Archive

25 November 2015, Volume 32 Issue 6 Previous Issue    Next Issue

---

For Selected:

Download Citations EndNote Ris BibTeX

Toggle Thumbnails

Select

Parallel Computing of First-principles Based Quantum Transport Simulations ZHANG Ruoxing, HOU Shimin, CHOU Qiang 2015, 32(6): 631-638.  Abstract ( )   PDF (3265KB) ( )   To solve long time-consuming problem in analysis of large-scale quantum transport systems based on first-principles calculations, we analyze hot spots of self-consistent iterations within the framework that combines non-equilibrium Green's function with density functional theory, namely DFT+NEGF method. Two parallel computing schemes based on MPI/OpenMP are proposed to deal with energy point integration and matrix inversion/multiplication. For energy point integral parallelism, sparse matrix as well as energy points should be assigned to each process over data initialization according to round-robin scheduling algorithm. Either MPI based ScaLAPACK subroutine or OpenMP based Intel MKL subroutine can be called to realize matrix inversion/multiplication parallelization. A sub-linear speedup ratio curve is obtained for energy point integral parallelism due to the fact that calculations related with different energy points are mutually independent. OpenMP parallelism adopts shared memory patterned data exchange mechanism and overhead of switching threads is rather small, and consequently it is better in computing efficiency but worse in code scalability than MPI implementation.

Select

Lattice Boltzmann Model for Compound Burgers-Korteweg-de Vries Equation DUAN Yali, CHEN Xianjin, KONG Linghua 2015, 32(6): 639-648.  Abstract ( )   PDF (1536KB) ( )   We develop a lattice Boltzmann model for compound Burgers-Korteweg-de Vries (cBKdV) equation. By properly treating dispersive term uxxx and applying Chapman-Enskog expansion, the governing equation is recovered correctly from lattice Boltzmann equation and local equilibrium distribution functions are obtained. Numerical experiments show that our results agree well with exact solutions and have better numerical accuracy compared with previous numerical results. This hence indicates that the model is satisfactory and efficient.

Select

A Finite Directional Difference Meshless Method for Diffusion Equations LV Guixia, SUN Shunkai 2015, 32(6): 649-661.  Abstract ( )   PDF (3557KB) ( )   An approach for numerically solving nonlinear diffusion equations on 2D scattered point distributions is developed with finite directional difference method. The approach yields stencils of minimal size using five neighboring points. And coefficients of discretization have explicit expressions. A scheme employing five-point formulae is proposed to discretize multimedia interface condition for discontinuous problems in which approximation to flux on interface is second-order accurate. The discretization methods show good performance in numerical examples with different computational domains and different point distributions.

Select

Simulation of High-Adiabat ICF Capsule Implosion GU Jianfa, DAI Zhensheng, YE Wenhua, GU Peijun, ZHENG Wudi 2015, 32(6): 662-668.  Abstract ( )   PDF (2512KB) ( )   High-foot high-adiabat implosion simulations of ICF ignition capsules are performed with one-dimensional multi-group radiation transport hydrodynamic code RDMG and two-dimensional few-group radiation diffusion hydrodynamic code LARED-S. Compared with low-adiabat implosion, high-foot implosions improve significantly stability of ablation front and ablator-fuel interface by increasing radiation drive temperature of the foot, leading to great reduction of hydrodynamic instability growth and hot-spot mix. Meanwhile, high-foot implosion decrease DT fuel compression. Final compression density and areal density of the main fuel at stagnation are decreased, causing lower neutron yield. A better stability of high-foot high-adiabat implosion is obtained at cost of reducing DT fuel compression.

Select

High-Performance Numerical Simulation of Jet in Cross-Flow Based on Lattice Boltzmann Method SHANGGUAN Yanqin, WANG Xian, LI Yueming 2015, 32(6): 669-676.  Abstract ( )   PDF (3305KB) ( )   Large eddy simulation (LES) with 1.12×108 meshes was performed on three-dimensional jet in cross-flow(JICF) using lattice Boltzmann method (LBM) and multiple Graphic Processing Units (Multi-GPUs). Reynolds number based on free-stream velocity and jet diameter is 4 000, streamwise inclination angle of jet is α=30°, and jet-to-cross-flow velocity ratio is set as 0.5. Validity and feasibility of LBM-LES on simulating jet in cross-flow were verified by reasonable qualitative and quantitative results. Fine coherent structures were captured by large-scaled simulation which benefits study on mixing mechanism of jet-in-cross-flow(JICF). Moreover, 6 K20M GPUs were adopted in simulation and it took 15 402 seconds for LBM-GPU solver to simulate 71 680 LBM steps resulting in calculated performance of 521. 24 MLUPS.

