CN113821885A - Aero-engine blade flow field computing platform based on grid automatic generation technology - Google Patents
Aero-engine blade flow field computing platform based on grid automatic generation technology Download PDFInfo
- Publication number
- CN113821885A CN113821885A CN202111036994.9A CN202111036994A CN113821885A CN 113821885 A CN113821885 A CN 113821885A CN 202111036994 A CN202111036994 A CN 202111036994A CN 113821885 A CN113821885 A CN 113821885A
- Authority
- CN
- China
- Prior art keywords
- interpolation
- flow field
- grid
- equal
- mapping
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
- G06F30/28—Design optimisation, verification or simulation using fluid dynamics, e.g. using Navier-Stokes equations or computational fluid dynamics [CFD]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/10—Geometric CAD
- G06F30/17—Mechanical parametric or variational design
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2113/00—Details relating to the application field
- G06F2113/06—Wind turbines or wind farms
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2119/00—Details relating to the type or aim of the analysis or the optimisation
- G06F2119/14—Force analysis or force optimisation, e.g. static or dynamic forces
Abstract
The invention discloses an aeroengine blade flow field computing platform based on a grid automatic generation technology, which comprises S1) reading aeroengine blade outline data and establishing a model; s2) adopting an automatic hexahedron mesh generation method, namely an overrun interpolation method of a mapping function method to divide the model into features, wherein the method comprises the steps of encrypting the one-dimensional contour line features; dividing a two-dimensional plane of the hexahedral mesh; dividing a single hexahedron; generating a regular hexahedron corner point; generating hexahedron units one by mapping the areas between the two-dimensional planes, the curved surface of the three-dimensional space and the area of the three-dimensional space; s3) combining all hexahedron units to generate finite element units, and finishing calculation of the flow field of the aeroengine blade. The invention adopts the line-based transfinite interpolation algorithm, realizes the generation of the grid with high precision and high quality, and has the advantages of high division flexibility, high generation speed and good grid quality for a single large complex model.
Description
Technical Field
The invention relates to the technical field of turbine machining, in particular to an aeroengine blade flow field computing platform based on a grid automatic generation technology.
Background
From the 60 s of the 20 th century, with the continuous development of computer technology, Computational Fluid Dynamics (CFD) technology is also continuously advanced in aerospace, ship water conservancy, chemical engineering and the like, and great achievement is achieved. Under the rapid development of the current fluid field, the CFD technology shows great application value more and more. The precondition for CFD to carry out high-quality calculation is to carry out high-quality mesh division, and the mesh division technology is a bridge between theory and experiment. When a group of complex models need to be processed, the traditional finite volume analysis software is difficult to directly divide the models, and the models are often divided in other software through various complex calculations and then introduced into common finite volume calculation software for processing. The biggest problem in meshing is that the amount of manual work is large, and the problem is also the bottleneck problem of the working efficiency of CFD.
When the turbine is used for aviation and navigation engines, the performance of the turbine directly influences the efficiency, noise and CO2 emission of airplanes and ships. Meanwhile, the device is widely applied to power generation of a gas turbine and has the characteristics of quick power generation reaction and low pollution. Meanwhile, the power generation of the gas turbine can be used as standby equipment for new energy power generation to compensate the power generation and solve the fluctuation problem of the new energy power generation. Therefore, the method has great significance for national aviation, navigation, energy generation and carbon neutralization. The turbine has complex, viscous and abnormal three-dimensional flow field accompanied by phenomena of multiphase flow, heat transfer, combustion and the like. The numerical calculation of the turbine flow field is subjected to simplified methods such as an S2 flow surface method, a finite difference solution, an S1 and an S2 flow surface method of a non-orthogonal curve coordinate system and the like. With the improvement of CFD algorithm and super computing capability, the method for solving the full three-dimensional flow field of the NS equation by using the finite volume method which is widely adopted at present can fully consider the complex phenomena of rotational flow, static and dynamic interference, transonic speed, supersonic speed and the like. The numerical platform for solving the turbine blade fluid is built by utilizing the characteristics of OpenFOAM, so that the complex flow field of the turbine can be optimized.
Disclosure of Invention
The invention aims to provide an aircraft engine blade flow field computing platform based on a grid automatic generation technology, which adopts a line-based overrun interpolation algorithm to realize high-precision and high-quality generation of grids and has the advantages of high division flexibility, high generation speed and good grid quality of a single large-scale complex model.
