Research on the Effects of in-Line Oscillatory Flow on the Vortex-Induced Motions of A Deep Draft Semi-Submersible in Currents

A Deep Draft Semi-submersible (DDS) under certain flow conditions could be subjected to Vortex-Induced Motions (VIM), which significantly influences the loads on and life fatigue of the moorings and the risers. To investigate the VIM of a DDS with four rectangular section columns in waves coupled with a uniform current, a numerical study using the computational fluid dynamics (CFD) method was conducted. The issues of the VIM of multi-column floaters can be conveniently converted to the issues of oscillating cylinders in fluid cross flows. This paper looks into the CFD numerical simulation of infinite cylinders having rectangular sections in a two-dimensional sinusoidal time-dependent flow field coupled with a uniform current. The resulted hydrodynamic forces and motion responses in different oscillatory flows plus currents both aligned in the same direction for the incidence of 135° of the DDS relative to the flow are compared with the ones in current only cases. The results show that the VIM response of this geometric arrangement of a DDS with four rectangular columns in a current combined with oscillatory flows is more evident than that in the current only case. The oscillatory flows and waves have the significant influence on the VIM response, forces and trajectory, in-plane motions of the DDS.     

Intense near-bed sediment motions in waves and currents

In the past few years, two-phase models have been developed to describe the detail behaviour of fluid/sediment interactions and transport under the sheet flow conditions. Due to the complexity of the governing equations and uncertainties in the formulations of various stress terms, few complete solutions of these equations are known and the validations are thus far limited to only a few experimental data. In this paper, the numerical predictions of the behaviour of sheet flows using an improved version of an earlier two-phase flow model [Coastal Eng. 36(2) (1999) 87] are described. Although the general structure of the model was retained, a number of improvements had been made to give better account the underlying physics of the flow in areas very close to the stationary bed. All key flow parameters have been predicted and analysed in order to gain insight into the processes. Calculated time-dependent as well as time-averaged concentrations are compared with experimental data from purely oscillatory flows and oscillatory flow plus a current. Good qualitative agreements between predictions and measurements were achieved for the time-dependent concentrations while the time-averaged concentrations are quantitatively accurate as well.     

Vertical structure of time-dependent currents in a mid-ocean ridge axial valley

Velocity measurements with high vertical resolution, and a two-dimensional linear quasianalytic model for subinertial oscillatory flows, are used to analyze the vertical structure of flow in the axial valley at Endeavour Segment, Juan de Fuca Ridge. At a site away from hydrothermal vents, observed semidiurnal flows are independent of depth, rectilinear and parallel to the valley axis, while subinertial flows are intensified and re-aligned along-valley toward the bottom. This behavior is consistent with solutions from the model, which show attenuation of subinertial across-valley flow with depth. This cross-flow attenuation is most pronounced for valleys with widths less than the internal Rossby radius of deformation. Reduction of across-valley flow with depth results in a weakened Coriolis force that cannot fully balance the along-valley pressure-gradient force. The resulting force imbalance yields a directly accelerated bottom-intensified along-valley flow. The importance of this physical process in other submarine valleys depends on their geometry, stratification and latitude. If active, this mechanism provides a dynamic background environment for the axial valley to which hydrothermal venting would add complexity. The strong vertical shears and spiraling flows observed within the axial valley for diurnal tidal and lower-frequency flows have important implications in the transport of hydrothermal vent fluid and the dispersal of larvae of vent organisms by bottom currents.     

海底滑坡对海洋单桩冲击力试验研究

为了研究海底滑坡对海洋单桩的冲击力大小,首先通过调整高岭土、粉砂的不同含量,得到不同流变特性、不同密度的碎屑流,采用Herschel-Bulkley模型和幂率模型对流体流变性质进行描述;随后利用自制海底滑坡模型槽,模拟碎屑流在不同流速和黏度下对模型桩的冲击;并结合流体力学理论,建立阻力系数与非牛顿流体雷诺数之间关系表达式。试验数据表明:碎屑流黏度和流速是影响海底滑坡冲击力的主要因素,海底滑坡冲击力随着泥浆黏度和流速的增加而增大。同时,考虑碎屑流剪切稀释特性,得到管桩阻力系数随雷诺数变化的拟合公式,为海洋桩基础设计提供参考。     

