共查询到20条相似文献，搜索用时 31 毫秒
1.
《Advances in water resources》2005,28(9):950-955
We describe an approach to parallelize Eulerian–Lagrangian localized adjoint methods such that no errors are introduced compared to the sequential case. This parallelization approach fully captures the hyperbolic features of the underlying problem. It uses an overlapping domain decomposition technique, and does not involve the introduction of artificial boundary conditions between subdomains. Implementation details on different parallel architectures are discussed. 相似文献
2.
Modified Mixed Lagrangian–Eulerian Method Based on Numerical Framework of MT3DMS on Cauchy Boundary 
下载免费PDF全文
下载免费PDF全文 Heejun Suk 《Ground water》2016,54(4):508-520
MT3DMS, a modular three‐dimensional multispecies transport model, has long been a popular model in the groundwater field for simulating solute transport in the saturated zone. However, the method of characteristics (MOC), modified MOC (MMOC), and hybrid MOC (HMOC) included in MT3DMS did not treat Cauchy boundary conditions in a straightforward or rigorous manner, from a mathematical point of view. The MOC, MMOC, and HMOC regard the Cauchy boundary as a source condition. For the source, MOC, MMOC, and HMOC calculate the Lagrangian concentration by setting it equal to the cell concentration at an old time level. However, the above calculation is an approximate method because it does not involve backward tracking in MMOC and HMOC or allow performing forward tracking at the source cell in MOC. To circumvent this problem, a new scheme is proposed that avoids direct calculation of the Lagrangian concentration on the Cauchy boundary. The proposed method combines the numerical formulations of two different schemes, the finite element method (FEM) and the Eulerian–Lagrangian method (ELM), into one global matrix equation. This study demonstrates the limitation of all MT3DMS schemes, including MOC, MMOC, HMOC, and a third‐order total‐variation‐diminishing (TVD) scheme under Cauchy boundary conditions. By contrast, the proposed method always shows good agreement with the exact solution, regardless of the flow conditions. Finally, the successful application of the proposed method sheds light on the possible flexibility and capability of the MT3DMS to deal with the mass transport problems of all flow regimes. 相似文献
3.
Under constant hydrodynamic conditions and assuming horizontal homogeneity, negatively buoyant particles released at the surface of the water column have a mean residence time in the surface mixed layer of h/w, where h is the thickness of the latter and w (?>?0) is the sinking velocity Deleersnijder (Environ Fluid Mech 6(6):541–547, 2006a). The residence time does not depend on the diffusivity and equals the settling timescale. We show that this behavior is a result of the particular boundary conditions of the problem and that it is related to a similar property of the exposure time in a one-dimensional infinite domain. In 1-D advection–diffusion problem with a constant and uniform velocity, the exposure time—which is a generalization of the residence time measuring the total time spent by a particle in a control domain allowing the particle to leave and reenter the control domain—is also equal to the advection timescale at the upstream boundary of the control domain. To explain this result, the concept of point exposure is introduced; the point exposure is the time integral of the concentration at a given location. It measures the integrated influence of a point release at a given location and is related to the concept of number of visits of the theory of random walks. We show that the point exposure takes a constant value downstream the point of release, even when the diffusivity varies in space. The analysis of this result reveals also that the integrated downstream transport of a passive tracer is only effected by advection. While the diffusion flux differs from zero at all times, its integrated value is strictly zero. 相似文献
4.
V. N. Zyryanov 《Water Resources》2011,38(3):261-273
A closed, ice-covered water body containing water with homogeneous density distribution is considered. No-slip conditions
are specified for flow velocity at the lower ice boundary and on the bed. Two variants of boundary conditions are considered
on the side boundary: the boundary is either a solid vertical wall with a finite liquid depth or the liquid wedges out to
zero depth values. Ice either is attached fast to the shore or is separated from it by an open-water zone. A basic fourth
order equation is derived for the amplitude of ice oscillations. The introduction of friction results in the appearance of
reduced depth. Eigenvalue problem is considered for evaluating seiche periods. For the case when the liquid wedges out at
the shore, the basic equation becomes singular at water body boundaries. A lake with a longitudinal depth profile, which can
be approximated by a parabola, is considered. The solution is found by the method of matched asymptotic expansions. In the
inner domain, beyond the boundary layers, the equation is reduced to Legendre equation, which yields a new relationship for
the spectrum of seiche oscillations both in the presence of ice and in an open lake with varying depth. Boundary layers appear
at the margins of the lake, where the liquid wedges out; a solution is found for these layers. 相似文献
5.
