The method in [1] has been extended to the case of rotational flow in this paper. A new method for dealing with the shock wave is presented. This method has the advantages of both the shock-fitting and the shock captu...The method in [1] has been extended to the case of rotational flow in this paper. A new method for dealing with the shock wave is presented. This method has the advantages of both the shock-fitting and the shock capturing methods. The direct problem and the mixed direct-inverse prob- lem of the rotational flow in a transonic plane cascade at both design and off design conditions are solved, and the results show that the present method has rapid convergence rate and high accuracy even for the flow with moderately strong shocks. The calculations have been carried out on the DPS-8 computer, and for the direct problem, only 50-80 iterations are needed, and 50-80 seconds of CPU time are required.展开更多
Based on the working of Lighthill and Hunt et al., in the present paper the author has established the topological rules adapting to analysing the skin-friction lines and the section streamlines in cascades. These rul...Based on the working of Lighthill and Hunt et al., in the present paper the author has established the topological rules adapting to analysing the skin-friction lines and the section streamlines in cascades. These rules are (1) for a rotor cascade without shroud band, the total number of nodal points equals that the saddle points on the skin-friction line vector fields in eachpitch range; (2) for an annular or straight cascade with no-clearances at blade ends, the total number of saddle points is two more than that of nodal points on the skin-friction line fields in a pitch; (3) the total number of saddles in the secondary flow fields on cross-sections in cascade is one less than that of nodes; (4) in the section streamline vector fields on a meridian surface penetrating a flow passage, and on leading and trailing edge sections, the total number of nodes is equal to that of saddles; (5) on the streamline vector fields of a blade-to-blade surface, the total number of nodes is one less than that of saddles.展开更多
According to Taylor's expansion formula, a kind of derivatives computation method is presented, which can achieve arbitrary order of accuracy. Then it is combined with a flux difference splitting technique and a f...According to Taylor's expansion formula, a kind of derivatives computation method is presented, which can achieve arbitrary order of accuracy. Then it is combined with a flux difference splitting technique and a flux limiter to construct the desiredscheme that is suitable for non-oscillatory shock-capturing calculation. TVD typeRunge-Kutta method is used for temporal discretization. Several steady and unsteadynumerical experiments demonstrate that the scheme is a robust solver for transonicflows with high accuracy and high resolution of shock wave structures.展开更多
A new type fully reversible combined blade is presented, which can fully reverse airflow during the inverse ventilation by simply reversion. It is suitable for reversible axial flow fans used in tunnel and mine ventil...A new type fully reversible combined blade is presented, which can fully reverse airflow during the inverse ventilation by simply reversion. It is suitable for reversible axial flow fans used in tunnel and mine ventilation. The optimal parameters such as overlap ratio and pitch ratio are determined through cascade experiment. Experiment results also show a big promotion of maximum lift coefficient C1,max and stall attack angle a1,max compared to the existing bi-directional symmetry airfoils.展开更多
This paper presents an experimental study of the three-dimensional turbulent flow fields in a lowspeed annular cascade of high turning angle turbine blades. Detailed measurements were performed on the blade surfaces a...This paper presents an experimental study of the three-dimensional turbulent flow fields in a lowspeed annular cascade of high turning angle turbine blades. Detailed measurements were performed on the blade surfaces and mid-streamsurface in the passage and at three axial planes downstream of the cascade by using wall static pressure taps, a five-hole probe and a hot-wire anemometer. The test data include static pressure distribution on blade surfaces, total pressure loss coefficient, mean flow velocity components, radial flow angle, turbulence intensity and Reynolds shear stress. Analyses of the three-dimensional cascade flow characteristics were made on the onset location of high loss vortices, the variation of pressure gradient inside the cascade passage and the properties of endwall boundary layers,total pressure loss distributions, secondary vortex turbulent dissipation and wake decay downstream of the cascade. These experimental results are valuable for revealing the details of the complex vortex flow structure in modern highly loaded axial turbomachines and validating the three-dimensional flow numerical computation codes.展开更多
