Archive

25 March 2003, Volume 20 Issue 2 Previous Issue    Next Issue

---

For Selected:

Download Citations EndNote Ris BibTeX

Toggle Thumbnails

Select

High Resolution Eulerian Adaptive Algorithm for Multi-fluid Interfaces BAI Jin-song, CHEN Sen-hua, LI Ping 2003, 20(2): 95-101.  Abstract ( )   PDF (418KB) ( )   Level Set (LS) function for tracking multi-fluid interfaces and the different materials corresponding to different leves are applied.These LS equations together with fluid dynamic Euler equations are solved independently.An adaptive mesh refinement (AMR) method in the context of Cartesian grids is discussed at the same time.Such an approach can be automatically adaptive in that it chooses the regions of refinement based on the behavior of the solution while for time dependent problems the regions of refinement must change with time.There are two main strategies that have been used to handle the data structures involved when a rectangular grid is being refined.One approach is to refine individual grid cells as needed,typically by splitting a single cell in two dimensions into four pieces.Each of these pieces may be further subdivided recursively,depending on how many levels of refinement are allowed.High resolution results can be given by using this method,and it can also track the shock front adaptively and save CPU greatly.

Select

Domain Decomposition Method Utilizing Chimera Grids in Conjunction with a Second Order Difference Scheme Based on an Entropy Condition DONG Hai-tao, ZHANG Li-dong, LEE Chun-Hian 2003, 20(2): 102-106.  Abstract ( )   PDF (252KB) ( )   A domain decomposition method based on chimera grids is proposed for Euler solver. The solver is developed based on a class of second order TVD difference schemes with a new entropy condition. The developed procedure is capable of handling the steady and unsteady flows over multibodies located separately, or in contact, or under relative motion with each other. The procedure is demonstrated by a test case of supersonic flows over two circular cylinders parted by different distances. A detail result is presented. The procedure is then applied to simulate the rotating flow fields over rockets with canard rudders.

Select

Incremental Harmonic Balance Computation Scheme for a Class of Piecewise-linear Circuit Systems XU Lei, LU Ming-wan, CAO Qing-jie 2003, 20(2): 107-110.  Abstract ( )   PDF (127KB) ( )   IHB computation scheme of the periodic solutions for a class of piecewise-linear circuit systems is derived and the relevant numerical simulation is carried out. Phase planes under some specific control parameter values, as well as the system response property within a band of control parameter values, are plotted with the results compared with those of the Runge-Kutta method. It is shown that the IHB computation scheme proves to be an effective way of analyzing nonlinear problems in engineering practice.

Select

Collar Grid and Virtual Grid Methods in Embedding Technique LI Xiao-wei, FAN Xu-ji, QIAO Zhi-de 2003, 20(2): 111-118.  Abstract ( )   PDF (444KB) ( )   The Collar grid method and the virtual grid method are jointly used in the embedding technique to solve the problem of finding the interpolating cells among the internal and external boundary points near the joint regions. With different boundary planes generated along different fixed surfaces, the Collar grid obtained using the hyperbolic partial differential equations can ensure to generate high-quality grids and to provide real interpolating cells for the boundary points on the joint regions. The main purpose of the virtual grid is to convert a solid wall boundary condition into an interface condition while no fluid flow computations are conducted within the virtual grid. The computational results of several configurations show that the currently developed embedding technique with Collar grid and virtual grid can effectively treat the geometrical configuration and can more accurately predict the flow over complex configurations with intersecting surfaces.

Select

Approximate Energies of the 1s2p-state of Helium by the Finite Element Method ZHENG Wei-ying, YING Lung-an 2003, 20(2): 119-122.  Abstract ( )   PDF (212KB) ( )   The finite element method is used to obtain numerical solutions of the Schrödinger equation for the Helium 1s2p-state. The relative errors of approximate energies are 10-6 for triplet and 10-4 for singlet which are slightly smaller than J. Shertzer's results for the Helium ground state[9]. The generalized eigenvalue problems obtained by FEM are symmetric for the ground state but unsymmetric for 1s2p-state. It becomes more difficult to solve the problems. Form the graphs of wave functions, it can be seen that it is reasonable to solve the Schrödinger equations in bounded domains.

Select

Magnetic Dipole Transitions of the K-like Isoelectronic Seguence YI You-gen, ZHENG Zhi-jian, TANG Yong-jian 2003, 20(2): 123-126.  Abstract ( )   PDF (187KB) ( )   A fully reIativistic multiconfiguration Dirac-Fock method with Breit and QED corrections is used to caIcuIate the magnetic dipoIe 3p63d2D5/2-2D3/2（Z=22~42）transition energy level separations,transition probabilities and oscillator strengths for the K-like ions. In calculation,significant core polarization effects,Breit and OED corrections are considered. The results are in agreement with recent experiments. A substantial part discrepancy disappears in the 2D5/2-2D3/2 separations when core excitation channels are taken into account. There are no transition probability and oscillator strength resuIts in the isoelectronic seguence until now.

