EI、Scopus 收录

2006 Vol. 38, No. 6

Display Method:
Numerical simulation of windward vortex shocks about supersonic slender
Dawei Chen, Xiaoliang Ma, Guowei Yang
Vortex shocks may be found inside the bow shock wave onsupersonic slender bodies at a high angles of attack. A numerical studyis performed by using Reynolds-averaged Navier-Stokes equations ,k\omega-sst and Badwin-Lomax turbulence models. It is shown that thevortex shocks are associated with the virtual double cone-like deflection of thesupersonic stream by the primary vortices.
2006, 38(6): 721-732. doi: 10.6052/0459-1879-2006-6-2005-251
Solving 2-D buoyancy-driven cavity flow on collocated meshes by perturbational finite volume scheme
Minguo Dai, Zhi Gao
The buoyancy-driven cavity flow in the case of$Ra=10^{3}$ to $Ra=10^{8}$ is numerically simulated with the upwind typeof perturbation finite volume (UPFV) scheme for the convective and diffusiveequations initiated by the second author. The UPFV scheme for diffusive andconvective equations takes the same concise expressions and uses thesame nodes as the upwind scheme of the first order and high approximateaccuracy of reconstruction can be achieved. Especially, theUPFV scheme is identical for uniform grid and non-uniform grid regardlessof the distance from the CV center to the common face. The results of numericalsimulations to buoyancy-driven cavity flows by UPFV are consistent withthese by spectral method and finite element method and DSC in the range from $Ra=10^{3}$ to$Ra=10^{8}$ with uniform or non-uniform structured grid. It shows that thePFV scheme is valid and can be applied for the heat convection and transferproblems.
2006, 38(6): 733-740. doi: 10.6052/0459-1879-2006-6-2005-072
Effect of blowing and suction on wall turbulent coherent structures
Zhenyan Xia, Nan Jiang, Zhendong Wang, Wei Shu
The streamwise velocity component in the wall turbulentboundary layer disturbed by blowing and suction was measured with hotwire anemometer. FFT transform and wavelet transform were adopted tostudy the effect on the turbulent energy spectrum. The results showthat the disturbance in low frequency makes the turbulent kinetic energy betransferred newly among different turbulent structures. One of theimpact is that the energy of large scale turbulent structures in innerboundary layer obviously decreases, especially in wavelet scale $j=-10$, whichcorresponds to coherent structures. The other effect shows that the energyof small scale turbulent structures increases synchronously near the wall. When themeasured position is leaving from the boundary layer, not only the energyof small structures keeps growing up but also the trend of increased energygradually extends to larger scale structures until the change adds up to themaximum in the borderline between inner and outer ($y_{v}=0.1$) boundarylayers. According to the above-mentioned facts, one may assumethat the large scale turbulent structures are broken up into many smallscale structures arising from the effect of low frequent blowing andsuction. Note that above the boundary layer at $y_{v}=0.15$(the outerboundary layer), the turbulent spectrum graphs with and without disturbanceare identical superposition, which reveals that the influence of blowing andsuction far from the wall boundary layer decays gradually until it disappearsin the outer boundary layer.Furthermore, by means of VITA method and wavelet transform to identify burstevents, it is clear that the disturbance not only reduces the burst intensityand delays the averaged burst period but also diminishes the scope ofconditionally averaged velocity signals and shortens the duration time,which indicates that the disturbance apparently restrains the burst processof coherent structures. Based upon the above described analysis, thedisturbance of blowing and suction is perfectly suited to controll theturbulent flow.In summary, the two methods, wavelet transform and FFT transform, canboth be regarded as perfect tools to analyze the turbulent energy spectrum.Compared with FFT, wavelet transform makes the energy spectrum graph muchsmooth due to its filter function. Another advantage is that itdecomposes theturbulent flow into different frequencies (or scales) which represent multiscales of turbulent structures including coherent structures and dissipativestructures in order to recognize their respective characteristics. Withfree parameters, wavelet transform is not only anobjective and effective utility to identify coherent structures and burstevents but also an important tool to validate and supplement theanalysis of the wall turbulent flow with VITA method.
2006, 38(6): 741-748. doi: 10.6052/0459-1879-2006-6-2005-109
Simulation of the motion of three-dimensional underwater explosion bubble using simple green function method
Xiongliang Yao, Aman Zhang
In this paper, it is assumed that the flow field isirrotational and incompressible at the underwater explosion pulsation phase. Thepotential theory is applied to deduce the governing equation of bubbleboundary motion. The underwater explosion bubble motion is studied byusing the boundary integral method with high order curve triangular-elements tomodel the three-dimensional bubble surface. The computational resultsare comparedwith Rayleigh-Plesset bubble model and experimental data, which showthat the high order surface element can precisely simulate the motion ofunderwater explosion bubble and is superior to linear elements.The gravity, non-gravity and rigid wall have an effect upon the bubblemotion. The paper has simulated these cases, and predicted the dynamicbehavior of underwater explosion bubble in a flow field, such as expansion,collapse, jetting, and also simulated kinetic characteristics ofthree-dimensional bubble around cylinder by establishing a three-dimensionalmodel, which reflects interactions between cylinder, free surface and theunderwater explosion bubble motion.
