A rotating machine grid generation method based on feature region topology configuration

Abstract

The application discloses a rotating machinery grid generation method based on feature region topology construction, and comprises the following steps: dynamically constructing a rotating machinery M-THETA plane topology; globally calculating and distributing key node information of a grid; generating a feature region grid; and generating a 3D grid through M-THETA surface spanwise stretching and stacking. The application specifically solves the technical problems of complex geometry preprocessing and high-quality rotating machinery structure grid automatic generation in the field of rotating machinery, and further accelerates the application of domestic numerical simulation technology in the whole machine simulation of an aero-engine.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of simulation, and particularly relates to a rotating machinery grid generation method based on feature region topology construction in the field of industrial simulation. BACKGROUND

[0002] With the development of modern aviation industry and the increasingly complex needs of the national defense industry, CFD simulation plays an increasingly important role in the industrial design of aviation engines, and is also a key technical support for the technology leap and digital research and development in the field of national defense aviation. In the face of industrial simulation of complex rotating parts such as compressors, fans and turbines of aviation engines, high-efficiency and high-quality grid generation technology is the core technology in the field of simulation calculation.

[0003] Rotating machinery simulation usually needs to face complex and shape-variable geometric models, and the structure grid generated by traditional manual grid drawing and recording artificial scripts has low quality and cannot meet the increasingly complex geometric model feature expression. More importantly, the grid drawing and generation efficiency is low, which cannot meet the new mode of industrial design pursuing high efficiency. Therefore, in order to shorten the grid generation cycle and reduce the cost of manual grid drawing, the application establishes a method of dynamically constructing rotating machinery M-THETA parameter plane topology and stacking grid generation, which uses a set of initial templates and feature regions to automatically generate grid. When facing other complex geometric models, the template can be dynamically transformed to quickly match the geometric model to construct topology, while ensuring the grid quality of the topology parameter plane and its three-dimensional grid. The application provides a low-cost and high-efficiency grid generation technology path for the field of rotating machinery simulation grid automatic generation, and further expands the application prospect of rotating machinery grid automatic generation topology technology. SUMMARY

[0004] The technical problem to be solved by the application is to provide a rotating machinery grid generation method based on feature region topology construction, which solves the problems in the prior art and expands the prospect of rotating machinery grid automatic generation topology.

[0005] In order to solve the above technical problems, the application is realized by the following way:

[0006] A rotating machinery grid generation method based on feature region topology construction, comprising the following steps:

[0007] S1, importing a geomTurbo geometry file;

[0008] S2, constructing a feature region O4H template topology;

[0009] S3, generating a 3D grid by B2B face spanwise stretching and stacking;

[0010] S4, complete the rotating machine surface mesh generation.

[0011] Further, the specific steps of step S2 are as follows:

[0012] Based on the feature region O4H template topology, the O-grid general template of the blade topology main part is generated, and the feature template of the flow channel region is adaptively generated according to the topology type. Through the combination of the main blade O-grid area and the flow channel feature area, the M-THETA feature topology of the current active layer position is adaptively formed. Based on the feature region O4H template topology, the O4H main part grid key node information calculation and the distribution point number calculation are included, and the grid generated based on the extension of the blade O-grid area main part is generated;

[0013] S21, O4H main part grid key node information calculation and distribution point number calculation, the key node calculation is performed based on the adaptive blade topology main part O-grid topology feature, the relative position of the key node is calculated in advance when the main part grid topology is generated, and the grid entity line is calculated according to the position information of the key node of the topology grid surface, so as to complete the construction of the blade area O-grid topology;

[0014] S22, the grid generated based on the extension of the blade topology main part O-grid in the feature region grid, which is mainly composed of constant part and OHGap gap layer, together constitutes the feature region of B2B surface (M-THETA parameter plane) flow channel, and accurately expresses the M-THETA surface topology type and the flow channel change rule in the span direction.

[0015] Further, the specific steps of step S21 are as follows:

[0016] S211, calculation of grid leading edge and trailing edge point positions

[0017] In order to ensure that the overall import and export grid of the M-THETA plane and the external O-grid layout are more standardized in the physical space, appropriate leading edge and trailing edge point positions need to be constructed, and the blade width at the leading edge and trailing edge point positions is calculated. The blade width of the leading edge and trailing edge point can be regarded as the diameter of the inscribed circle of the blade at the leading edge and trailing edge point. A plurality of groups of sampling points near the leading edge and trailing edge point are calculated to obtain the center of the circle, the minimum circle is taken as the final inscribed circle, and the maximum blade width is calculated as the maximum limit of the diameter of the inscribed circle. Then, the grid leading edge and trailing edge point positions are obtained according to the ratio of the relative control distance to the length of the blade pressure surface and the suction surface.