In order to achieve the purpose, the invention adopts the technical scheme that: an aviation engine blade flow field computing platform based on a grid automatic generation technology comprises the following steps,
s1) reading profile data of the aero-engine blade, and establishing a model, wherein the aero-engine blade comprises a stator blade and a rotor blade;
s2) the model is characterized by using an automatic hexahedral mesh generation method, i.e., an overrun interpolation method of a mapping function method, which includes the steps of,
s21) encrypting the one-dimensional contour line characteristics;
s22) dividing the two-dimensional plane of the hexahedral mesh;
s23) dividing the single hexahedron;
s24) generating regular hexahedron corner points;
s25) topological mapping comprises area mapping between two-dimensional planes, curved surface mapping of three-dimensional space and area mapping of three-dimensional space, and hexahedron units are generated one by one;
s3) combining all hexahedron units to generate finite element units, and finishing calculation of the flow field of the aeroengine blade.
As a further optimization, S2 further includes S26, and the quality of the hexahedral cell generated by the grid is evaluated, and the degree of abnormality of the hexahedral cell is determined by comparing the difference in the infinitesimal volume amplification factor at each integration point by the element determinant.
Compared with the prior art, the invention has the following beneficial effects: the transfinite interpolation algorithm is completely based on mesh division of lines, the quality problem of the generation of the curved surface mesh is automatically solved in the algorithm, namely the generated curved surface and the outer contour line are certainly overlapped, and therefore a large amount of manual interactive parameter adjusting processes brought by commercial software for ensuring the quality of the curved surface are avoided. In short, the program only needs to adjust the node distribution on the model lines to enable the model mesh to be generated automatically, and the high-quality curved surface of the module surface enables the whole model mesh to have higher quality, so that the accuracy of finite volume method calculation is improved.
Drawings
Fig. 1 illustrates a modular meshing method of the present invention.
Fig. 2 is a schematic diagram of a two-dimensional planar area map.
Fig. 3 is a schematic diagram of surface mapping in three-dimensional space.
Fig. 4 is a schematic diagram of region mapping in three-dimensional space.
Fig. 5 is a schematic diagram of a hexahedral region transformation.
FIG. 6a is a block diagram of outer flow field lines for a turbine stator blade.
FIG. 6b is a block diagram of outer flow field lines for a turbine rotor blade.
FIG. 7a is a simple grid model of the turbine stator blade outer flow field.
FIG. 7b is a simple grid model diagram of the turbine rotor blade outer flow field.
FIG. 8 is a diagram of a stator blade outer flow field grid model.
FIG. 9 is a diagram of a rotor blade outer flow field grid model.
Fig. 10 is a bottom view of a grooved flow field grid.
FIG. 11 is a histogram of a determinant detection of a rotor blade outer flow field lattice.
FIG. 12 is a histogram of the maximum-minimum interior angle detection of the rotor blade outer flow field grid
FIG. 13 is a bar graph of rotor blade outer flow field grid aspect ratio detection.
Fig. 14 is a streamline cloud diagram of a visualization of the flow field LIC.
FIG. 15 is a pressure cloud of a flow field cross-section.
Detailed Description
The following are specific embodiments of the present invention and are further described with reference to the drawings, but the present invention is not limited to these embodiments.
At present, the grid division software has low intelligence degree, most software needs manual fine adjustment after grid division, and when a divided model needs to be adjusted, all work needs to be implemented again.
Taking the Geomagic Studio grid division software as an example, the general division idea is as follows: 1. simplifying a geometric model by three-dimensional software; 2. generating a triangular patch; 3. manually dividing each two-dimensional unit; 4. generating three-dimensional grid cells from the two-dimensional cells; 5. and adjusting the model, generating an IGES file, and exporting calculation.
The cells divided by the steps can generate a finite volume model for calculation, and the method is suitable for meshing a small fluid model. The current grid division technology mainly comprises the following steps: automated mesh technology, overlapping mesh technology, cartesian mesh technology, adaptive mesh encryption technology. In these technical aspects, a lot of difficulties are not solved completely, and for large and complicated model engineering with millions of units at all, manual partitioning by using such software is difficult. Before a truly universal grid division method is not found, grids are divided by using other means by a mode of grasping model features through modeling calculation according to the special part of the model, and the fineness of grid division can be adjusted at will according to the specific situation of the model. The modular grid division method is characterized in that various characteristics of the model are finely classified, the more characteristics of the extracted model are, the more detailed the classification is, and when the model needs to be changed in the later stage, the more convenient the detail is changed.
The invention uses the C + + platform to program, selects the hexahedral mesh to divide the model, and the flow chart of the compiling algorithm is shown in figure 1. According to various characteristics of the model, the model is modularized, namely divided into a plurality of hexahedron modules. And finely controlling the node distribution of the hexahedron surface one-dimensional contour line according to the characteristics of the whole model. And for each hexahedron module, a uniform automatic mesh generation algorithm is adopted for mesh division. The invention adopts the transfinite interpolation method of the mapping function method to realize automatic grid generation.