Sand ripples generated by regular oscillatory flow

The prediction of ripple geometry is a necessary precursor to the prediction of sand transport under waves for ripple regime conditions. The paper begins with a comparison of four existing methods for predicting the geometry of sand ripples generated by oscillatory flow. The comparison points to substantial differences between ripple dimensions predicted by the methods, especially for field-scale conditions. Ripple geometry experiments carried out in a large oscillatory flow tunnel are then described. The experiments involved a range of sand sizes and sinusoidal and asymmetric flows with periods and velocities typical of field conditions. Comparison of measured and predicted ripple geometries leads to the recommendation that the method of Mogridge, Davies and Willis be used to predict ripple geometry for field-scale oscillatory flows. The Nielsen method yields good predictions of ripple length, but the rapid fall-off in ripple steepness predicted by the Nielsen method at high mobility number is not supported by the measurements. The lengths and heights of symmetric ripples produced by sinusoidal flows are found to be similar to the lengths and heights of asymmetric ripples produced by “equivalent” asymmetric flows. Three-dimensional ripples occur with fine sand in long-period flows typical of field conditions. The dimensions of these ripples cannot be predicted using methods developed for two-dimensional ripples. Previously suggested criteria for predicting the occurrence of three-dimensional ripples fail when tested against a wide range of flow and sand conditions. The occurrence of three-dimensional ripples and the effects of ripple and flow history on ripple geometry require further research.     

Three-dimensional instability of shear flows with inflection-free velocity profiles in stratified media with a high Prandtl number

In stably stratified media with a Prandtl number Pr ≫ 1, vertical scales of the density (ℓ) and horizontal velocity variation (L) are quite different, ℓ/L = O(Pr^−1/2) ≪ 1, and this influences the flow stability. In particular, shear flows without inflection points on the velocity profile are unstable even in an ideal incompressible fluid. The maximum instability growth rate for sufficiently small ℓ/L is of the same order as in homogeneous mixing layers, with mainly three-dimensional rather than two-dimensional oscillations increasing in a wide range of parameters. This paper focuses on the three-dimensional instability of such flows. It is shown that the spectrum of unstable oscillations is essentially anisotropic in the case of a relatively weak stratification when the bulk Richardson number J ≤ O[(ℓ/L)^3/2]. The results of the asymptotic analysis are illustrated by calculations for a model flow in a two-layer medium (ℓ = 0) as well as for flows with values of ℓ/L corresponding to a temperature or salinity stratification of the water.     

Calculation of wave-induced turbulent flows in estuaries

The k-ε turbulence model which relates the eddy viscosity to turbulent kinetic energy, k, and to the rate of its dissipation, ε, and determines the distribution of these two quantities from modeled transport equations, is employed in calculating the vertical structure of wave-induced turbulent flows in two-dimensional estuaries. The empirical constants in this model are given the standard values cited in the literature and used successfully for calculating a large variety of steady flows; they are not tuned to the oscillating flows. The free surface elevation and the eddy-viscosity distribution are also calculated as a part of the solution. First, calculations performed for oscillatory laminar flows are compared with analytic solutions to ensure the proper performance of the numerical scheme. The turbulent flow in a laboratory-estuary model with one end closed, and in the Humber Estuary is simulated with the numerical model, and the results are compared with the corresponding flume and field measurements. The influence of frequency and roughness is thereby investigated. The results indicate that the time and space variation of the velocity field and the phase lag between the surface slope and the horizontal velocity can be predicted satisfactorily in wave-induced turbulent flows.     