[0018] S212, calculation of blade position in flow channel

[0019] The blade has four positions in the flow channel, specifically: blade to upper cycle region AREA2, blade to lower cycle region AREA3, blade leading edge to inlet region AREA1, and blade trailing edge to outlet region AREA4.

[0020] The distance from the blade to the upper / lower cycle is calculated by taking the geometric discrete points of the upper / lower cycle boundary and the discrete points of the blade suction / pressure surface, obtaining the two discrete points that are closest to each other, and then calculating the distance between the upper / lower cycle points by using the coordinate discrete points. At the same time, the point that is closest to the endpoint of the upper / lower cycle boundary to the blade inlet position is determined.

[0021] The distance from the blade to the inlet is calculated by obtaining the midpoint at the inlet, the distance from the midpoint to the endpoint of the blade leading edge, and the horizontal distance from the inlet to the blade leading edge is obtained based on the length of the midpoint and the endpoint of the leading edge. This gives us the location and size of the nodes on the blade.

[0022] S213, Calculation of Periodic Boundary Node Positions

[0023] When generating the mesh, the positions of the periodic boundary nodes are pre-calculated, and the periodic boundary mesh line entities are subsequently calculated using these positions. In the main mesh of O4H, there are 4 periodic boundary point positions that need to be calculated for each periodic boundary. When matching periodicity, the positions may be related to each other. The 4 periodic boundary point positions include position 0 as the inlet end face, position 1 as the boundary surface corresponding to the upper and lower periodic boundaries from the blade leading edge position, position 2 as the boundary surface corresponding to the upper and lower periodic boundaries from the blade trailing edge position, and position 3 as the outlet end face.

[0024] 1) For positions 0 and 3, the presence or absence of a constant component needs to be considered:

[0025] If there is no constant at the entrance, position 0 is 0.0; if there is a constant at the entrance, position 0 is the intersection of the last constant line at the entrance and the periodic boundary line.

[0026] If there is no constant at the exit, position 3 is 1.0; if there is a constant at the exit, position 1 is the intersection of the foremost constant line at the exit and the periodic boundary line.

[0027] 2) For positions 1 and 2, directly use the method of calculating the nearest point from the midline endpoint to the periodic boundary, and take the nearest point as the positions of positions 1 and 2. If the obtained positions 1 and 2 intersect with positions 0 and 3 or are too close, then positions 1 and 2 need to be optimized as follows:

[0028] 1. Cross-check: Determine if position 2 is greater than position 3;

[0029] 2. Distance Judgment: Calculate the horizontal distance from the trailing edge to position 3, and determine whether the ratio of the distance from position 2 to position 3 to this horizontal distance is less than a given threshold;

[0030] 3. If either of the above two conditions is met, for position 2, take the position of the corresponding upper cycle boundary line based on the midpoint of the M coordinates of the trailing edge point and position 3, and use it as the new optimized position;

[0031] S214. Calculation of intermediate node positions at entrance and exit boundaries.

[0032] When generating the mesh, the positions of the intermediate nodes at the inlet and outlet boundaries are pre-calculated. These positions are then used to calculate the mesh line entities at the inlet and outlet boundaries. The methods for determining the two intermediate node positions at the inlet and outlet boundaries are as follows:

[0033] 1) Uniform distribution of entrances and exits: The location is calculated based on the ratio of the number of points on the three lines of the entrance and exit boundaries;

[0034] 2) The overall entrance and exit are dense in the middle and gradually sparse towards both ends. The thickness of the top layer of the boundary layer mesh is based on the distance between the two nodes at the front and rear edges after advancement. This thickness is used as the width of the dividing point where the entrance and exit are divided into 3 segments. The width of the remaining two ends is allocated according to the number of points.