Select

Symplectic Integrators in Quantum Systems with Time-dependent External Fields LIU Xiao-Yan, LIU Xue-shen, DING Pei-zhu 2003, 20(2): 127-129.  Abstract ( )   PDF (147KB) ( )   In a quantum system,when the Hamiltonian operator is time-dependent,"artificial" variables are introduced to construct the symplectic integrators with arbitrary high order accuracy.As an example,the time-evolution of an electron in the infinite deep potential well interacting with an animated laser field is investigated.The computed results coincide with the theory and can preserve the norm,which show that the methods are reasonable.

Select

Finite Element Analysis on Traffic Flow Problem(Ⅲ) ZHANG Peng, LIU Ru-xun 2003, 20(2): 130-136.  Abstract ( )   PDF (307KB) ( )   Two influential traffic flow models formulated by Zhang H M and Wu Zheng are improved to include the lane-change problem,and the relations between the two models and the traditional speed-density hypothesis model are discussed.Runge-Kutta Discontinuous Galerkin Finite Element Method is generalized and applied to these three models for numerical solutions through which the traffic signal problem,lane-change problem and non-equilibrium traffic flow phenomena are well simulated.

Select

Surface Segregation of Binary Alloys with Monte Carlo Simulation DENG Hui-qiu, HU Wang-yu, SHU Xiao-lin 2003, 20(2): 137-141.  Abstract ( )   PDF (210KB) ( )   The Monte Carlo method, which can be used to simulate the surface segregation of binary alloys, is discussed. Using our analytic modified EAM potentials, the surface concentrations and concentration depth profiles of Pd-Au binary alloys are studied with Monte Carlo simulation based on the Grand Canonical Ensemble Statistical Rule. Simulation results show that the topmost surface is enriched with Au, and a damped oscillation of Au concentration is found in the whole composition range. The results are in agreement with the available experiment data and other theoretical values.

Select

Analysis of Two-phase Continuous Porous Media with FEM-EFGM Coupled Method ZHANG Yan-jun, WANG Si-jing, WANG En-zhi 2003, 20(2): 142-146.  Abstract ( )   PDF (232KB) ( )   In this paper, Finite element method (FEM) coupled with Element free Galerkin method (EFGM), is developed it, not only takes advantage of FEM, but also gives EFGM local technique and element-free character. Both make it has more advantages in calculating. The paper introduces EFGM in details, and meantime dedues discrete equations of consolidation. It gives two examples of calculating and analyzing, the results of calculating point out that the implementation will achieve more accurate in the deformation of consolidation, another discovery is that an accurate solution of fluid flux can be obtained. In particularly, combining with FEM the results is more effective, in conclusion, EFGM is powerful complementary of FEM.

Select

Calculation of Mixed Cells in Simulating High-speed Collision Problems with Two-dimensional Eulerian Hydrodynamic Method FENG Qi-jing, HE Chang-jiang, ZHANG Min, YU Zhi-lu 2003, 20(2): 147-152.  Abstract ( )   PDF (264KB) ( )   The algorithm proposed here can be used in 2D two-step plane and axisymmetric Eulerian hydrodynamic method to deal with the interface of a mixed cell. It takes advantage of the idea of Youngs method: in a mixed cell, the interface is regarded as a straight line. The area fraction of eight cells surrounding the mixed cell is used to determine the normal orientation of the interface, the area fraction or the volume fraction of the mixed cell is used to determine the equation (position) of the interface, and then the flux passing across the boundary of the cell according to this interface and the area fraction or the volume fraction of the cell at next time is calculated. In the last part of this report, some numerical results are presented, which include the precision tests of interface calculation, the comparison with SLIC and the calculations of comprehensive problems.

Select

SPH Simulation of One-dimensional Shock Tube Problems XU Li, SUN Jin-shan 2003, 20(2): 153-156.  Abstract ( )   PDF (179KB) ( )   Smoothed particle hydrodynamics(SPH) is a free-Lagrangian computational method.Simulating shock tube problems is a common measure to test SPH methods and SPH codes.The fundamentals of SPH are presented and the implementation of SPH for 1-D shock tube problem simulation is discussed.Some numerical results and explanations are provided.