2006, 38(6): 749-759. doi: 10.6052/0459-1879-2006-6-2005-598
A wave model based on the boussinesq equations
Xiaozhou Ma, Guohai Dong, Bin Teng
An alternative method to derive a set of fully nonlinearBoussinesq equations up to the order of $O(\mu^{2}$, $\varepsilon^3\mu^{2})$ ispresented. The linear dispersion relation and the shoaling gradient of theequations are improved by adding some dispersive terms. The lineardispersion relation of the enhanced equations is the Pad\'{e} [4,4]expansion of the linear Stokes dispersion relation, the accuracy of which isacceptable even when the relative water depth is as large as 1.0. Itsnonlinear property and shoaling gradient are also improved. Thehorizontal one-dimensional equations are solved with a predictor-correctorfinite difference scheme and a fully nonlinear Boussinesq wave model isestablished, which enjoys high computational efficiency and reliability. Thenumerical model is verified by simulating the transformation of wavespropagating over a submerged bar. The numerical results are verified againstthe laboratory experimental data, and their agreement is found to be very good.
2006, 38(6): 760-766. doi: 10.6052/0459-1879-2006-6-2005-144
Study on nonlinear wave motions and wave forces on ship sections against vertical quay in a harbor
Daguo Wang, Zhili Zou, Chun'an Tang
A two-dimensional time-domain coupled numerical model isdeveloped to calculate the nonlinear wave forces acting on a ship sectionagainst vertical quay in a harbor. The fluid domain is divided into an innerdomain and an outer domain. The outer domain is the area between the leftside of the ship section and the incident boundary, where the flow canbe expressed byBoussinesq equations. The rest area is the inner domain, which is the domainbeneath the ship section plus the domain between the right side of the shipsection and the vertical quay wall. The flow in the inner domain can be expressed byNewton's Second Law. Matching conditions on the interface between the innerdomain and the outer domain are the continuation of volume flux and theequality of pressures. The numerical results of the coupled model agree wellwith the experimental data for wave elevations and wave forces. The naturalfrequency of the fluid motions in the gap between a ship section, seabed andvertical quay wall is derived. It is shown from the experimental data andnumerical results that the resonance waves in the gap are induced by the first orhigher harmonics of incident waves, and the first or higher order horizontalwave forces on the ship section against vertical quay wall increasegreatly whenthe frequency of the harmonics of incident waves is close to the naturalfrequency of fluid motions in the gap.
2006, 38(6): 767-775. doi: 10.6052/0459-1879-2006-6-2005-418
Sub-grid scale combustion models for two-phase combustion flow in an annular combustor
Yingwen Yan, Jianxing Zhao, Jingyong Zhang, Yong Liu
An Eulerian-Lagrangian two phase large-eddy simulation (LES)is described for calculating the instantaneous gas-liquid two-phase turbulentreacting flow in an annular combustor by using different sub-grid scalecombustion models. The k-equation sub-grid scale model is used to simulatethe sub-grid eddy viscosity. The six-heat flux model is applied topredictions of the radiant heat transfer. The EBU combustion sub-grid scalemodel(E-A model), the turbulent combustion second-order moment algebraicmodel(SOM-A Model) and the second-order moment transport model (SOMmodel) are,respectively, employed to determine the chemical reaction rate.Three-dimensional body fitted grid is generated by an elliptic gridgeneration procedure. The gas phase is solved with SIMPLE algorithm andhybrid scheme in the non-staggered grid system. The liquid phase equations aresolved in a Lagrangian frame of reference by PSIC algorithm. From simulationresults, the exchange of mass, momentum and energy between gas and fueldroplet can be obtained. Experimental and predicted results are shown to bein qualitative agreement, and the comparison shows that the large-eddysimulation method in an arbitrary curvilinear coordinates may be used toinvestigate two-phase reacting flows and the combustion process by usingdifferent sub-grid scale combustion models. SOM model is better than E-AModel and SOM-A Model in the simulations.