[0035] S215, O4H main body point count calculation

[0036] The edge lines of the O4H initial type topology template are divided into 7 groups. The number and distribution of points in each group are the same. The determination of the number of points is divided into two methods: a given number of points and a given minimum number of points followed by calculation based on the actual model. The given number of points method is mainly used for the relatively fixed geometry of the HorizonMiddle, VerticalMiddle, and OutO edge lines in the O4H initial type topology template. The given minimum number of points followed by calculation based on the actual model method is mainly used for the HorizonFront and HorizonBack edge lines in the O4H initial type topology template that change drastically and require the number of points to be calculated based on the length of the flow channel in the flow direction, or the VerticalUp and VerticalDown edge lines whose corresponding geometric scale is affected by the number of periodic blades.

[0037] A reference length is set based on the geometric model. If the calculated geometric length of a certain part is too long relative to the reference length, the number of points is increased; if it is too short, the number of points is decreased. At the same time, the number of topological points is calculated using a logarithmic function.

[0038] S216, Topology Construction Methods

[0039] Topology construction refers to the computation of line entities. These line entities need to represent the shape of ideal mesh lines. Lines mapped to geometric edges are obtained by sampling a small number of point coordinates on the geometric edges as mesh line entities. During mesh generation, a projection function is used to project these points onto the geometric edges. The shape of mesh lines within the mesh is calculated using boundary lines, thus generating the initial mesh. The specific steps are as follows:

[0040] 1) Initial topology template generation mainly involves 4 steps:

[0041] 1. Create all points in the initial topology state;

[0042] 2. Create all lines in the initial topology state;

[0043] 3. Modify the grid lines to match the specific topology type;

[0044] 4. Based on the specific topology type, create all faces but do not generate interior points on the faces;

[0045] 2) Boundary layer mesh generation requires calculating the distribution of the four lines on the wall and advancing the points on the blade outwards. Boundary layer mesh generation typically uses a normal-vector-guided advancement method, which needs to consider issues such as mesh intersection handling and mesh orthogonality. A relatively simple method for advancing the frame lines is chosen for boundary layer mesh generation, as follows:

[0046] 1. Calculate the total propulsion height H;

[0047] 2. Advance all nodes on the four lines of the wall by a height H along their normal direction to obtain a new line composed of a series of points;

[0048] 3. Intersections may occur in the new lines; intersection handling will be performed.

[0049] 4. Assign the distribution of the four solid lines on the wall to the lines after they have been advanced;

[0050] 5. Calculate and set the support line distribution;

[0051] 6. Generate points inside each grid block;

[0052] The core of this method is cross-intersection processing. There are two possible cross-intersection situations: one is self-intersection within a line; the other is cross-intersection between lines. The case of self-intersection within a line is as follows: when there is a concave point on the line to be propelled (in the state of turbine blades), the propulsion result may produce self-intersection. The method of removing self-intersection in this invention is: start from the starting point of the line after propulsion and calculate the turning angle at each point. When the turning angle at a certain point is greater than 90°, remove this point, and repeat until no point is removed.

[0053] 3) Flow channel mesh generation

[0054] 1. Entrance and exit grid lines entity

[0055] Based on the intermediate node positions of the inlet and outlet, calculate and extract the corresponding discrete points on the inlet and outlet as the inlet and outlet grid line entities;

[0056] 2. Periodic boundary grid line entity

[0057] Based on the positions of the upper and lower periodic boundary lines, calculate and determine the four intersection points that intersect with the inlet and outlet. These four intersection points are the maximum range of the left and right discrete points of the upper and lower periodic boundary lines. The discrete points within this range are the discrete point areas of the periodic boundary grid line entities.

[0058] 3. Entrance and exit horizontal center grid lines solid

[0059] Calculate using the upper and lower periodic solid lines and two endpoints. Given the number of sampling points, calculate a series of vectors within the periodic boundary of the two points. Translate the vectors to the corresponding periodic boundary node positions. The endpoint of the vectors is the sampling point required for the grid line solid.

[0060] 4. Internal vertical grid line entity

[0061] For the geometric model of rotating machinery with excessive blade tilt angle, the vertical grid lines of the generated M-THETA topology leading to the periodic boundary will point to the boundary layer region, and then from the boundary layer region to the periodic boundary. The distribution of the vertical grid line entities is consistent with the distribution of the horizontal intermediate grid lines at the inlet and outlet.

[0062] Furthermore, the specific steps of step S22 are as follows:

[0063] S221, constant partial mesh generation

[0064] In the constant part, one edge of a mesh face directly uses the edge of the original O4H main body, while the other three edges are generated based on the O4H exit and the actual exit position. The mesh distribution of the constant part is as follows: the distribution of the rightmost exit edge is the same as the distribution at the exit of the O4H main body, and the distribution of the upper and lower mesh lines is calculated to transition evenly with the horizontal mesh size at the exit of the O4H main body.