Resonant interaction of waves of continuous and discrete spectra in the simplest model of a stratified shear flow

The equations of dynamics of eddy—wave disturbances of two-dimensional stratified flows in an ideal incompressible fluid that are written in a Hamiltonian form are used to study the resonant interaction of waves of discrete and continuous spectra. A gravity—shear wave generated at a jump of the density and vorticity of the undisturbed flow and a wave generated at a weak vorticity jump, which is similar to a wave of a continuous spectrum, participate in the interaction. The equations are written in terms of normal variables to obtain the system of evolution equations for the amplitudes of the interacting waves. The stability condition for eddy—wave disturbances is derived within the framework of the linear theory. It is shown that a cubic nonlinearity may lead to the stabilization of unstable disturbances if the coefficient of the nonlinear term is positive.     

下凹内孤立波致流场结构及其影响因素的实验研究

为进一步探究海洋内孤立波诱导流场对海洋工程结构物以及潜航器的影响,本文采用重力塌陷方法和粒子图像测速(Particle Image velocimetry,PIV)技术在大型分层流水槽中进行内孤立波造波以及内部流速场测量,定量分析了下凹型内孤立波诱导流场结构及其影响因素。研究表明:在密度分层流体中,PIV技术可实现对大幅面内孤立波诱导流场的精细测量以及波动结构特征的准确描述;水平流速在上下层方向相反且在跃层处最小,其剪切作用在波谷附近最强;垂向流动在波前和波后分别为上升和下沉流,两者流速值在距离波谷1/4~1/2波长位置达到最大;在相同内孤立波振幅条件下,上下层流体密度差越大、厚度比越小,则波致流场越强;随着振幅增大,流场结构与Kd V、e Kd V和MCC理论模型对应波幅适用范围的描述相吻合。     

The dimensions of sand ripples in full-scale oscillatory flows

New large-scale experiments have been carried out in two oscillatory flow tunnels to study ripple regime sand suspension and net sand transport processes in full-scale oscillatory flows. The paper focuses on ripple dimensions and the new data are combined with existing data to make a large dataset of ripple heights and lengths for flows with field-scale amplitudes and periods. A feature of the new experiments is a focus on the effect of flow irregularity. The combined dataset is analysed to examine the range of hydraulic conditions under which oscillatory flow ripples occur, to examine the effects of flow irregularity and ripple three-dimensionality on ripple dimensions and to test and improve existing methods for predicting ripple dimensions.The following are the main conclusions. (1) The highest velocities in a flow time-series play an important role in determining the type of bedform occurring in oscillatory flow. Bedform regime is well characterised by mobility number based on maximum velocity in the case of regular flow and based on the mean of the highest one tenth peak velocities in the case of irregular flow. (2) For field-scale flows, sand size is the primary factor determining whether equilibrium ripples will be 2D or 3D. 2D ripples occur when the sand D₅₀ ≥ 0.30 mm and 3D ripples occur when D₅₀ ≤ 0.22 mm (except when the flow orbital diameter is low). (3) Ripple type (2D or 3D) is the same for regular and irregular flows and ripple dimensions produced by equivalent regular and irregular flows follow a similar functional dependence on mobility number, with mobility number based on maximum velocity in the case of regular flow and based on the mean of the highest one tenth velocities in the case of irregular flow. For much of the ripple regime, ripple dimensions have weak dependency on mobility number and ripple dimensions are similar for regular and irregular flows with the same flow orbital amplitude. However, differences in ripples produced by equivalent regular and irregular flows become significant at the high mobility end of the ripple regime. (4) Ripple dimensions predicted using the Wiberg and Harris formulae are in poor agreement with measured ripple dimensions from the large-scale experiments. Predictions based on the Mogridge et al. and the Nielsen formulae show better overall agreement with the data but also show systematic differences in cases of 3D ripples and ripples generated by irregular flows. (5) Based on the combined large-scale data, modifications to the Nielsen ripple dimension equations are proposed for the heights and lengths of 2D ripples. The same equations apply to regular and irregular flows, but with mobility number appropriately defined. 3D ripples are generally smaller than 2D ripples and estimates of 3D ripple height and length may be obtained by applying multipliers of 0.55 and 0.73 respectively to the 2D formulae. The proposed modified Nielsen formulae provide an improved fit to the large-scale data, accounting for flow irregularity and ripple three-dimensionality.     