W.A. Mulder 《Geophysical Prospecting》2020,68(9):2857-2866
Simulations of wave propagation in the Earth usually require truncation of a larger domain to the region of interest to keep computational cost acceptable. This introduces artificial boundaries that should not generate reflected waves. Most existing boundary conditions are not able to completely suppress all the reflected energy, but suffice in practice except when modelling subtle events such as interbed multiples. Exact boundary conditions promise better performance but are usually formulated in terms of the governing wave equation and, after discretization, still may produce unwanted artefacts. Numerically exact non-reflecting boundary conditions are instead formulated in terms of the discretized wave equation. They have the property that the numerical solution computed on a given domain is the same as one on a domain enlarged to the extent that waves reflected from the boundary do not have the time to reach the original truncated domain. With a second- or higher-order finite-difference scheme for the one-dimensional wave equation, these boundary conditions follow from a recurrence relation. In its generalization to two or three dimensions, a recurrence relation was only found for a single non-reflecting boundary on one side of the domain or two of them at opposing ends. The other boundaries should then be zero Dirichlet or Neumann. If two non-reflecting boundaries meet at a corner, translation invariance is lost and a simple recurrence relation could not be found. Here, a workaround is presented that restores translation invariance by imposing classic, approximately non-reflecting boundary conditions on the other sides and numerically exact ones on the two opposing sides that otherwise would create the strongest reflected waves with the classic condition. The exact ones can also be applied independently. As a proof of principle, the method is applied to the two-dimensional acoustic wave equation, discretized on a rectangular domain with a second-order finite-difference scheme and first-order Enquist–Majda boundary conditions as approximate ones. The method is computationally costly but has the advantage that it can be reused on a sequence of problems as long as the time step and the sound speed values next to the boundary are kept fixed. 相似文献
6.
The inherent heterogeneity of geological media often results in anomalous dispersion for solute transport through them, and how to model it has been an interest over the past few decades. One promising approach that has been increasingly used to simulate the anomalous transport in surface and subsurface water is the fractional advection–dispersion equation (FADE), derived as a special case of the more general continuous time random walk or the stochastic continuum model. In FADE, the dispersion is not local and the solutes have appreciable probability to move long distances, and thus reach the boundary faster than predicted by the classical advection–dispersion equation (ADE). How to deal with different boundaries associated with FADE and their consequent impact is an issue that has not been thoroughly explored. In this paper we address this by taking one-dimensional solute movement in soil columns as an example. We show that the commonly used FADE with its fractional derivatives defined by the Riemann–Liouville definition is problematic and could result in unphysical results for solute transport in bounded domains; a modified method with the fractional dispersive flux defined by the Caputo derivatives is presented to overcome this problem. A finite volume approach is given to numerically solve the modified FADE and its associated boundaries. With the numerical model, we analyse the inlet-boundary treatment in displacement experiments in soil columns, and find that, as in ADE, treating the inlet as a prescribed concentration boundary gives rise to mass-balance errors and such errors could be more significant in FADE because of its non-local dispersion. We also discuss a less-documented but important issue in hydrology: how to treat the upstream boundary in analysing the lateral movement of tracer in an aquifer when the tracer is injected as a pulse. It is shown that the use of an infinite domain, as commonly assumed in literature, leads to unphysical backward dispersion, which has a significant impact on data interpretation. To avoid this, the upstream boundary should be flux-prescribed and located at the upstream edge of the injecting point. We apply the model to simulate the movement of Cl− in a tracer experiment conducted in a saturated hillslope, and analyse in details the significance of upstream-boundary treatments in parameter estimation. 相似文献
7.