[0065] S222, OHGap Mesh Generation

[0066] The OHGap mesh is divided into an O-mesh that surrounds the blade wall and an innermost H-mesh. The O-mesh is generated directly using the boundary layer mesh generation method in the topology construction method. The H-mesh in the middle needs to distinguish whether the number of nodes on the pressure surface and suction surface of the blade are the same. If they are the same, a mesh surface is generated based on the top line of the O-mesh after the push. If they are not the same, a Connector needs to be generated at the mesh line entity above the O-mesh. The number of points of this Connector is the same as the number of points of the mesh line below the O-mesh, so that a mesh surface can be formed.

[0067] When generating the O-grid around the blade wall and the innermost H-grid, if the number of nodes on the pressure and suction surfaces is the same, node matching is completed. The OHGap mesh is directly advanced to generate the feature region, and its corresponding 3D mesh is periodic. After full-cycle replication, the pre-mesh boundary is well connected, and the mesh quality is high. Conversely, if the 3D mesh has periodic node mismatch, the mesh quality and mesh solution will be affected after full-channel periodic replication.

[0068] Furthermore, the specific method for step S3 is as follows:

[0069] When constructing a 3D mesh by stretching and stacking M-THETA surfaces, a 3D template needs to be stretched from the 2D M-THETA surface at the hub. The stretching height of the M-THETA surface along the x-direction is consistent with the width of the meridional plane. The height of the 3D mesh is controlled by the width of the meridional plane. Then, according to the specific position of the M-THETA in the meridional plane corresponding to the activation layer (the meridional plane range ratio is [0.0, 1.0]) (for example, the M-THETA in the middle of the meridional plane is the activation layer 0.5), the above-constructed B2B (M-THETA) surfaces are placed layer by layer into this 3D template. Then, the M-THETA surface mesh data stacked in the template is mapped onto the stretched 3D mesh template through surface mapping to form the final 3D stacked mesh.

[0070] Compared with the prior art, the present invention has the following beneficial effects:

[0071] 1) The M-THETA mesh topology dynamic construction technology based on topology templates can dynamically generate M-THETA surface topology for complex geometric models of rotating machinery (such as fans, turbines, splitters, cooling holes, etc.) by initializing templates without additional manual operation. That is, only one initial template is needed and dynamic template transformation can be performed to obtain other complex types of topologies such as fans, turbines, splitters, cooling holes, etc. This method greatly improves the versatility of the original mesh template and the efficiency of automatic mesh generation, and reduces R&D costs and user costs.

[0072] 2) The automatic 3D mesh generation technology based on spanwise stretching and stacking can effectively restore the global 3D mesh from the local parametric plane. The stacked 3D mesh can well express the local details of the topology of the M-THETA surface, which technically improves the difficulty of accurately expressing the global mesh of rotating machinery by parametric plane mesh. Attached Figure Description

[0073] Figure 1 This is a schematic diagram of the process steps of the present invention;

[0074] Figure 2 This is a schematic diagram showing the position of the blades in the flow channel according to the present invention;

[0075] Figure 3 This is a schematic diagram of the position information of the M-THETA plane periodic boundary points in this invention;

[0076] Figure 4 This is a schematic diagram illustrating the calculation of inlet and outlet node location information according to the present invention.

[0077] Figure 5 This is a schematic diagram of the topological edge relationship between O4H and LH in this invention;

[0078] Figure 6 This is a schematic diagram illustrating the generation of the internal feature mesh OHGap portion of the present invention. Detailed Implementation

[0079] The specific embodiments of the present invention will be further described in detail below with reference to the accompanying drawings and specific examples.

[0080] like Figure 1 As shown, a method for generating rotating mechanical meshes based on feature region topology construction includes the following steps:

[0081] S1. Import geomTurbo geometry file;

[0082] S2. Topology construction based on feature region O4H template, the specific steps are as follows:

[0083] The topology construction based on the feature region O4H template refers to generating a general O-grid template for the main body of the blade topology and a feature template for the flow channel region that is adaptively generated according to the topology type. By combining the main blade O-grid area and the flow channel feature area, the M-THETA feature topology at the current active layer position is adaptively formed. The topology construction based on the feature region O4H template includes the calculation of key node information and distribution point number of the O4H main body grid and the feature region grid generated by extending the main body of the blade O-grid area.