Select

Lattice Boltzmann Simulations of the Klinkenberg Effect in Porous Media LIU Yue-wu, ZHOU Fu-xin, YAN Guang-wu 2003, 20(2): 157-160.  Abstract ( )   PDF (128KB) ( )   The numerical simulation of lattice Boltzmann method (LBM) is one of the most efficient methods to investigate the complex porous media structure, particularly the Klinkenberg effect. It is very useful to deal with the related complex boundary problems.The problems of gas flow through porous media are studied by using the lattice Boltzmann methods. Comparison between the numerical simulation results and the experimental results is carried out. It is shown that the lattice Boltzmamn method is one of the most efficient methods to simulate the problems of gas flow through complex porous media.

Select

Forward Modeling of Induction Well Logging Tools in Dipping Boreholes and Their Response CHIMEDSURONG Z, WANG Hong-nian 2003, 20(2): 161-168.  Abstract ( )   PDF (387KB) ( )   The transmitters of induction well logging tools can be simplified as a series of magnetic dipoles, whose fields formed in dipping boreholes can be decomposed as the sum of transversal electrical and transversal magnetic waves (TE and TM waves), and solved completely from the two scalar partial differential equations. A new process is first given to decompose the electromagnetic field into TE and TM waves through the tensor expression of Maxwell equations in the cylinder system, and then recurrence formula for amplitudes of the field in each bed and generalized reflection and transmission coefficients on each boundary are given. On the bases, the analytic solution of the field in each bed is obtained. Finally, using numerical results, the influence of dipping angles of boreholes, bed thickness and thinly laminated sand/shale on the response of the deep induction well logging tools is investigated.

Select

Electromagnetic Scattering Calculation Based on Iterative QL-Born Approximation ZHAO Lian-feng, CAO Jun-xin, PAN Xian-jun, TANG Jian-guang 2003, 20(2): 169-172.  Abstract ( )   PDF (184KB) ( )   Integral equation method is one of effective methods of numerical simulation and inversion calculation for multi-dimensional electromagnetic response.Based on the linearization of the scattering electromagnetic integral equation, conductance imaging using iterative Born approximation is effective.Aiming at the characteristics that iterative Born approximation inversion depends on the initial model and that QL approximation doesn't need initial value, the two step method,iterative QL-Born approximation method,is put forward,which takes the result of QL inversion as the initial model of the iterative Born approximation method and avoids contrived initialization.The iterative QL-Born approximation method is effective and testified by numerical experiments

Select

Effects of the HOB and the Burst Yield on the Properties of NEMP MENG Cui, CHEN Yu-sheng, ZHOU Hui 2003, 20(2): 173-177.  Abstract ( )   PDF (219KB) ( )   Finite difference method is used to simulate the physical process numerically during which the electromagnetic signal is generated by the interaction of nuclear-explosion-induced currents with the geomagnetic field. The EMP waveforms above burst point are researched. The earth-based coordinates and the local spherical coordinates are introduced to illustrate propagating properties of the nuclear electromagnetic pulse along the orbit direction of satellites with various explosion height and yield. Taking into account the attenuation and dispersion of the ionospheric plasma, the FFT method is used to extrapolate the strength of the DEMP. With cosmic noise considered, the peak electric field strength with the yield over 1 kt could be detected at a distance of 104km.

Select

Three Dimensional Fluid Flow and Heat Transfer Based on Parallelepiped Elements TONG Zhong-xiang, ZHANG Jian-bang, QIANG Xiao-yi, WANG Xu 2003, 20(2): 178-182.  Abstract ( )   PDF (232KB) ( )   Because the number of tetrahedral elements in three dimensional field division is too great and will need too much computer memory and run time,the control volume finite element method based on parallelepiped elements is presented. In this method,the number of elements is much smaller and the number of points is egual to that in using tetrahedral elements. The divergence theorem can be used to translate the volume integral into the surface integral corresponding to the same control volume of grid points so that the upwind interpolated scheme for convection variables can be used. Through three test problems,it is proved that this method is valid for numerical simulation of three dimensional fluid flow and heat transfer under the limit of computer memory and run speed.

Select

Numerical Methods for One-dimensional Compressible Multi-fluid Flow MA Dong-jun, SUN De-jun, YIN Xie-yuan 2003, 20(2): 183-188.  Abstract ( )   PDF (284KB) ( )   A high-resolution interface tracking method is applied to simulate compressible multi-fluid flows with general equation of state in one-dimension. It captures the interface by using Level Set technique with conservative numerical discretization at interface. It uses high-order PPM scheme and two-shock approximation for GEOS Riemann problem. Results show that the computation is high-order accurate and there are no numerical oscillation and smearing at multi-fluid interface.