The basic idea of the overrun interpolation is to construct a mapping between regions based on a continuous boundary function. For example, in the two-dimensional case, a region D on the physical plane (x, y) is transformed into a regular region on the calculation plane (xi, eta)
Furthermore, the solution can be simply and regularly solved on the (xi, eta) plane and then mapped back to the region D to form the required grid. The method comprises the following steps:
as shown in FIG. 2, [0,1 ] in the ξ η coordinate system]×[0,1]Is defined as a region
A square area having a curved edge in the xy coordinate system is defined as an area D. Assuming that there exists an ideal coordinate mapping u (ξ, η) → (x, y) between two coordinate systems, u being a coordinate mapping function describing the correspondence between coordinates (ξ, η) and (x, y), i.e., u (ξ, η), and u:
and
are respectively regions
And the boundary of D. The expression of the ideal mapping function u (ξ, η) is unknown, but the boundary node mapping of the two regions is known, such as the regions in FIG. 3Each node on the boundary in u (0, η) in D corresponds to a region
Of the node (c) is located on the boundary of the η coordinate axis. In other words, u (0, η), u (1, η), u (ξ,0), u (ξ,1) in the region D can be understood as the continuous boundary functions mentioned above. From the ideal mapping function u (ξ, η), a local region coordinate map F can be constructed:
f is a description of two regions
And D, F (xi, eta) is the corresponding relation between coordinates (xi, eta) and (x, y). Here, F (xi, eta) and u (xi, eta) are in the region
The same mapping relation is satisfied on the boundary with D, i.e.
Therefore, F can be understood as an interpolation function of u with infinite points on the boundary that satisfy u, called an overrun interpolation function. The construction of the overrun interpolation function has various forms, and a general simple linear overrun interpolation function construction process is provided:
(1) a linear interpolation between the boundaries u (0, η) and u (1, η) in the direction of the coordinate ξ is constructed:
Lξ＝(1-ξ)u(0,η)+ξu(1,η)，
(2) a linear interpolation between the boundaries u (ξ,0) and u (ξ,1) in the direction of the coordinate η is constructed:
Lη＝(1-η)u(ξ,0)+ηu(ξ,1)，
(3) constructing a bilinear interpolation based on four corner points u (0,0), u (1,0), u (0,1) and u (1, 1):
(4) the linear over-limit interpolation function is:
F(ξ,η)＝Lξ+Lη-Lξη，
in the formula, xi is more than or equal to 0 and less than or equal to 1, eta is more than or equal to 0 and less than or equal to 1, and F (xi, eta) obviously meets the requirement.
The description of the overrun interpolation for a general two-dimensional quadrilateral plane is described above, and the overrun interpolation for a three-dimensional hexahedron needs to be divided into two steps: 1. transfinite interpolation of the hexahedral surface three-dimensional curved surface; 2. and (4) overrun interpolation inside the hexahedron.
As shown in FIG. 3, the ξ η ζ coordinate system is in the ξ η plane [0,1 ]]×[0,1]Is defined as a region
A square area having a curved edge in the xyz coordinate system is defined as an area D. Assuming that there exists an ideal coordinate mapping u (ξ, η, ζ) → (x, y, z) between two coordinate systems, u is a coordinate mapping function describing the correspondence of coordinates (ξ, η, ζ) and (x, y, z), i.e., u (ξ, η, z), and u:
and
are respectively regions
And the boundary of D. The expression of the ideal mapping function u (ξ, η, ζ) is unknown, but the boundary node mapping relationships of the above two regions, i.e., the continuous boundary functions u (0, η,0), u (1, η,0), u (ξ,0,0), u (ξ,1,0) are known. From the ideal mapping function u (ξ, η, ζ), a local region coordinate map F can be constructed:
f is a description of two regions
And D, i.e., F (ξ, η, ζ), with respect to the correspondence between the coordinates (ξ, η, ζ) and (x, y, z). Here, F (xi, eta, zeta) and u (xi, eta, zeta) are in the region
The same mapping relation is satisfied on the boundary with D, i.e.