[0084] The meridional flow channel of a rotating machine is a set of M-THETA topologies distributed in the meridional region. These topologies need to be stacked longitudinally along the x-direction to form a 3D mesh area. The active layer represents the specific location of the M-THETA topology along the x-direction (its coordinates in the z, y plane (or m_theta parameter plane) within the meridional plane during stacking (the meridional range is [0.0, 1.0]). Its range is the vertical width of the meridional plane (the corresponding position ratio within the meridional plane is [0.0, 1.0]). For example, the M-THETA topology stacked as the top layer of the 3D mesh, because it is within the meridional region... The topmost layer, which is the 3D mesh that caps in the x-direction, has its M-THETA topology activation layer at position 1.0 in the meridional region. Similarly, the bottommost M-THETA topology activation layer is at position 0.0 in the meridional region, also called the activation layer 0.0. The activation layers of the stacked M-THETA topologies in the middle correspond to the range of (0.0, 1.0). For example, the middle section of the meridional region (0.5 scale position area) is the 0.5 activation layer of the M-THETA topology. All M-THETA topologies are stacked layer by layer along the x-direction according to the meridional range to form a 3D mesh.

[0085] The calculation of key node information and distribution points of the main body mesh of S21 and O4H adopts the blade O-mesh topology feature based on adaptive calculation of key nodes. In order to ensure that the 3D mesh does not undergo too large abrupt changes in the meridional spanwise direction, the relative positions of key nodes are pre-calculated when generating the main body mesh topology. Based on the position information of key nodes on the topology mesh surface (such as the position of nodes at the leading and trailing edges of the mesh, the position of periodic boundaries, and the position of intermediate nodes at the inlet and outlet of the flow channel), the mesh solid lines are calculated to complete the construction of the blade region O-mesh topology.

[0086] S22, the feature region mesh is a mesh generated by extending the main part of the blade O mesh region. Its structure is mainly composed of two parts: the constant part and the OHGap gap layer. Together, they constitute the characteristic region of the B2B surface (M-THETA parameter plane) flow channel, accurately expressing the M-THETA surface topology type and the flow channel variation law in the span direction.

[0087] Furthermore, the specific steps of step S21 are as follows:

[0088] S211 Calculation of the positions of the leading and trailing edges of the mesh

[0089] To ensure that the overall inlet and outlet mesh and outer O mesh layout of the M-THETA plane are more standardized in physical space, it is necessary to construct appropriate leading and trailing edge point positions and calculate the blade width near the leading and trailing edge point positions. The position is not controlled by the leading and trailing edge point positions, but by the relative control distance. The relative control distance is a normalized value of the relative blade width, that is, a certain proportion of the blade thickness of the geometric model. The blade width at the leading and trailing edge points can be approximated as the diameter of the inscribed circle of the blade at the leading and trailing edge points. Multiple sets of sampling points near the leading and trailing edge points are used to calculate the center of the circle. The smallest circle is used as the final inscribed circle. At the same time, the maximum blade width is calculated as the maximum limit of the inscribed circle diameter. Then, the position of the leading and trailing edge points of the mesh is obtained according to the ratio of the relative control distance to the length of the blade pressure surface and suction surface.

[0090] S212. Calculation of blade position in the flow channel

[0091] like Figure 2 As shown, the blade has four positions in the flow channel, specifically: from the blade to the upper cycle region AREA2, from the blade to the lower cycle region AREA3, from the blade leading edge to the inlet region AREA1, and from the blade trailing edge to the outlet region AREA4.

[0092] Distance from the blade to the upper / lower cycle ( Figure 2 In the middle (S1, S2), based on the geometric discrete point of the upper cycle boundary and the discrete point of the blade suction surface, the two discrete points that are closest to each other are obtained. Then, the distance between the upper and lower cycle points is calculated by using the coordinate discrete points to obtain the blade upper / lower cycle distance S1. At the same time, the point closest to the leading edge point A of the upper / lower cycle boundary is also determined. Similarly, the blade lower cycle distance S2 is obtained by using the geometric discrete point and the discrete point of the blade pressure surface.

