dynamically whilst maintaining an accurate solution. In particular, during the initial growth of the convective boundary layer a high resolution is required compared to the subsequent stage of decaying turbulence. More specifically, the number of grid cells varies by two orders of magnitude over the course of the simulation. For this specific DNS case, the adaptive solver was not yet more efficient than the more traditional solver that is dedicated to these types of flows. However, the overall analysis shows that the method has a clear potential for numerical investigations of the most challenging atmospheric cases.

Good representation of turbulence in urban canopy models is necessary for accurate prediction of momentum and scalar distribution in and above urban canopies. To develop and improve turbulence closure schemes for one-dimensional multi-layer urban canopy models, turbulence characteristics are investigated here by analyzing existing large-eddy simulation and direct numerical simulation data. A range of geometries and flow regimes are analyzed that span packing densities of 0.0625 to 0.44, different building array configurations (cubes and cuboids, aligned and staggered arrays, and variable building height), and different incident wind directions (0∘ and 45∘ with regards to the building face). Momentum mixing-length profiles share similar characteristics across the range of geometries, making a first-order momentum mixing-length turbulence closure a promising approach. In vegetation canopies turbulence is dominated by mixing-layer eddies of a scale determined by the canopy-top shear length scale. No relationship was found between the depth-averaged momentum mixing length within the canopy and the canopy-top shear length scale in the present study. By careful specification of the intrinsic averaging operator in the canopy, an often-overlooked term that accounts for changes in plan area density with height is included in a first-order momentum mixing-length turbulence closure model. For an array of variable-height buildings, its omission leads to velocity overestimation of up to 17%. Additionally, we observe that the von Kármán coefficient varies between 0.20 and 0.51 across simulations, which is the first time such a range of values has been documented. When driving flow is oblique to the building faces, the ratio of dispersive to turbulent momentum flux is larger than unity in the lower half of the canopy, and wake production becomes significant compared to shear production of turbulent momentum flux. It is probable that dispersive momentum fluxes are more significant than previously thought in real urban settings, where the wind direction is almost always oblique.

The paper reviews recent advances in our understanding about the dynamics of thermally driven winds over sloping terrain. Major progress from recent experiments, both in the field and in the laboratory, are outlined. Achievements from numerical modelling efforts, including both parameterized turbulence and large eddy simulation approaches, up to direct numerical simulations, are also reviewed. Finally, theoretical insights on the nature of turbulence in such winds are analyzed along with applications which benefit from progress in understanding of these flows. Open questions to be faced for further investigations are finally highlighted.

The interior oceans of several icy moons are considered as affected by rotation. Observations suggest a larger heat transport around the poles than at the equator. Rotating Rayleigh‐Bénard convection (RRBC) in planar configuration can show an enhanced heat transport compared to the non‐rotating case within this “rotation‐affected” regime. We investigate the potential for such a (polar) heat transport enhancement in these subglacial oceans by direct numerical simulations of RRBC in spherical geometry for Ra = 106 and 0.7 ≤ Pr ≤ 4.38. We find an enhancement up to 28% in the “polar tangent cylinder,” which is globally compensated by a reduced heat transport at low latitudes. As a result, the polar heat transport can exceed the equatorial by up to 50%. The enhancement is mostly insensitive to different radial gravity profiles, but decreases for thinner shells. In general, polar heat transport and its enhancement in spherical RRBC follow the same principles as in planar RRBC. Plain Language Summary The icy moons of Jupiter and Saturn like for example, Europa, Titan, or Enceladus are believed to have a water ocean beneath their ice crust. Several of them show phenomena in their polar regions like active geysers or a thinner crust than at the equator, all of which might be related to a larger heat transport around the poles from the underlying ocean. We simulate the flow dynamics and currents in these subglacial ocean by high‐fidelity simulations, though still at less extreme parameters than in reality, to study the heat transport and provide a possible explanation of such a “polar heat transport enhancement.” We find that the heat transport around the poles can be up to 50% larger than around the equator, and that the believed properties of the icy moons and their oceans would allow polar heat transport enhancement. Therefore, our results may help to improve the understanding of ocean currents and latitudinal variations in the oceanic heat transport and crustal thickness on icy moons. Key Points The polar heat transport in spherical rotating Rayleigh‐Bénard convection experiences an enhancement by rotation The influence of rotation differs at low latitudes: the heat flux is reduced and compensates the polar enhancement on the global average In combination, this strengthens the latitudinal variation between polar and equatorial heat flux for Prandtl numbers larger than unity

Cloud microphysical processes occur at the smallest end of scales among cloud-related processes and thus must be parameterized not only in large-scale global circulation models (GCMs) but also in various higher-resolution limited-area models such as cloud-resolving models (CRMs) and large-eddy simulation (LES) models. Instead of giving a comprehensive review of existing microphysical parameterizations that have been developed over the years, this study concentrates purposely on several topics that we believe are understudied but hold great potential for further advancing bulk microphysics parameterizations: multi-moment bulk microphysics parameterizations and the role of the spectral shape of hydrometeor size distributions; discrete vs “continuous” representation of hydrometeor types; turbulence-microphysics interactions including turbulent entrainment-mixing processes and stochastic condensation; theoretical foundations for the mathematical expressions used to describe hydrometeor size distributions and hydrometeor morphology; and approaches for developing bulk microphysics parameterizations. Also presented are the spectral bin scheme and particle-based scheme (especially, super-droplet method) for representing explicit microphysics. Their advantages and disadvantages are elucidated for constructing cloud models with detailed microphysics that are essential to developing processes understanding and bulk microphysics parameterizations. Particle-resolved direct numerical simulation (DNS) models are described as an emerging technique to investigate turbulence-microphysics interactions at the most fundamental level by tracking individual particles and resolving the smallest turbulent eddies in turbulent clouds. Outstanding challenges and future research directions are explored as well.