F can therefore be understood as an interpolation function of u with an infinite number of points on the boundary that satisfy u, called an overrun interpolation function. Because in the square area
Where ζ is a constant 0, F (ξ, η, ζ) can be represented as F (ξ, η); regarding the construction of the overrun interpolation function, a general construction process of a simple linear overrun interpolation function is given here:
(1) a linear interpolation between the boundaries u (0, η,0) and u (1, η,0) in the direction of the coordinate ξ is constructed:
Lξ＝(1-ξ)u(0,η,0)+ξu(1,η,0)，
(2) a linear interpolation between the boundaries u (ξ,0,0) and u (ξ,1,0) in the direction of the coordinate η is constructed:
Lη＝(1-η)u(ξ,0,0)+ηu(ξ,1,0)，
(3) constructing a bilinear interpolation based on four corner points u (0,0,0), u (1,0,0), u (0,1,0), u (1,1, 0):
(4) the linear over-limit interpolation function is:
F(ξ,η)＝Lξ+Lη-Lξη，
in the formula, xi is more than or equal to 0 and less than or equal to 1, eta is more than or equal to 0 and less than or equal to 1, and F (xi, eta) obviously meets the requirement.
As shown in FIG. 4, [0,1 ] in the ξ η ζ coordinate system]×[0,1]×[0,1]Is marked as a region
And a hexahedral region having a curved surface in the xyz coordinate system is denoted as region D, and all mesh surfaces in the figure are discrete surfaces of the two regions, respectively. Similarly, assume that there is an ideal coordinate mapping u (ξ, η, ζ) → (x, y, z), i.e., u (ξ, η, z), between the two coordinate systems, and that u:
and
are respectively regions
And D. The expression of the ideal mapping function u (ξ, η, ζ) is also unknown, but the node positions of all the hexahedral surface mesh divisions can be determined by a surface interpolation function obtained by the overrun interpolation of the three-dimensional surface of the hexahedral surface, so that the known boundary surface functions are all in the figure, i.e., the surface node mapping relations of the above two regions, i.e., the required continuous boundary functions u (0, η 1, ζ), u (1, η, ζ), u ( η 0,0, ζ), u (ξ,1, ζ), u (ξ, η,0), u (ξ, η,1) are all known. From the ideal mapping function u (ξ, η, ζ), a local region coordinate map F can be constructed:
f is a description of two regions
And D with respect to the correspondence between the coordinates (ζ, η, ζ) and (x, y, z),i.e., F (ξ, η, ζ). Here, F (xi, eta, zeta) and u (xi, eta, zeta) are in the region
And D satisfy the same mapping relationship on the surface, i.e.
Since F (xi, eta, zeta) has infinite points on the surface of the hexahedron to satisfy the coordinate mapping relation represented by u, the F (xi, eta, zeta) can be regarded as an overrun interpolation function of the hexahedron. The construction process of the linear interpolation function about the hexahedron is as follows:
(1) a linear interpolation between the surfaces u (0, η, ζ) and u (1, η, ζ) in the direction of the coordinate ξ is constructed:
Lξ＝(1-ξ)u(0,η,ζ)+ξu(1,η,ζ)，
(2) a linear interpolation between the boundaries u (ξ,0, ζ) and u (ξ,1, ζ) in the direction of the coordinate η is constructed:
Lη＝(1-η)u(ξ,0,ζ)+ηu(ξ,1,ζ)，
(3) a linear interpolation between the boundaries u (ξ, η,0) and u (ξ, η,1) in the ζ direction of the coordinate is constructed:
Lζ＝(1-ζ)u(ξ,η,0)+ζu(ξ,η,1)，
(4) the construction is based on bilinear interpolation between boundaries u (0,0, ζ), u (0,1, ζ), u (1,0, ζ), u (1,1, ζ) in the direction of coordinates ξ and η:
(5) the construction regards bilinear interpolation between the boundaries u (ξ,0,0), u (ξ,0,1), u (ξ,1,0), u (ξ,1,1) in the coordinate η and ζ direction:
(6) the construction is based on bilinear interpolation between boundaries u (0, η,0), u (0, η,1), u (1, η,0), u (1, η,1) in the zeta and xi directions:
(7) constructing a trilinear interpolation based on eight corner points u (0,0,0), u (0,0,1), u (0,1,0), u (0,1,1), u (1,0,0), u (1,0,1), u (1,1,0), u (1,1, 1):
Lξηζ＝(1-ξ)[(1-η)[(1-ζ)u(0,0,0)+ζu(0,0,1)]+η[(1-ζ)u(0,1,0)+ζu(0,1,1)]+ξ[(1-η)[(1-ζ)u(1,0,0)+ζu(1,0,1)]+η[(1-ζ)u(1,1,0)+ζu(1,1,1)]]
(8) the hexahedral over-limit interpolation function is:
in the formula, xi is more than or equal to 0 and less than or equal to 1, eta is more than or equal to 0 and less than or equal to 1, zeta is more than or equal to 0 and less than or equal to 1, and the linear structure of F (xi, eta, zeta) meets the requirement.