Double diffusive convection in a rotating system

Linear instability theory of double diffusive convection is generalized to a rotating system. For sufficiently large rotation, rotation has a stabilizing effect, and viscosity stabilizes most of oscillatory modes but destabilizes direct modes. The following behavior is found in oscillatory modes in the case of destabilizing temperature distribution and stabilizing solute distribution. With the fixed destabilizing temperature distribution (1) the system is stable when the stabilizing solute gradient is large, (2) it becomes unstable with respect to oscillatory modes when the gradient is less than a certain value. (3) When the gradient is reduced further, the system becomes stable again. This anomalous phenomenon happens only when the Prandtle number is less than unity and larger than the ratio of the diffusivity of solute and that of temperature, under the existence of rotation.     

A two-phase numerical model for sediment transport prediction under oscillatory sheet flows

To predict sediment transport under oscillatory sheet flow condition, especially for fine sand, is still a challenging research subject in coastal engineering. This paper describes a newly-developed numerical model based on two-phase theory with the use of a one-equation turbulence closure, and its applications in predicting fine sediment suspension in near-prototype oscillatory sheet flow conditions. Model results were compared with comprehensive laboratory measurements of flow velocity and sediment concentration under both symmetrical and asymmetrical oscillatory sheet flows from a large-scale water tunnel. Good agreements between the model results and measurements were achieved and the results demonstrated that the model is capable of reproducing detailed characteristics of sediment entrainment process in the sheet flow regime. The comparisons also revealed the fact that the concentration peaks at flow reversal is associated with the strong vertical sediment transport flux in the pickup layer, which has been widely observed in many laboratory experiments. The effects of flow reversal events on total sediment transport were also discussed.     

基于MPS方法模拟低雷诺数方柱绕流问题

柱体绕流问题是流体力学领域一个非常经典的问题。当流体流经柱体时,由于黏性的存在,会发生许多复杂的流动现象,如流动分离、涡旋周期性生成与脱落等,经常被作为标准验证算例。同时,柱体绕流广泛存在于实际工程中,并在一定工况下可能对工程产生巨大危害,因此对柱体绕流进行深入研究具有重要意义。研究中,拟将一种无网格类方法——半隐式移动粒子方法(moving particle semi-implicit method,简称MPS)引入到柱体绕流问题的数值研究中,并对不同雷诺数下二维方柱绕流问题进行数值模拟。首先,使用基于MPS方法自主开发的MLParticle-SJTU求解器,结合入口边界条件和出口边界条件,模拟了雷诺数Re分别为40、200和1 000时均匀来流条件下的方柱绕流。随后,将模拟的绕流结果与文献中试验和数值计算结果进行了对比,结果吻合较好,并且在雷诺数为200和1 000时,可以清晰地捕捉到方柱尾流中的卡门涡街现象,验证了MPS方法在柱体绕流问题模拟上的有效性和适用性。     

Numerical simulation of oscillatory flows over a rippled bed by immersed boundary method

In this paper, a well-developed numerical model based on the immersed boundary (IB) method is used to study oscillatory flows over a bed with large-amplitude ripples in a systematic manner. The work shows that the complex flow over the rippled bed can be numerically dealt with in Cartesian coordinate by the IB method and that the IB method is able to provide main features of the flows near the ripples. An accurate simulation of vortices generation as a result of flow separation at the rippled bed is obtained. It is found that the oscillatory flows start to separate during the flow deceleration when the Keulegan–Carpenter (KC) number is small. The steady streaming for various ripple steepness is simulated and the criterion for separating the single and double structure streaming is also discussed. Moreover, a new type of steady streaming which consists of a pair of embedded recirculations in the vicinity of the ripple trough is obtained for relatively steep ripples in this work. The numerical results, including the steady streaming in particular, may be helpful to improve the understanding of the sediment transport and the seabed evolution with natural ripples under sea waves.     