A class of Compact Finite Volume Schemes (CFVS) with small support stencils are developed based on a new reconstruction method for cell face variables. The accumulative errors of these schemes for a scalar wave equat...A class of Compact Finite Volume Schemes (CFVS) with small support stencils are developed based on a new reconstruction method for cell face variables. The accumulative errors of these schemes for a scalar wave equation are analyzed and compared. The established compact schemes are proved suitable for steady and unsteady flow simulations by several numerical experiments in this paper.展开更多
The tip-clearance flow in a cascade was numerically simulated by solving theRANS equations of incompressible fluids. The computational model was based upon the artificialcompressibility formulation proposed by Chorin....The tip-clearance flow in a cascade was numerically simulated by solving theRANS equations of incompressible fluids. The computational model was based upon the artificialcompressibility formulation proposed by Chorin. The Baldwin-Lomax turbulence model was used to makethe governing equations closed. For the specific structure of tip-clearance flow, a multi-block gridstructure was adopted to facilitate numerical compulations. The comparison of numerical resultswith experimental data indicates that the present method is capable of simulating tip-clearanceflows with satisfactory accuracy.展开更多
A high-order accurate explicit scheme is proposed for solving Euler/Reynolds-averaged Navier-Stokes equations for steady and unsteady flows, respectively. Baldwin-Lomax turbulence model is utilized to obtain the turbu...A high-order accurate explicit scheme is proposed for solving Euler/Reynolds-averaged Navier-Stokes equations for steady and unsteady flows, respectively. Baldwin-Lomax turbulence model is utilized to obtain the turbulent viscosity. For the explicit scheme, the Runge-Kutta time-stepping methods of third orders are used in time integration, and space discretization for the right-hand side (RHS) terms of semi-discrete equations is performed by third-order ENN scheme for inviscid terms and fourth-order compact difference for viscous terms. Numerical experiments suggest that the present scheme not only has a fairly rapid convergence rate, but also can generate a highly resolved approximation to numerical solution, even to unsteady problem.展开更多
A new type of third\|order upwind finite volume implicit scheme is proposed for solving two/three\|dimensional Euler/Reynolds\|averaged Navier\|Stokes equations for steady flow. The fundamental form of the implicit sc...A new type of third\|order upwind finite volume implicit scheme is proposed for solving two/three\|dimensional Euler/Reynolds\|averaged Navier\|Stokes equations for steady flow. The fundamental form of the implicit scheme is based on the LU\|TVD finite volume scheme with the hybrid flux splitting technique. The third\|order ENN scheme's numerical flux is used to calculate the inviscid terms of Navier\|Stokes equations.A fourth\|order accurate symmetric compact difference is applied to its viscous terms. The Baldwin\|Lomax turbulence model is used to calculate the turbulent viscosity. Numerical experiments suggest that the proposed scheme not only has a fairly rapid convergence rate, but also can generate a highly resolved approximation to the numerical solution.展开更多
文摘The method in [1] has been extended to the case of rotational flow in this paper. A new method for dealing with the shock wave is presented. This method has the advantages of both the shock-fitting and the shock capturing methods. The direct problem and the mixed direct-inverse prob- lem of the rotational flow in a transonic plane cascade at both design and off design conditions are solved, and the results show that the present method has rapid convergence rate and high accuracy even for the flow with moderately strong shocks. The calculations have been carried out on the DPS-8 computer, and for the direct problem, only 50-80 iterations are needed, and 50-80 seconds of CPU time are required.
文摘Based on the working of Lighthill and Hunt et al., in the present paper the author has established the topological rules adapting to analysing the skin-friction lines and the section streamlines in cascades. These rules are (1) for a rotor cascade without shroud band, the total number of nodal points equals that the saddle points on the skin-friction line vector fields in eachpitch range; (2) for an annular or straight cascade with no-clearances at blade ends, the total number of saddle points is two more than that of nodal points on the skin-friction line fields in a pitch; (3) the total number of saddles in the secondary flow fields on cross-sections in cascade is one less than that of nodes; (4) in the section streamline vector fields on a meridian surface penetrating a flow passage, and on leading and trailing edge sections, the total number of nodes is equal to that of saddles; (5) on the streamline vector fields of a blade-to-blade surface, the total number of nodes is one less than that of saddles.