The mesh partition aims to serve the computation of the finite volume method, and the quality of the mesh, namely the rationality of the mesh geometry, influences the computation precision. The evaluation criteria of the grid quality are many, such as internal angle, skewness, aspect ratio, warping amount and the like. Although parameters for evaluating the grid quality are numerous, the emphasis of grid quality evaluation is different in different application scenarios. The invention relates to fluid calculation simulation of a propeller external flow field, and the grid quality mainly takes a determinant and an internal angle as evaluation standards. In the hexahedral mesh, the internal angle can be understood as the internal angle of the surface quadrangle of the hexahedral unit, so that the calculation is easy, and the determinant represents the malformation degree of the hexahedral unit, so that the calculation is complex. The calculation method of the hexahedral unit determinant is as follows:
as shown in fig. 5, to transform a set of arguments (ξ, η, ζ) in the computation space into a new set of arguments (x, y, z) in the physical space, the transformation is represented by the following equation:
it is known that a point (x) in the coordinate system xyz1,y1,z1)、(x2,y2,z2)、(x3,y3,z3)、(x4,y4,z4)、(x5,y5,z5)、(x6,y6,z6)、(x7,y7,z7)、(x8,y8,z8) Corresponding to points (-1, -1, -1), (1,1, -1), (-1, -1,1), (1,1,1), (-1,1,1) respectively in the xi η ζ coordinate system.
Meanwhile, the hexahedron in the coordinate system ξ η ζ is constructed by connecting 8 known vertexes in a straight line, and the coordinate system xyz is a standard cube, and the mapping relation between two hexahedron areas needs to be satisfied. Thus, the mapping function can be constructed simply using tri-linear interpolation:
in the formula (I), the compound is shown in the specification,
for a transformation in three-dimensional space, the jacobian J can be expressed as:
a specific expression of the jacobian is thus obtained, the value of which represents the magnification factor of the volume of the infinitesimal after transformation, i.e. the infinitesimal volume at a single point in the coordinate system ξ η ζ is proportional to the magnification after transformation. In the principle of calculus transformation, the jacobian represents the ratio of n-dimensional volumes before and after transformation, for example, the three-dimensional calculus is as follows:
dxdydz＝Jdξdηdζ，
in practice, the determinant, also called relative determinant, of a hexahedral unit is a ratio, i.e. the ratio of the minimum jacobian value to the maximum jacobian value at all integration points within the unit. Therefore, the element determinant determines the degree of the deformation of the hexahedral element by comparing the difference of the infinitesimal volume amplification coefficients at each integration point.
In order to embody the universality of the algorithm, the invention is based on the model modularization method and the transfinite interpolation theory, and takes the stator blade and the rotor blade of the turbine as examples respectively to carry out free mesh division on the flow field outside the blade. Firstly, an external flow field model wire frame of a stator blade and a rotor blade is obtained by adopting a manual division mode. As shown in fig. 6, the stator vane has a simpler structure, and the outer flow field thereof is divided into 3 layers in the longitudinal direction; and the rotor blade is a complex structure with a groove at the top, so the longitudinal direction of the external flow field is divided into 5 layers. In fact, the wire frame of the out-of-blade flow field has divided the entire model into several hexahedral modules. If the specific characteristics of the whole flow field are not considered, simple uniform grid division can be performed for each module. As shown in fig. 7, a simple grid of the off-blade flow field is generated. Since the top of the rotor blade is a groove, the crescent part in fig. 7(b) is also covered by the flow field grid.
In order to ensure the reasonability and accuracy of fluid calculation, the division requirements of the flow field grids are very delicate, such as division of a fluid boundary layer, control of grid size gradual change and the like. Thus, simple mesh models clearly do not meet the fluid computation needs. The modular meshing method provided herein needs to determine specific characteristics of the entire model in order to preset node distribution for the one-dimensional contour line of each module, and then performs hexahedral meshing based on an overrun interpolation method. Fig. 8 shows the final meshing of the stator vane outer flow field. In order to meet the requirements of model characteristics and calculation accuracy, the integral grid density is greatly improved. The locally enlarged area in the diagram is a flow field grid containing the splicing part of the multiple modules at the front end of the blade and the boundary layer of the blade, and the grid division can be seen in consideration of the requirements of grid size gradual change and boundary layer grid expansion. Similarly, according to the modular meshing method, fig. 9 shows a final mesh generation result of the rotor blade outer flow field, because the rotor blade includes the grooves, the model features of the entire outer flow field are more complex, and the division of the modules is more detailed, for example, at a locally enlarged mesh position in the figure, i.e., at the tail of the groove, due to the groove boundary, more mesh expansion and gradual change are required to meet the demand of fluid calculation. Fig. 10 shows a bottom view of the flow field grid at the groove part, in which the raised grid part is the region completely corresponding to the blade groove, and the whole is crescent, and meanwhile, the grid division details at the head and the tail of the groove are partially enlarged.