The MARS-3D model in conjunction with the particle tracking module Ichthyop is used to study circulation and tracer dynamics under a variety of forcing conditions in the eastern English Channel, and in the Boulogne-sur-Mer harbour (referred to hereafter as BLH). Results of hydrodynamic modelling are validated against the tidal gauge data, VHF radar surface velocities and ADCP measurements. Lagrangian tracking experiments are performed with passive particles to study tracer dispersal along the northern French coast, with special emphasis on the BLH. Simulations revealed an anticyclonic eddy generated in the harbour at rising tide. Tracers, released during flood tide at the Liane river mouth, move northward with powerful clockwise rotating current. After the high water, the current direction changes to westward, and tracers leave the harbour through the open boundary. During ebb tide, currents convergence along the western open boundary but no eddy is formed, surface currents inside the harbour are much weaker and the tracer excursion length is small. After the current reversal at low water, particles are advected shoreward resulting in a significant increase of the residence time of tracers released during ebb tide. The effect of wind on particle dispersion was found to be particularly strong. Under strong SW wind, the residence time of particles released during flood tide increases from 1.5 to 6 days. For release during ebb tide, SW wind weakens the southward tidally induced drift and thus the residence time decreases. Similar effects are observed when the freshwater inflow to the harbour is increased from 2 to 10 m3/s during the ebb tide flow. For flood tide conditions, the effect of freshwater inflow is less significant. We also demonstrate an example of innovative coastal management targeted at the reduction of the residence time of the pathogenic material accidentally released in the harbour. 相似文献
8.
The MARS-3D model in conjunction with the particle tracking module Ichthyop is used to study circulation and tracer dynamics under a variety of forcing conditions in the eastern English Channel, and in the Boulogne-sur-Mer harbour (referred to hereafter as BLH). Results of hydrodynamic modelling are validated against the tidal gauge data, VHF radar surface velocities and ADCP measurements. Lagrangian tracking experiments are performed with passive particles to study tracer dispersal along the northern French coast, with special emphasis on the BLH. Simulations revealed an anticyclonic eddy generated in the harbour at rising tide. Tracers, released during flood tide at the Liane river mouth, move northward with powerful clockwise rotating current. After the high water, the current direction changes to westward, and tracers leave the harbour through the open boundary. During ebb tide, currents convergence along the western open boundary but no eddy is formed, surface currents inside the harbour are much weaker and the tracer excursion length is small. After the current reversal at low water, particles are advected shoreward resulting in a significant increase of the residence time of tracers released during ebb tide. The effect of wind on particle dispersion was found to be particularly strong. Under strong SW wind, the residence time of particles released during flood tide increases from 1.5 to 6 days. For release during ebb tide, SW wind weakens the southward tidally induced drift and thus the residence time decreases. Similar effects are observed when the freshwater inflow to the harbour is increased from 2 to 10 m3/s during the ebb tide flow. For flood tide conditions, the effect of freshwater inflow is less significant. We also demonstrate an example of innovative coastal management targeted at the reduction of the residence time of the pathogenic material accidentally released in the harbour.
相似文献9.
Under constant hydrodynamic conditions and assuming horizontal homogeneity, negatively buoyant particles released at the surface of the water column have a mean residence time in the surface mixed layer of h/w, where h is the thickness of the latter and w ( > 0) is the sinking velocity Deleersnijder (Environ Fluid Mech 6(6):541–547, 2006a). The residence time does not depend on the diffusivity and equals the settling timescale. We show that this behavior is a result of the particular boundary conditions of the problem and that it is related to a similar property of the exposure time in a one-dimensional infinite domain. In 1-D advection–diffusion problem with a constant and uniform velocity, the exposure time—which is a generalization of the residence time measuring the total time spent by a particle in a control domain allowing the particle to leave and reenter the control domain—is also equal to the advection timescale at the upstream boundary of the control domain. To explain this result, the concept of point exposure is introduced; the point exposure is the time integral of the concentration at a given location. It measures the integrated influence of a point release at a given location and is related to the concept of number of visits of the theory of random walks. We show that the point exposure takes a constant value downstream the point of release, even when the diffusivity varies in space. The analysis of this result reveals also that the integrated downstream transport of a passive tracer is only effected by advection. While the diffusion flux differs from zero at all times, its integrated value is strictly zero.
相似文献10.
A Lagrangian perturbation method is applied to develop a method of moments for solute flux through a three-dimensional nonstationary flow field. The flow nonstationarity stems from medium nonstationarity and internal and external boundaries of the study domain. The solute flux is described as a space-time process where time refers to the solute flux breakthrough through a control plane (CP) at some distance downstream of the solute source and space refers to the transverse displacement distribution at the CP. The analytically derived moment equations for solute transport in a nonstationarity flow field are too complicated to solve analytically, a numerical finite difference method is implemented to obtain the solutions. This approach combines the stochastic model with the flexibility of the numerical method to boundary and initial conditions. The developed method is applied to study the effects of heterogeneity and nonstationarity of the hydraulic conductivity and chemical sorption coefficient on solute transport. The study results indicate all these factors will significantly influence the mean and variance of solute flux. 相似文献
11.