2006, 38(6): 776-784. doi: 10.6052/0459-1879-2006-6-2005-040
On the numerical simulation of perturbation's propagation and development in traffic flow
Zheng Wu
The propagation and development of perturbations in variousdensity traffic flows are numerically simulated. The studies include thecharacteristics of some traffic flow models such as PW model, 1-d pipe fluidmodel and the unilateral propagation model, the effects of equilibriumfunctions in the models, and the influences of difference schemes on thenumerical results. It is found that we can obtain results quite close to the measured data by using the appropriate traffic model, equilibrium function anddifference scheme for the traffic perturbation problems, especially thepropagation speed of the perturbation waves. A group of calculation schemessuitable to various density traffic flows are proposed.
2006, 38(6): 785-791. doi: 10.6052/0459-1879-2006-6-2005-392
Bifurcation creep buckling of viscoelastic laminated circular cylindrical shells
Fan Peng, Hong Xiang, Yiming Fu
The prebuckling creep deformation and bifurcation creepbuckling for viscoelastic symmetrically laminated circular cylindricalshells subjected to axial compression are analyzed. Boltzmann hereditaryconstitutive relationship is used to model the viscoelastic behavior oflamina and the instantaneous elastic response is assumed whenbifurcation occurs. The governing equations of prebuckling creep deformationand bifurcation buckling are based on DMV shallow shell theory and the K\'{a}rm\'{a}n-Donnell geometrically nonlinear relationship. Thenumerical procedure and analytical deduction based on viscoelasticapproachand quasi-elastic approach, respectively, are performed to investigate thebehavior of prebuckling creep deformation. A semi-analytical solutionstrategy is adopted to obtain the critical time of bifurcation buckling.Particular attention is placed on the analysis of glass/epoxy matrixcomposite circular cylindrical shells, results of prebuckling creepdeformation show that the time-dependent evolution of flexural deformationand circumferential membrane force is strongly dependent on the lay-upconfiguration of cylindrical shells. Both of them decrease in the case ofcross-ply conditions and increase for symmetric angle lay-up mode withtime. Theboundary condition at two ends of short laminated cylindrical shells withsymmetric angle lay-up plays a key role on the stability and the bifurcationcreep buckling may not be observed for simply supported ends.The increase of the ratio of radius to thickness will deteriorate the longterm stability of laminated cylindrical shells with constrained ends.
2006, 38(6): 792-798. doi: 10.6052/0459-1879-2006-6-2005-534
Perturbed magnetic fields generated by deformation of structures in earth magnetic field
Fei QinDongmei Yan, Xiaofeng Zhang
Development of magnetism based non-destructive test (NDT) technologiesrequires quantitative relations between deformation, stress,damage or fracture of structures and their induced magnetic field which canbe measured as indicators to assess integrity of the structures. Based onlinearized magnetoelastic theory, the perturbed magnetic fieldsinduced by mechanical stress and deformation in steel or otherferromagnetic structures working in the environment of the Earth magneticfield, are investigated theoretically in this paper.Governing equations and boundary conditions to determine the perturbedfields are derived. The effect of mechanicaldeformation on the magnetic fields is taken into account by couplingstructural displacement into continity conditions of the perturbed magneticfield on the boundary of the structure. When the applied magnetic field isweak, such as the Earth's magnetic field, the effect of magnetic fields onstructural deformation can be neglected. This greatly simplifies thecoupling equations of magnetomechanical interactions. It shows that thenormal projection of displacement gradient on the structure boundary plays adominating role in the perturbed magnetic fields.As an example, the perturbed field of a half-plane magnetized structurecaused by a point force is calculated by the Fourier transform method. Thecalculated magnetic intensity component normal to the boundary of thestructure assumes a symmetric distribution about the point where theforces is acted andreaches its maximum at that point while the component tangentto the boundary is asymmetric and inverses its direction sharply at thatpoint. The magnetic flux density of the perturbed fields isproportional to the magnitude of the applied force. These features provide apossible way for NDT technologies.
2006, 38(6): 799-806. doi: 10.6052/0459-1879-2006-6-2005-246
On structural optimization for non-probabilistic reliability based on convex models
Zhan Kang, Yangjun Luo
This paper discusses the setting and numerical solution ofthe non-probabilistic reliability-based structural optimization problem.Based on the convex model description for parameter scatters, the optimaldesign of non-deterministic structures is formulated as a nestedoptimization problem, in which the inner loop concerns a Min-max problem forevaluation of the reliability index. A performance measure-based method isproposed, where the feasibility of a design is determined throughminimization of the performance function value within the parameter domain.The expensive computation of the non-probabilistic reliability index, asneeded in the conventional approach, is thus avoided. The proposed approachis applicable to structural optimization problems accounting for deviationsof material properties, geometrical dimensions and loading conditions.Moreover, the design sensitivity of the minimum performance function valuecan be readily evaluated. Numerical examples show the validity andefficiency of the present method.