Generally, to ensure the accuracy of fluid calculation, the division of the flow field grid needs to be very fine, which also results in that the number of cells divided by the whole model often reaches the level of millions. Such as the outer flow field grid of the rotor blade described above, the number of cells has reached over 200 ten thousand. In the face of such a large number of cells, the check of the grid quality is very important. In order to evaluate the quality of the flow field grid more conveniently and timely, the invention further develops the function of grid quality check based on a calculation framework generated by the grid.
In order to ensure the accuracy of grid quality check, the method compares the results with the commercial software ICEM CFD and improves the algorithm. For the complex flow field grid shown in fig. 9, taking the determinant with more important grid quality parameters as an example, fig. 11 shows a comparison of the detection results between the current checking algorithm program and the ICEM. In the figure, the arrows in the bar graph indicate that the data is far larger than the value range of the current ordinate axis, and it can be seen that the distribution contrast of the number of the cells of the two sides with respect to the grid determinant is basically consistent. In fact, the calculation on different platforms will generate certain precision errors, so there will be some errors in the specific distribution quantity, but the error ratio is very small and basically indistinguishable in the figure. The present invention continues to show the comparison of the detection results of the internal angle and the aspect ratio in the grid quality parameter with fig. 12 and fig. 13, respectively, and similarly, the cell proportion distribution of the two sides with respect to the grid parameter is substantially consistent, and with respect to the aspect ratio of the grid quality parameter, the ratio of the shortest side to the longest side of the grid cell is referred to.
The method realizes the turbine blade modular grid division method by using the transfinite interpolation method, and can effectively realize the finite volume grid division of a large complex model by showing the effect of using the characteristic value to control the model to divide the fineness degree. Different from the algorithm of commercial software ICEM for generating the grid based on the surface, the invention adopts the line-based transfinite interpolation algorithm, realizes the generation of the grid with high precision and high quality, and has the advantages of high division flexibility, high generation speed and good grid quality for a single large-scale complex model.
The invention adopts a three-dimensional steady-state Reynolds average k-omega SST turbulence model to simulate the flow of a two-stage turbine. As shown in fig. 14 to 15, information interaction is handled using a hybrid interface (mixingPlanes) at the mover-stator interface, a periodic boundary (cyclic ggi) is used in the radial direction to simplify the model, and the steadyCompressableRFFoam solver of Foam-extended 3.1 is used to solve the flow field information.
The specific embodiments described herein are merely illustrative of the spirit of the invention. Various modifications or additions may be made to the described embodiments or alternatives may be employed by those skilled in the art without departing from the spirit or ambit of the invention as defined in the appended claims.
Claims (9)
1. An aviation engine blade flow field computing platform based on a grid automatic generation technology is characterized by comprising the following steps,
s1) reading the profile data of the aero-engine blade and establishing a model;
s2) characterizing the model by using an automatic hexahedral mesh generation method, which includes the steps of,
s21) encrypting the one-dimensional contour line characteristics;
s22) dividing the two-dimensional plane of the hexahedral mesh;
s23) dividing the single hexahedron;
s24) generating regular hexahedron corner points;
s25), topology mapping is carried out, and hexahedron units are generated one by one;
s3) combining all hexahedron units to generate finite element units, and finishing calculation of the flow field of the aeroengine blade.
2. The aero-engine blade flow field computing platform based on grid automatic generation technology as claimed in claim 1, wherein the grid automatic generation method is an overrun interpolation method of a mapping function method.
3. The aero engine blade flow field calculation platform based on mesh automatic generation technology according to claim 1 or 2, wherein the topological mapping in S2 comprises an area mapping between two-dimensional planes, a curved surface mapping in a three-dimensional space, and an area mapping in a three-dimensional space.