[0093] The distance from the blade to the inlet / outlet is calculated by obtaining the midpoint Q at the inlet, the distance from the midpoint Q to the endpoint A of the blade's leading edge, and the horizontal distance from the inlet to the blade's leading edge is obtained based on the length QA*cosθ between the midpoint and the endpoint of the leading edge. For most models, the distance from the leading (tail) edge of the blade to the blade inlet and the distance from the leading (tail) edge point to the blade outlet are similar. Based on this distance, the distance from the leading (tail) edge point to the flow channel outlet can be obtained similarly, thus confirming the approximate location and size of the nodes on the blade.

[0094] S213, Calculation of Periodic Boundary Node Positions

[0095] like Figure 3As shown, when generating the mesh, the positions of the periodic boundary nodes are pre-calculated, and the periodic boundary mesh line entities are subsequently calculated using these positions. In the main mesh of O4H, there are 4 periodic boundary point positions that need to be calculated for each periodic boundary. When matching periodicity, the positions may be related to each other. The 4 periodic boundary point positions include position 0 as the inlet end face, position 1 as the boundary surface corresponding to the upper and lower periodic boundaries from the blade leading edge position, position 2 as the boundary surface corresponding to the upper and lower periodic boundaries from the blade trailing edge position, and position 3 as the outlet end face.

[0096] 1) For positions 0 and 3, the presence or absence of a constant component needs to be considered:

[0097] If there is no constant at the entrance, position 0 is 0.0; if there is a constant at the entrance, position 0 is the intersection of the last constant line at the entrance and the periodic boundary line.

[0098] If there is no constant at the exit, position 3 is 1.0; if there is a constant at the exit, position 1 is the intersection of the foremost constant line at the exit and the periodic boundary line.

[0099] 2) For positions 1 and 2, directly use the method of calculating the nearest point from the midline endpoint to the periodic boundary, and take the nearest point as the positions of positions 1 and 2 (shortest perpendicular segment principle). If the obtained positions 1 and 2 intersect with positions 0 and 3 or are too close, then positions 1 and 2 need to be optimized as follows:

[0100] 1. Cross-check: Determine if position 2 is greater than position 3;

[0101] 2. Distance Judgment: Calculate the horizontal distance from the trailing edge to position 3, and determine whether the ratio of the distance from position 2 to position 3 to this horizontal distance is less than a given threshold. A good topological block should have the same length on opposite sides. When the difference between the two ends is too large, Laplace optimization is required.

[0102] 3. If either of the above two conditions is met, for position 2, take the position of the corresponding upper cycle boundary line based on the midpoint of the M coordinates of the trailing edge point and position 3, and use it as the new optimized position;

[0103] S214. Calculation of intermediate node positions at entrance and exit boundaries.

[0104] like Figure 4 As shown, when generating the mesh, the positions of the intermediate nodes at the inlet and outlet boundaries are pre-calculated, and these positions are subsequently used to calculate the mesh line entities at the inlet and outlet boundaries. The method for determining the two positions of the intermediate nodes at the inlet and outlet boundaries is as follows:

[0105] 1. Uniform distribution of entrances and exits: The location is calculated based on the ratio of the number of points on the three lines of the entrance and exit boundaries;

[0106] 2. The overall entrance and exit are dense in the middle and gradually sparse towards both ends. The thickness of the top layer of the boundary layer mesh is determined by the distance between the two nodes at the front and rear edges after advancement. This thickness is used as the width of the dividing point where the entrance and exit are divided into 3 segments. The width of the remaining two ends is allocated according to the number of points.

[0107] The construction of key node information for the M-THETA plane can ensure the integrity of the model blade topology and is the technical basis for generating 3D mesh parameter planes.

[0108] S215, O4H Main Body Point Calculation Method

[0109] The edges of the initial type O4H topology template are divided into 7 groups. The number and distribution of edges in each group are the same. The correspondence of the edges needs to be optimized after topology optimization. Taking the transformed topology of O4H (double L topology, Low-High angle) as an example:

[0110] like Figure 5 As shown, at this time, VerticalUp only has one edge above the entrance. The two edges LINE1 and LINE2 in the figure become corresponding to the HorizonFront edge (the number of nodes matches). In order to match the periodicity, the number of points of HorizonFront and HorizonBack must be the same. That is to say, the number of points of the grid edges from 1 to 10 in the figure are all the same. At the same time, in the blade surface at this time, [11, 12<-->13] is one pair of opposite edges, and [2<-->3] is another pair of opposite edges.