Select

Adaptive Moving Mesh Kinetic Scheme for Reactive Fluids ZHEN Yaxin, NI Guoxi 2015, 32(6): 677-684.  Abstract ( )   PDF (1719KB) ( )   We concern extension of gas-kinetic scheme of BGK type to reactive fluids, and develop an adaptive moving meshes BGK scheme (AMMBGK). We derive systems from a mass fraction BGK model for detonation fluids, including both inviscid and viscous reactive flow systems. Then, based on a BGK type scheme and splitting method that splits system into fluid part and part of energy released from reaction process, we present a mass fraction BGK scheme on moving meshes for reactive flows. Numerical results validate availability of the gas-kinetic scheme in simulation of reactive fluids.

Select

Characterization of Fracture Network by Volume Fracturing in Horizontal Wells and Coupled Flow Model JIA Pin, CHENG Linsong, HUANG Shijun, TIAN Xiaofeng 2015, 32(6): 685-692.  Abstract ( )   PDF (2743KB) ( )   For reservoirs with high rock brittleness coefficient and uniform development of natural fractures, three basic modes are developed to characterize fracture network created by volume fracturing in horizontal wells. Flow from reservoir to wellbore is divided to two parts:reservoir flow and network flow. Principle of potential superposition is used to derive reservoir flow equation. Finitedifference method is adopted to establish flow equation within finite conductivity network. Star-Delta transformation is used to tackle interplay of flow between hydraulic and natural fractures. A comprehensive flow model is presented by coupling matrix equations of two flows. It indicates that as stimulated horizontal length is kept constant, effect of quantity of fracture stages and perforation clusters in each stage on production is more significant than that of half-length and flow conductivity of hydraulic fracture, which is, in turn, more dominant than that of density and flow conductivity of natural fractures. Finally, a field example shows that little difference exists between measured and calculated results.

Select

An Improved Particle Swarm Optimization for Precocious Phenomenon in Nonlinear System Engineering XIAO Yuan, CUI Guomin, PENG Fuyu, ZHOU Jing 2015, 32(6): 693-700.  Abstract ( )   PDF (2426KB) ( )   By analyzing mechanism of premature phenomenon in particle swarm optimization (PSO), we found nature of premature convergence and proposed a "leap" strategy to jump out of local minimum, making halted particles "renewed" when they are trapped into a local optimum. The strategy is applied to nonlinear programming and results are encouraging. The improved PSO solves efficiently premature convergence of the algorithm applying in nonlinear optimizations and improves global search ability of PSO.

Select

CPML Absorbing Boundary Condition in Modeling of Transient Electromagnetic Fields JIANG Yannan, LIU Wen, WANG Jiao, ZHANG Wencui 2015, 32(6): 701-708.  Abstract ( )   PDF (2115KB) ( )   A scheme of convolutional perfectly matched layer (CPML) absorbing boundary condition (ABC) is proposed, which is used to truncate finite-difference time-domain (FDTD) method modeling transient electromagnetic (TEM) response. It derives specially CPML formulation dealing explicitly with divergence of magnetic induction and conceives an algorithm calculating convolutional term of z component. Then, the scheme is validated by numerical modeling of a homogeneous half-space model. The results indicate that efficient calculation of TEM is achieved.

Select

Classification of Epilepsy Based on Lempel-Ziv Complexity and EMD XIA Deling, MENG Qingfang, NIU Hegong, WEI Yingda, LIU Haihong 2015, 32(6): 709-714.  Abstract ( )   PDF (2316KB) ( )   Taking non-stationary and nonlinearity of epilepsy signals into consideration, we proposed a method for detection of epilepsy, based on Lempel-Ziv (LZ) complexity and empirical mode decomposition (EMD). EMD first decomposed epilepsy signals into a set of intrinsic mode functions (IMFs). Then calculated complexity of each IMF. Bonn dataset was utilized for evaluating the method. Experimental results showed that the highest accuracy could be achieved to 95. 25%. It has advantages of high accuracy, strong adaptability and so on.