4. The calculation platform for the flow field of the aero-engine blade based on the grid automatic generation technology as claimed in claim 3, wherein the construction process of the linear overrun interpolation function of the area mapping between the two-dimensional planes is,
a linear interpolation between the boundaries u (0, η) and u (1, η) in the direction of the coordinate ξ is constructed:
Lξ＝(1-ξ)u(0,η)+ξu(1,η)，
a linear interpolation between the boundaries u (ξ,0) and u (ξ,1) in the direction of the coordinate η is constructed:
Lη＝(1-η)u(ξ,0)+ηu(ξ,1)，
constructing a bilinear interpolation based on four corner points u (0,0), u (1,0), u (0,1) and u (1, 1):
the linear over-limit interpolation function is:
F(ξ,η)＝Lξ+Lη-Lξη，
wherein xi is more than or equal to 0 and less than or equal to 1, eta is more than or equal to 0 and less than or equal to 1, and F (xi, eta) satisfies the following formula,
5. the calculation platform for the flow field of the blades of the aero-engine based on the grid automatic generation technology as claimed in claim 4, wherein the construction process of the linear transfinite interpolation function of the curved surface mapping of the three-dimensional space is,
a linear interpolation between the boundaries u (0, η,0) and u (1, η,0) in the direction of the coordinate ξ is constructed:
Lξ＝(1-ξ)u(0,η,0)+ξu(1,η,0)，
a linear interpolation between the boundaries u (ξ,0,0) and u (ξ,1,0) in the direction of the coordinate η is constructed:
Lη＝(1-η)u(ξ,0,0)+ηu(ξ,1,0)，
constructing a bilinear interpolation based on four corner points u (0,0,0), u (1,0,0), u (0,1,0), u (1,1, 0):
the linear over-limit interpolation function is:
F(ξ,η)＝Lξ+Lη-Lξη，
wherein xi is more than or equal to 0 and less than or equal to 1, eta is more than or equal to 0 and less than or equal to 1, and F (xi, eta) satisfies the following formula,
6. the calculation platform for the flow field of the blades of the aero-engine based on the grid automatic generation technology as claimed in claim 5, wherein the construction process of the linear transfinite interpolation function of the area mapping of the three-dimensional space is,
a linear interpolation between the surfaces u (0, η, ζ) and u (1, η, ζ) in the direction of the coordinate ξ is constructed:
Lξ＝(1-ξ)u(0,η,ζ)+ξu(1,η,ζ)，
a linear interpolation between the boundaries u (ξ,0, ζ) and u (ξ,1, ζ) in the direction of the coordinate η is constructed:
Lη＝(1-η)u(ξ,0,ζ)+ηu(ξ,1,ζ)，
a linear interpolation between the boundaries u (ξ, η,0) and u (ξ, η,1) in the ζ direction of the coordinate is constructed:
Lζ＝(1-ζ)u(ξ,η,0)+ζu(ξ,η,1)，
the construction is based on bilinear interpolation between boundaries u (0,0, ζ), u (0,1, ζ), u (1,0, ζ), u (1,1, ζ) in the direction of coordinates ξ and η:
the construction regards bilinear interpolation between the boundaries u (ξ,0,0), u (ξ,0,1), u (ξ,1,0), u (ξ,1,1) in the coordinate η and ζ direction:
the construction is based on bilinear interpolation between boundaries u (0, η,0), u (0, η,1), u (1, η,0), u (1, η,1) in the zeta and xi directions:
constructing a trilinear interpolation based on eight corner points u (0,0,0), u (0,0,1), u (0,1,0), u (0,1,1), u (1,0,0), u (1,0,1), u (1,1,0), u (1,1, 1):
Lξηζ＝(1-ξ)[(1-η)[(1-ζ)u(0,0,0)+ζu(0,0,1)]+η[(1-ζ)u(0,1,0)+ζu(0,1,1)]+ξ[(1-η)[(1-ζ)u(1,0,0)+ζu(1,0,1)]+η[(1-ζ)u(1,1,0)+ζu(1,1,1)]]
(8) the hexahedral over-limit interpolation function is:
F(ξ,η,ζ)＝Lξ+Lη+Lζ-Lξη-Lηζ-Lζξ+Lξηζ，
wherein xi is more than or equal to 0 and less than or equal to 1, eta is more than or equal to 0 and less than or equal to 1, zeta is more than or equal to 0 and less than or equal to 1, and the linear structure of F (xi, eta, zeta) satisfies the following formula,
7. the platform for calculating the flow field of the blades of the aero-engine based on the grid automatic generation technology as claimed in claim 1, wherein the step S2 further comprises a step S26, the quality of the hexahedral cells generated by the grid is evaluated, and the degree of deformity of the hexahedral cells is judged by comparing the difference of the infinitesimal volume amplification coefficients at each integral point through a unit determinant.