Velocity and Sediment Concentration Profiles in Sediment-Laden Flows

A theoretical analysis of velocity profiles in sediment-laden flows is presented by means of Prandtl-Karman mixing length theorem. The study shows that the upward velocity of liquid-phase caused by settling sediment leads to the invalidity of the log-law and Rouse equation. The theoretical analysis takes into account the upward velocity and shows: 1) the mean velocity in sediment-laden flows follows the log-law, but the Karman constant reduces in the main flow region, 2) sediment concentration reduces the mixing length of fluid particles, 3) flow resistance reduces with the presence of sediment concentration, and 4) the sediment concentration profile deviates from the well know Rouse equation. The experimental data agree well with the equations derived on the basis of non-zero wall velocity. It is found that the wall-normal velocity should not be neglected for density gradient flows because it induces more than for pure water flows.     

Measurements of sheet flow transport in acceleration-skewed oscillatory flow and comparison with practical formulations

Near-bed oscillatory flows with acceleration skewness are characteristic of steep and breaking waves in shallow water. In order to isolate the effects of acceleration skewness on sheet flow sand transport, new experiments are carried out in the Aberdeen Oscillatory Flow Tunnel. The experiments have produced a dataset of net transport rates for full-scale oscillatory flows with varying degrees of acceleration skewness and three sand sizes. The new data confirm previous research that net transport in acceleration-skewed flow is non-zero, is always in the direction of the largest acceleration and increases with increasing acceleration skewness. Large transport rates for the fine sand conditions suggest that phase lag effects play an important role in augmenting positive net transport. A comparison of the new experimental data with a number of practical sand transport formulations that incorporate acceleration skewness shows that none of the formulations performs well in predicting the measured net transport rates for both the fine and the coarser sands. The new experimental data can be used to further develop practical sand transport formulations to better account for acceleration skewness.     

Wave-induced sediment transport and onshore sandbar migration

The 25-m onshore migration of a nearshore sandbar observed over a 5-day period near Duck, NC, is simulated with a simplified, computationally efficient, wave-resolving single-phase model. The modeled sediment transport is assumed to occur close to the seabed and to be in phase with the bottom stress. Neglected intergranular stresses and fluid–granular interactions, likely important in concentrated flow, are compensated for with an elevated (relative to that appropriate for a clear fluid) model roughness height that gives the best fit to the observed bar migration. Model results suggest that when mean-current-induced transport is small, wave-induced transport leads to the observed onshore bar migration. Based on the results from the simplified phase-resolving model, a wave-averaged, energetics-type model (e.g., only moments of the near-bottom velocity field are required) with different friction factors for oscillatory and mean flows is developed that also predicts the observed bar migration. Although the assumptions underlying the models differ, the similarity of model results precludes determination of the dominant mechanisms of sediment transport during onshore bar migration.     

The inception of sheet flow in oscillatory flow

A simple model is developed to study the inception of sheet flow in oscillatory flow based on the available experimental data. The inception of sheet flow in oscillatory flow is well defined by the simple model: A/d=KA²ω/ν+B, where A is the semi-excursion of wave orbital motion near the bed, d is the grain size, ω is the angular frequency, ν is the kinematic viscosity of water, and K and B are the coefficients and dependent on sediment properties only. The inception velocity of sheet flow derived from the model is shown to be the function of grain size d, oscillatory period T and specific sediment density s. For a given sediment, the inception velocity is found to increase sharply initially with T and then approach a constant at T>6.0 s. The present model is quite simple and gives good agreement with the available experimental data.     

The inception of sheet flow in oscillatory flow

A simple model is developed to study the inception of sheet flow in oscillatory flow based on the available experimental data. The inception of sheet flow in oscillatory flow is well defined by the simple model: A/d=KA²ω/ν+B, where A is the semi-excursion of wave orbital motion near the bed, d is the grain size, ω is the angular frequency, ν is the kinematic viscosity of water, and K and B are the coefficients and dependent on sediment properties only. The inception velocity of sheet flow derived from the model is shown to be the function of grain size d, oscillatory period T and specific sediment density s. For a given sediment, the inception velocity is found to increase sharply initially with T and then approach a constant at T>6.0 s. The present model is quite simple and gives good agreement with the available experimental data.