The non‐Darcy flow behavior in unfilled and fully filled rough fractures has been investigated thoroughly for decades. Natural fractures usually be partially filled with porous media due to long‐term internal and external dynamic disturbances and water flow erosion. However, how to evaluate the nonlinear seepage characteristics of the partially filled rough fracture (PFRF) is never scrutinized. In the present study, 2D direct numerical simulations are conducted to investigate the non‐Darcy behavior for flow through 6 PFRF models under different filling conditions (i.e., filling degree, roughness, and hydraulic conductivity capability of the filling medium), which included coupling between the free and seepage flow in the unfilled and filled region, respectively. The results show that the critical non‐Darcy factor E = 0.02 can be used to judge the onset of the non‐Darcy regime in PFRF, which is also the critical point of eddy region (ER) volume mutation. The higher the filling degree and the weaker the hydraulic conductivity of the filled medium, the more significant the non‐Darcy behavior. Furthermore, the variation of ER volume can be divided into three stages with the increasing inertia effect, that is, stable stage (E < 0.02), slow nonlinear growth stage (0.02 ≤ E ≤ 0.1), and rapid linear growth stage (E > 0.1). In addition, a new model for quantitatively evaluating the non‐Darcy behavior of PFRF is developed via gene expression programming based on 735 simulation data set, which is further employed to establish the relationship between the friction factor. Plain Language Summary The nonlinear flow behavior in partially filled rough fractures (PFRF) was investigated by using 2D simulations. The simulation results revealed a critical non‐Darcy factor of 0.02 to determine the onset of non‐Darcy behavior in PFRF. Higher filling degrees and weaker hydraulic conductivity of filling region intensified the non‐Darcy effect. The volume of eddy regions exhibited three stages of variation with increasing inertia effect (i.e., stable stage, slow nonlinear growth stage, and rapid linear growth stage). We developed a new model based on gene expression programming to quantitatively evaluate PFRF's non‐Darcy behavior. This study enhances the understanding of fluid flow in partially filled rough fractures, benefiting for advancing the understanding of nonlinear seepage and solute transport in partially filled rough fractures. Key Points Critical non‐Darcy factor E = 0.02 can be used to determine the onset of the non‐Darcy regime Volume evolution of eddy region can be classified into stable, slow nonlinear growth, and rapid linear growth stages A nonlinear seepage model is developed for partially filled rough fractures

We demonstrate that the inference of galaxy stellar masses via spectral energy distribution (SED) fitting techniques for galaxies formed in the first billion years after the Big Bang carries fundamental uncertainties owing to the loss of star formation history (SFH) information from the very first episodes of star formation in the integrated spectra of galaxies. While this early star formation can contribute substantially to the total stellar mass of high-redshift systems, ongoing star formation at the time of detection outshines the residual light from earlier bursts, hampering the determination of accurate stellar masses. As a result, order-of-magnitude uncertainties in stellar masses can be expected. We demonstrate this potential problem via direct numerical simulation of galaxy formation in a cosmological context. In detail, we carry out two cosmological simulations with significantly different stellar feedback models, which span a significant range in SFH burstiness. We compute the mock SEDs for these model galaxies at z = 7 via calculations of 3D dust radiative transfer, and then backward fit these SEDs with prospector SED fitting software. The uncertainties in derived stellar masses that we find for z > 7 galaxies motivate the development of new techniques and/or priors for SFH to model star formation in the early Universe.

Time integration of the unsteady Reynolds-averaged Navier–Stokes (URANS) equations is the principal approach used in numerical weather prediction. This approach represents a balanced compromise between accuracy and computational cost. The URANS equations require the flow to be decomposed into an ensemble mean and excursions that are presumed to be entirely related to turbulence, thereby enabling conventional closure schemes to be used to describe their statistics. Implicit in such a decomposition is the assumption of a spectral gap between the unsteadiness in the mean flow and the scales of turbulence. Modelling challenges arise when some of the unresolved fluctuations are related to non-turbulent, structured motions that can also blur the spectral gap and render conventional closure schemes ineffective. This work seeks to clarify modelling issues that occur when unresolved fluctuations include submesoscale motions and persistent secondary circulations related to surface heterogeneities. Because submeso motions and persistent secondary circulations are not random, new theoretical tactics are discussed to represent their effects on URANS transport. By reviewing the interpretation of fluctuating terms in the URANS equations, we suggest the use of large-eddy simulations, direct numerical simulations and field measurements to guide the development of closure schemes that explicitly include fluxes due to submeso motions and persistent secondary circulations.

This study numerically investigates the vortex-induced vibration (VIV) of an elastically mounted rigid cylinder by using Reynolds-averaged Navier-Stokes (RANS) equations with computational fluid dynamic (CFD) tools. CFD analysis is performed for a fixed-cylinder case with Reynolds number (Re) = 104 and for a cylinder that is free to oscillate in the transverse direction and possesses a low mass-damping ratio and Re = 104. Previously, similar studies have been performed with 3-dimensional and comparatively expensive turbulent models. In the current study, the capability and accuracy of the RANS model are validated, and the results of this model are compared with those of detached eddy simulation, direct numerical simulation, and large eddy simulation models. All three response branches and the maximum amplitude are well captured. The 2-dimensional case with the RANS shear-stress transport k-w model, which involves minimal computational cost, is reliable and appropriate for analyzing the characteristics of VIV.