Select

Monte-Carlo Modelling of Impurity Transport in EAST Tokamak ZHA Xuejun, ZHONG Dejun, WANG Fuqiong, CHEN Yiping, LU Hongwei, HU Liqun 2015, 32(6): 715-721.  Abstract ( )   PDF (1856KB) ( )   Basic idea and physical models in DIVIMP are introduced and DIVIMP modelling of C impurity production and transport properties in edge plasma of EAST machine with ohmic heating are provided. Simulation results show favorable consistency with experimental measurements, demonstrating that the models coupled in DIVIMP are suitable for interpretive and predictive modelling of impurity transport in EAST machine. Transport properties of W impurity in EAST with heating power as high as 8MW are also predicted by DIVIMP modelling.

Select

Diode Phenomenon of Acoustic Wave Propagation in a Two-Layer Nonlinear System GUO Fengli, HU Dongsheng, XU Jiang 2015, 32(6): 722-728.  Abstract ( )   PDF (2172KB) ( )   In one-dimensional nonlinear system, linear parts of acoustic wave meet wave equation and nonlinear part is assumed similar to that of nonlinear Schrödinger equation. Transmission coefficients of two incident acoustic waves from opposite directions are calculated respectively. It is shown that this model has unidirectional penetrability for acoustic wave within certain frequency and amplitude, and expresses behaviors of diode. Increasing nonlinear intensity leads to a remarkable enhancement of unidirectional penetrability.

Select

Thyristor Latching Current and Maintain Current TAN Wei, LI Jianqing 2015, 32(6): 729-734.  Abstract ( )   PDF (1318KB) ( )   Thyristor latching current and maintaining current have important effects on its working state. With thyristor conduction characteristics latching current is derived from carrier density; Based on current gain of NPN transistor, maintaining current is obtained. Compared with simulation results of Sentaurus TCAD software, calculated results have good agreement. It shows correctness of the formula.

Select

A Quadratic Memristor-based Fourth-order Chaotic Circuit YU Shicheng, ZENG Yicheng, LI Zhijun 2015, 32(6): 735-743.  Abstract ( )   PDF (5923KB) ( )   Compared with active piecewise linear(PWL) and cubic smooth memristor model adopted in most memristor-based chaotic circuits, charge-control quadratic memristor model is more practical, which is derived from a passive device fabricated by HP Labs. Based on the model and a topological dual structure of Chua's chaotic circuit, a memristor-based fourth-order chaotic circuit was designed. As is verified by theoretical analysis, simulation and circuit realizations, the charge-control quadratic memristor-based chaotic circuit has complex dynamical behaviors relying on initial state of memristor. Meanwhile, non-linear dynamical phenomenon of state transition is generated along with variation of initial state, time and system parameters. An eye of scroll and an eye of limit cycle are observed in phase diagrams of chaotic circuits

Select

Ground States of Rotating Bose-Einstein Condensates in an Annular Trap LIU Yan, ZHANG Suying 2015, 32(6): 744-750.  Abstract ( )   PDF (2998KB) ( )   Thomas-Fermi approximation (TFA) and imaginary-time propagation method are used to study ground states of rotating Bose-Einstein condensates in an annular trap. Ground state density profiles of condensates experience a transition from vortex lattice phase to giant vortex phase with increase of angular frequency or with increase of width and center height of trap potential. Particularly, ground state density profiles change from a disc shape into an annulus shape with the increase of width and center height of trap potential, when angular frequency is zero. Finally, comparison between ground state density profiles, obtained by analytical method and numerical method is made. They coincide with each other.

Select

Thermodynamic Properties of Fermi Gas Trapped in Both Gravity Field and Magnetic Field TIAN Qingsong, MEN Fudian, CHEN Xinlong 2015, 32(6): 751-756.  Abstract ( )   PDF (1767KB) ( )   With semi-classical approximation, thermodynamic properties of Fermi gas trapped in both gravity field and magnetic field are studied. By using theoretical analysis and numerical simulation, influence of gravity field on thermodynamic properties of the system in strong magnetic field is analyzed. It shows that, compared with the case of strong magnetic field only, gravity field makes the energy, chemical potential reduced. With rising temperature, influence of gravity field on chemical potential is gradually enlarged. There is a maximal influence of gravity field on heat capacity. Gravity field makes oscillation of heat capacity almost unchangeable while oscillation center of chemical potential shift down.