2006, 38(6): 807-815. doi: 10.6052/0459-1879-2006-6-2005-629
Mechanism of neurodynamics on attention and memory
Rubin Wang, Zhikang Zhang, Jing Yu
In this paper we propose a new nonlinear stochastic dynamicevolution model for phase encoding in population of neuronal oscillatorswith different phases, and study the neural informationprocessing in cerebral cortex and dynamic evolution under the action ofdifferent stimulation signals. The evolution of the averagednumber density along with time in the space of three dimensions is described indifferent clusters of neuronal oscillator firing action potential atdifferent phase spaces by means of numerical analysis. The resultsof numerical analysis show that the dynamic model proposed in this paper canbe used to describe the mechanism of neurodynamics of attention andmemory, and it is proved that only the neural dynamic model in ahigh-dimension space can adequatelydescribe dynamic characteristics of the neural population, and muchuseful neural information may be lost in the early models of stochasticdynamics for phase coding.
2006, 38(6): 816-824. doi: 10.6052/0459-1879-2006-6-2005-600
On the mechanism of the interaction between flexible panels and sound waves in flow ducts
Xiaoyu Wang, Xiaofeng Sun
The control of sound propagation in ducts has always been a problem ofconcern. One of new attempts is to make use of the interaction between asound wave and a flexible panel to reduce the sound propagation. This paperpresents a general model to describe the interaction between flexiblepanelsand sound waves with the emphasis on the effect of flow Mach number, higherorder modes and annular or circular geometrical parameters on the control ofsound propagation, which is not much studied in the current literature. Based ona comparison with the existing results, some important parametric studyis made in the present paper. Further research work issuggested after a detailed discussion and analysis on thenumerical results.
2006, 38(6): 825-830. doi: 10.6052/0459-1879-2006-6-2005-290
Molecular dynamics simulations of nanocrystalline ta during uniaxial tensile deformation process
Zhiliang Pan, Yulong Li
Using molecular dynamics simulations, the microstructureevolution of nanocrystalline Ta under uniaxial tensile loading isinvestigated. The results show that stress-induced phase transitions fromBCC to FCC and HCP structures may take place during the deformation process. Themaximum of the fraction of FCC atoms varies with the tensile strengthlinearly, by which we can find a critical stress when the phase transitiontakes place. The higher the strain rate is, the later the phase transitionoccurs. Intergranular fracture is also found during the deformationprocess. Crack-formation strain is influenced not by the mean grain size, butalso by the strain rate and the simulation temperature.
2006, 38(6): 831-834. doi: 10.6052/0459-1879-2006-6-2005-607
Theoretical solutions of stress distribution in casing-cement and stratum system
Youquan Yin, Zhaowei Chen, Ping'en Li
The theoretical solutions of a casing-cement-stratum system in theformation stress field are studied using the elasticity theory. The problem isdivided into two relatively simple subproblems.The cased hole system is a cylindrical layer structure and theloading state of each layer is similar. The solutions of two planestrain basic problems can be used to obtain the analytic solutions oftwo subproblems through the continuity condition of displacement. The theoretical solutionsfor the stress distribution in casing--cement and stratum system arefinally obtained by applying the superposition principle.
2006, 38(6): 835-842. doi: 10.6052/0459-1879-2006-6-2005-625
Improved genetic algorithm for shape optimization of truss structures
Wenyan Tang, Qingke Yuan
This paper presents an improved genetic algorithm (GA) tominimize the weight of a truss with discrete sizing, continuous shape variables.Because of the nature of discrete and continuous variables, mixed codingschemes are proposed, including binary and float coding, integer and floatcoding. Surrogate reproduction is developed to select good individuals tomating pool on the basis of constraint and fitness values, taking fullaccount of the character of the constrained optimization. This paper proposes anew strategy of creating next population by competing between parent andoffspring populations based on constraint and fitness values; so that thelifetime of the excellent gene is prolonged. Standard examples are solved,numerical solutions are shown to be better than those in the literature.
2006, 38(6): 843-849. doi: 10.6052/0459-1879-2006-6-2006-056
Study on dynamic responses of high-speed maglev vehicle/guideway coupling system under random irregularity
Jin Shi, Qingchao Wei, Chuanfeng Wan, Yashi Deng
In this paper, the maglev vehicle/guideway dynamic model isestablished based on the EMS maglev vehicle and the elevated-guidewaytechnology, the PSD model for guideway irregularities is discussed, and theseparated iterative procedure and its corresponding numerical algorithm areapplied in simulations. The effect of random irregularity on dynamiccharacteristics of the maglev system is studied numerically, and the randomresponses of the vehicle and the guideway are analyzed. Finally therandom vibrationcharacteristics of the vehicle/guideway system are obtained. It is onthis basis that control suggestions are proposed for guideway irregularityin maglev.
2006, 38(6): 850-857. doi: 10.6052/0459-1879-2006-6-2005-493