Sébastien Blaise Benjamin de Brye Anouk de Brauwere Eric Deleersnijder Eric J. M. Delhez Richard Comblen 《Ocean Dynamics》2010,60(3):535-554
At high Peclet number, the residence time exhibits a boundary layer adjacent to incoming open boundaries. In a Eulerian model,
not resolving this boundary layer can generate spurious oscillations that can propagate into the area of interest. However,
resolving this boundary layer would require an unacceptably high spatial resolution. Therefore, alternative methods are needed
in which no grid refinement is required to capture the key aspects of the physics of the residence time boundary layer. An
extended finite element method representation and a boundary layer parameterisation are presented and tested herein. It is
also explained how to preserve local consistency in reversed time simulations so as to avoid the generation of spurious residence
time extrema. Finally, the boundary layer parameterisation is applied to the computation of the residence time in the Scheldt
Estuary (Belgium/The Netherlands). This timescale is simulated by means of a depth-integrated, finite element, unstructured
mesh model, with a high space–time resolution. It is seen that the residence time temporal variations are mainly affected
by the semi-diurnal tides. However, the spring–neap variability also impacts the residence time, particularly in the sandbank
and shallow areas. Seasonal variability is also observed, which is induced by the fluctuations over the year of the upstream
flows. In general, the residence time is an increasing function of the distance to the mouth of the estuary. However, smaller-scale
fluctuations are also present: they are caused by local bathymetric features and their impact on the hydrodynamics. 相似文献
12.
Michela De Dominicis Giovanni Leuzzi Paolo Monti Nadia Pinardi Pierre-Marie Poulain 《Ocean Dynamics》2012,62(9):1381-1398
Ocean transport and dispersion processes are at the present time simulated using Lagrangian stochastic models coupled with Eulerian circulation models that are supplying analyses and forecasts of the ocean currents at unprecedented time and space resolution. Using the Lagrangian approach, each particle displacement is described by an average motion and a fluctuating part. The first one represents the advection associated with the Eulerian current field of the circulation models while the second one describes the sub-grid scale diffusion. The focus of this study is to quantify the sub-grid scale diffusion of the Lagrangian models written in terms of a horizontal eddy diffusivity. Using a large database of drifters released in different regions of the Mediterranean Sea, the Lagrangian sub-grid scale diffusion has been computed, by considering different regimes when averaging statistical quantities. In addition, the real drifters have been simulated using a trajectory model forced by OGCM currents, focusing on how the Lagrangian properties are reproduced by the simulated trajectories. 相似文献
13.
《Advances in water resources》2004,27(4):323-334
A forward particle tracking Eulerian Lagrangian localized adjoint method (ELLAM) is applied to the multicomponent reactive transport problem using a split operator approach. Two split operator algorithms are compared, the Strang algorithm and the sequential non-iterative algorithm (SNIA). The reaction equations are integrated using a coupled predictor corrector algorithm with adaptive time stepping. Reaction time steps are adjusted at the inflow boundary to reflect the actual time of transport inside the solution domain.Results show that split operator ELLAM formulations are competitive with direct or fully coupled ELLAM solutions for reactive transport problems. The SNIA algorithm is more accurate than the Strang splitting algorithm when large time steps are used. The reaction algorithm employed dominates computational effort in runs with large time step sizes. To illustrate the use of the method in practical problems, the model is fitted to aerobic aniline degradation data from laboratory scale column experiments. Model inversion is achieved using non-linear regression with a shuffled complex evolution optimization algorithm and parameter uncertainty is assessed using a Bayesian uncertainty analysis procedure. 相似文献
14.