文摘According to Taylor's expansion formula, a kind of derivatives computation method is presented, which can achieve arbitrary order of accuracy. Then it is combined with a flux difference splitting technique and a flux limiter to construct the desiredscheme that is suitable for non-oscillatory shock-capturing calculation. TVD typeRunge-Kutta method is used for temporal discretization. Several steady and unsteadynumerical experiments demonstrate that the scheme is a robust solver for transonicflows with high accuracy and high resolution of shock wave structures.
文摘A new type fully reversible combined blade is presented, which can fully reverse airflow during the inverse ventilation by simply reversion. It is suitable for reversible axial flow fans used in tunnel and mine ventilation. The optimal parameters such as overlap ratio and pitch ratio are determined through cascade experiment. Experiment results also show a big promotion of maximum lift coefficient C1,max and stall attack angle a1,max compared to the existing bi-directional symmetry airfoils.
文摘This paper presents an experimental study of the three-dimensional turbulent flow fields in a lowspeed annular cascade of high turning angle turbine blades. Detailed measurements were performed on the blade surfaces and mid-streamsurface in the passage and at three axial planes downstream of the cascade by using wall static pressure taps, a five-hole probe and a hot-wire anemometer. The test data include static pressure distribution on blade surfaces, total pressure loss coefficient, mean flow velocity components, radial flow angle, turbulence intensity and Reynolds shear stress. Analyses of the three-dimensional cascade flow characteristics were made on the onset location of high loss vortices, the variation of pressure gradient inside the cascade passage and the properties of endwall boundary layers,total pressure loss distributions, secondary vortex turbulent dissipation and wake decay downstream of the cascade. These experimental results are valuable for revealing the details of the complex vortex flow structure in modern highly loaded axial turbomachines and validating the three-dimensional flow numerical computation codes.
文摘A class of Compact Finite Volume Schemes (CFVS) with small support stencils are developed based on a new reconstruction method for cell face variables. The accumulative errors of these schemes for a scalar wave equation are analyzed and compared. The established compact schemes are proved suitable for steady and unsteady flow simulations by several numerical experiments in this paper.
文摘The tip-clearance flow in a cascade was numerically simulated by solving theRANS equations of incompressible fluids. The computational model was based upon the artificialcompressibility formulation proposed by Chorin. The Baldwin-Lomax turbulence model was used to makethe governing equations closed. For the specific structure of tip-clearance flow, a multi-block gridstructure was adopted to facilitate numerical compulations. The comparison of numerical resultswith experimental data indicates that the present method is capable of simulating tip-clearanceflows with satisfactory accuracy.
基金The project supported by the National Natural Science Foundation of China under Contract No.59576007 and 19572038
文摘A high-order accurate explicit scheme is proposed for solving Euler/Reynolds-averaged Navier-Stokes equations for steady and unsteady flows, respectively. Baldwin-Lomax turbulence model is utilized to obtain the turbulent viscosity. For the explicit scheme, the Runge-Kutta time-stepping methods of third orders are used in time integration, and space discretization for the right-hand side (RHS) terms of semi-discrete equations is performed by third-order ENN scheme for inviscid terms and fourth-order compact difference for viscous terms. Numerical experiments suggest that the present scheme not only has a fairly rapid convergence rate, but also can generate a highly resolved approximation to numerical solution, even to unsteady problem.
文摘A new type of third\|order upwind finite volume implicit scheme is proposed for solving two/three\|dimensional Euler/Reynolds\|averaged Navier\|Stokes equations for steady flow. The fundamental form of the implicit scheme is based on the LU\|TVD finite volume scheme with the hybrid flux splitting technique. The third\|order ENN scheme's numerical flux is used to calculate the inviscid terms of Navier\|Stokes equations.A fourth\|order accurate symmetric compact difference is applied to its viscous terms. The Baldwin\|Lomax turbulence model is used to calculate the turbulent viscosity. Numerical experiments suggest that the proposed scheme not only has a fairly rapid convergence rate, but also can generate a highly resolved approximation to the numerical solution.