8. The aero-engine blade flow field calculation platform based on grid automatic generation technology according to claim 7, wherein the hexahedral unit determinant is constructed by the process of,
the mapping function is constructed using tri-linear interpolation:
for a transformation in three-dimensional space, the jacobian J is represented as:
jacobian represents the ratio of n-dimensional volumes before and after transformation, n is 3:
dxdydz＝Jdξdηdζ。
9. the grid auto-generation technology-based aeroengine blade flow field computing platform of claim 1, wherein the aeroengine blade comprises a stator blade and a rotor blade.
Applications Claiming Priority (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202121604578 | 2021-07-15 | ||
| CN202121604578X | 2021-07-15 |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN113821885A true CN113821885A (en) | 2021-12-21 |
Family
ID=78921773
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111036994.9A Pending CN113821885A (en) | 2021-07-15 | 2021-09-06 | Aero-engine blade flow field computing platform based on grid automatic generation technology |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113821885A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117421517A (en) * | 2023-12-18 | 2024-01-19 | 中国空气动力研究与发展中心高速空气动力研究所 | Poisson equation source term rapid calculation method for background schlieren measurement density field |
-
2021
- 2021-09-06 CN CN202111036994.9A patent/CN113821885A/en active Pending
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117421517A (en) * | 2023-12-18 | 2024-01-19 | 中国空气动力研究与发展中心高速空气动力研究所 | Poisson equation source term rapid calculation method for background schlieren measurement density field |
| CN117421517B (en) * | 2023-12-18 | 2024-03-01 | 中国空气动力研究与发展中心高速空气动力研究所 | Poisson equation source term rapid calculation method for background schlieren measurement density field |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Secco et al. | Efficient mesh generation and deformation for aerodynamic shape optimization | |
| US7191161B1 (en) | Method for constructing composite response surfaces by combining neural networks with polynominal interpolation or estimation techniques | |
| US6606612B1 (en) | Method for constructing composite response surfaces by combining neural networks with other interpolation or estimation techniques | |
| CN105122248A (en) | Method for modelling a part, in particular a turbine blade | |
| CN110083968B (en) | Compressor characteristic prediction method based on correction of gas seal leakage influence numerical model | |
| CN112818573B (en) | Method for acquiring boundary layer non-local variable information for unstructured grid | |
| CN114077802B (en) | Particle modeling method using shape function interpolation to replace kernel function approximation | |
| CN113821885A (en) | Aero-engine blade flow field computing platform based on grid automatic generation technology | |
| Trompoukis et al. | Continuous adjoint‐based shape optimization of a turbomachinery stage using a 3D volumetric parameterization | |
| Li et al. | Multidisciplinary design optimization based on parameterized free-form deformation for single turbine | |
| Seshadri et al. | Supporting multi-point fan design with dimension reduction | |
| WO2020054086A1 (en) | Simulation method, simulation method using mbd program, numerical analysis device, numerical analysis system for mdb, numerical analysis program, and mbd program | |
| Arbabi et al. | Aerodynamic inverse blade design of axial compressors in three-dimensional flow using a commercial CFD program | |
| Meheut | Multidisciplinary Adjoint-based optimizations in the MADELEINE project: overview and main results | |
| Chen et al. | Integrated optimization method for transonic turbine based on direct free-form deformation and prescreening differential evolutionary optimizer | |
| CN108664742B (en) | Multi-objective optimization method for nacelle parametric design | |
| CN113486512A (en) | Flutter analysis method for functional gradient variable-thickness blade model | |
| CN115081100A (en) | Training method of nacelle design model, and manufacturing method and system of nacelle | |
| Hu et al. | Quasi-Developable B-Spline Surface Design with Control Rulings | |
| Merchant et al. | A cad-based blade geometry model for turbomachinery aero design systems | |
| Lee et al. | A strain energy minimization method for generating continuous NURBS-based motion curves in free-form surface machining | |
| Sadrehaghighi | Dynamic & Adaptive Meshing | |
| Milli et al. | Full-parametric design system to improve the stage efficiency of a high-fidelity HP turbine configuration | |
| Dong et al. | Study and Verification of Large-Scale Parallel Mesh Generation Algorithm for Centrifugal Pump | |
| CN111008492B (en) | High-order unit Euler equation numerical simulation method based on Jacobian-free matrix |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| TA01 | Transfer of patent application right |
Effective date of registration: 20220715 Address after: 215000 room 901-5, floor 9, building 6, No. 27, Zigang Road, science and education new town, Taicang City, Suzhou City, Jiangsu Province Applicant after: Suzhou nuerfei Technology Co.,Ltd. Address before: No. 1239 Siping Road, Yangpu District, Shanghai 200082 Applicant before: Duan Penghao |
|
| TA01 | Transfer of patent application right |