In this study, a theoretical approach is used to investigate the scattering problem of circular holes under a scalene triangle on the surface. The wave displacement function is obtained by solving the Helmholtz equation that meets the zero-stress boundary conditions by adopting the method of separation of variables. Based on the complex function, multi-polar coordinate method, and region-matching technique, algebraic equations are established at auxiliary boundaries and free boundaries conditions in a complex domain. The auxiliary circle is used to solve the singularity of the reflex angle at the triangle corner. Then, according to sample statistics, the least squares method is used instead of the Fourier expansion method to solve the undetermined coefficient of the algebraic equations by discrete boundary. Numerical results show that the continuity of the auxiliary boundaries and the accuracy of the zero-stress boundaries are adequate, and the displacement of the free surface and the stress of the circular hole are related to the shape of the triangle, the position of the circular hole, the direction of the incident wave, and the frequency content of the excitation. Finally, time-domain responses are calculated by FFT based on the frequency domain theory, and the results reveal the wave propagation mechanism in a complicated structure.
相似文献15.
Membrane polarization occurs in sediments with different surface area of capillaries (pores) and is regarded as a slow type of polarization. This phenomenon is the foundation of the well known methods of induced polarization (IP): time domain and frequency domain induced polarization. The characteristic parameters of induced polarization which are required for studying physical properties of rocks are measured in the laboratory. Data measured in the laboratory confirmed the distinctions of IP processes at time-on and time-off. Additionally linear dependence of voltage and applied current is not always observed. This paper presents the first step of studying: theoretical consideration for time-on and mathematical modeling of membrane polarization, ion concentrations of electrolyte in the pores of different models of pores space, and arising voltage. The problem of concentration of ions along the pores can be solved using the diffusion equation with specified initial and boundary conditions. Reduced boundary conditions for time-on show that transient concentrations at the boundaries are linear with time. It allows obtaining the analytical solution for this equation. Mathematical modeling has been performed for different combinations of pores. It is shown that if electrical current flows from the pores with greater transfer numbers to the pores with smaller transfer numbers, an excess of ions will be observed at this boundary. If the difference of transfer numbers is negative, there is a decrease in the concentration of ions at the vicinity of the boundary. This decrease will continue until the concentration at this boundary reaches zero. In this case the galvanic chain will be interrupted and electrical current flowing through the sample does not penetrate to this cell. The duration of the process of ions distribution in the pore and time of blockage t 0 is proportional to the radii of contacted pores and inversely proportional to the transfer number difference and square of the current flowing through this cell. It was shown by both laboratory measurement and field processes that induced polarization relates to low porous rocks with small transfer number differences. 相似文献
16.
We derive a governing second-order acoustic wave equation in the time domain with a perfectly matched layer absorbing boundary condition for general inhomogeneous media. Besides, a new scheme to solve the perfectly matched layer equation for absorbing reflections from the model boundaries based on the rapid expansion method is proposed. The suggested scheme can be easily applied to a wide class of wave equations and numerical methods for seismic modelling. The absorbing boundary condition method is formulated based on the split perfectly matched layer method and we employ the rapid expansion method to solve the derived new perfectly matched layer equation. The use of the rapid expansion method allows us to extrapolate wavefields with a time step larger than the ones commonly used by traditional finite-difference schemes in a stable way and free of dispersion noise. Furthermore, in order to demonstrate the efficiency and applicability of the proposed perfectly matched layer scheme, numerical modelling examples are also presented. The numerical results obtained with the put forward perfectly matched layer scheme are compared with results from traditional attenuation absorbing boundary conditions and enlarged models as well. The analysis of the numerical results indicates that the proposed perfectly matched layer scheme is significantly effective and more efficient in absorbing spurious reflections from the model boundaries. 相似文献
17.
《Advances in water resources》1998,22(4):319-332
The equation describing the ensemble-average solute concentration in a heterogeneous porous media can be developed from the Lagrangian (stochastic–convective) approach and from a method that uses a renormalized cumulant expansion. These two approaches are compared for the case of steady flow, and it is shown that they are related. The cumulant expansion approach can be interpreted as a series expansion of the convolution path integral that defines the ensemble-average concentration in the Lagrangian approach. The two methods can be used independently to develop the classical form for the convection–dispersion equation, and are shown to lead to identical transport equations under certain simplifying assumptions. In the development of such transport equations, the cumulant expansion does not require a priori the assumption of any particular distribution for the Lagrangian displacements or velocity field, and does not require one to approximate trajectories with their ensemble-average. In order to obtain a second-order equation, the cumulant expansion method does require truncation of a series, but this truncation is done rationally by the development of a constraint in terms of parameters of the transport field. This constraint is less demanding than requiring that the distribution for the Lagrangian displacements be strictly Gaussian, and it indicates under what velocity field conditions a second-order transport equation is a reasonable approximation. 相似文献
18.
基于双程波动方程的逆时偏移被认为是目前最好的偏移成像技术,更适合于复杂构造成像.然而,大计算量和大存储量使得逆时偏移的计算成本很高而无法用于大数据量的地震成像.本文分析了目前常用的存储策略,并分别在空间和时间上对存储策略进行了研究:空间上,根据有限差分格式,在边界存储策略的基础上通过修改波场逆向传播的边界条件,提出了有效边界存储策略.该策略可在不增加任何计算量的情况下大幅降低逆时偏移对存储量的需求;在时间上,使用checkpointing技术对有效边界存储策略进行了改进,使叠前逆时偏移在增加少量计算量的情况下进一步降低存储量需求.Marmousi模型测试结果表明了有效边界存储策略的有效性和优越性. 相似文献
19.
偏移成像是VSP数据处理中的一个重要环节,常规的VSP成像方法通常利用VSP-CDP转换或Kirchhoff偏移,均存在保幅性差及成像精度低等问题,而波动方程叠前深度偏移被认为是对地下复杂构造进行成像的精确偏移方法.任意广角波动方程作为一种高精度的空间域单程波波动方程,同时由于只含有二阶偏导数项,易于数值实现,与其他单程波波动方程相比,具有更大的成像倾角,因此是偏移成像的有力工具之一.本文将AWWE推广应用到VSP数据成像中,实现了VSP时空域高角度单程波方程偏移.首先从三维标量任意广角波动方程出发,推导了完全匹配层吸收边界条件,在基本不增加计算量的前提下有效地压制了边界反射成像噪音,同时利用非线性反演算法优选参考速度来提高平方根算子的近似程度,从而提高高角度地层的成像精度.模型数值模拟实验验证了该方法的有效性,同时表明该方法在陡倾角构造情况下能取得很好的成像效果.最后对某地区实际观测的VSP资料进行了偏移成像,并与地面地震偏移结果进行了对比,显示出VSP波动方程偏移在成像分辨率上的优势. 相似文献
20.
Application of a perfectly matched layer in seismic wavefield simulation with an irregular free surface 总被引：2，自引：0，他引：2 下载免费PDF全文
下载免费PDF全文 Haiqiang Lan Jingyi Chen Zhongjie Zhang Youshan Liu Jianguo Zhao Ruiqi Shi 《Geophysical Prospecting》2016,64(1):112-128
Recently, an effective and powerful approach for simulating seismic wave propagation in elastic media with an irregular free surface was proposed. However, in previous studies, researchers used the periodic condition and/or sponge boundary condition to attenuate artificial reflections at boundaries of a computational domain. As demonstrated in many literatures, either the periodic condition or sponge boundary condition is simple but much less effective than the well‐known perfectly matched layer boundary condition. In view of this, we intend to introduce a perfectly matched layer to simulate seismic wavefields in unbounded models with an irregular free surface. We first incorporate a perfectly matched layer into wave equations formulated in a frequency domain in Cartesian coordinates. We then transform them back into a time domain through inverse Fourier transformation. Afterwards, we use a boundary‐conforming grid and map a rectangular grid onto a curved one, which allows us to transform the equations and free surface boundary conditions from Cartesian coordinates to curvilinear coordinates. As numerical examples show, if free surface boundary conditions are imposed at the top border of a model, then it should also be incorporated into the perfectly matched layer imposed at the top‐left and top‐ right corners of a 2D model where the free surface boundary conditions and perfectly matched layer encounter; otherwise, reflections will occur at the intersections of the free surface and the perfectly matched layer, which is confirmed in this paper. So, by replacing normal second derivatives in wave equations in curvilinear coordinates with free surface boundary conditions, we successfully implement the free surface boundary conditions into the perfectly matched layer at the top‐left and top‐right corners of a 2D model at the surface. A number of numerical examples show that the perfectly matched layer constructed in this study is effective in simulating wave propagation in unbounded media and the algorithm for implementation of the perfectly matched layer and free surface boundary conditions is stable for long‐time wavefield simulation on models with an irregular free surface. 相似文献