[0111] The determination of the number of points is divided into two methods: a given number of points and a given minimum number of points followed by calculation based on the actual model. The given number of points method is mainly used for HorizonMiddle, VerticalMiddle, and OutO, whose geometry is relatively fixed in the edge of the O4H initial type topology template. The given minimum number of points followed by calculation based on the actual model is mainly used for HorizonFront and HorizonBack, whose edge of the O4H initial type topology template changes drastically and whose number of points needs to be calculated based on the length of the flow channel in the flow direction, or for VerticalUp and VerticalDown, whose corresponding geometric scale is affected by the number of periodic blades.

[0112] A reference length is set based on the geometric model. If the calculated geometric length of a certain part is too long relative to the reference length, the number of points is increased; if it is too short, the number of points is decreased (a minimum number of points needs to be limited). At the same time, the number of topological points is calculated using a logarithmic function.

[0113] Taking the point calculation for the HorizonFront section as an example (the principle is the same for the rest):

[0114] Let the blade length be bL and the length from the blade leading edge to the inlet be iL.

[0115] The ratio of the lengths of these two parts is ratio = iL / bL

[0116] The formula for calculating the points in HorizonFront is as follows:

[0117] V = ratio * e * control + e

[0118] PointNumber=log(V)*minPtsNum

[0119] In the formula, control is used to adjust the rate of increase or decrease of points (different positions may require different rates of increase or decrease, i.e., growth rate); +e is used to make the function graph pass through the logical (0, 1) point, so that when the ratio is close to 0 (relative to the reference length is very small), the minimum number of points minPtsNum will be taken; ratio*e is used to speed up the basic rate of increase or decrease of points. In the test of generating M-THETA surfaces, the point distribution of the resulting mesh topology when the ratio is 0-10 is more suitable for pre-mesh generation.

[0120] 1. The horizontal ratio is calculated as follows: the reference length is the blade length, the lengths used for comparison are the lengths from the blade to the leading edge and the blade to the trailing edge, and the value of control depends on whether topology optimization is involved.

[0121] 2. The vertical ratio is calculated as the ratio of the periodic length to the defined standard periodic length. The control also needs to distinguish whether there is topology optimization. When the blade's leading edge transitions to the middle, the number of points is more concentrated, and the number of points in the blade section will also increase significantly.

[0122] For the distribution calculation method, first determine the grid line distribution of the blade part, that is, the grid lines of the leading and trailing edges are evenly distributed, and the grid lines of the pressure and suction surfaces transition from the leading and trailing edges to the middle; then the distribution of the periodic boundary grid lines parallel to the grid lines of the pressure and suction surfaces is the same as the distribution of the top layer of the boundary layer grid; finally, after advancing the boundary layer grid, there is the thickness of the top layer of the boundary layer grid, and the remaining grid lines are distributed according to the principle of transitioning evenly with the boundary layer grid.

[0123] According to the principle of node distribution in the structural mesh blocks, the node distribution of the exit section must be consistent with that of the inlet section. Furthermore, to ensure orthogonality, the node distribution in the middle section of the inlet and outlet must also be consistent with the distribution of the outer O-grid close to the blade wall in the horizontal direction. This ensures that the distribution of opposite edges of each mesh block in the horizontal and vertical directions of the blade is consistent and that the entire blade topology is periodic, providing technical support for subsequent periodic boundary optimization.

[0124] S216, Topology Construction Methods

[0125] Topology construction refers to the computation of line entities. These line entities need to represent the shape of ideal mesh lines. Lines mapped to geometric edges are obtained by sampling a small number of point coordinates on the geometric edges as mesh line entities. During mesh generation, a projection function is used to project these points onto the geometric edges. The shape of mesh lines within the mesh is calculated using boundary lines, thus generating the initial mesh. The specific steps are as follows:

[0126] 1) Initial topology template generation mainly involves 4 steps:

[0127] 1. Create all points in the initial topology state;

[0128] 2. Create all lines in the initial topology state;

[0129] 3. Modify the grid lines to match the specific topology type;

[0130] 4. Based on the specific topology type, create all faces but do not generate interior points on the faces;

[0131] 2) Boundary layer mesh generation requires calculating the distribution of the four lines on the wall and advancing the points on the blade outwards. Boundary layer mesh generation typically uses a normal-vector-guided advancement method, which needs to consider issues such as mesh intersection handling and mesh orthogonality. A relatively simple method for advancing the frame lines is chosen for boundary layer mesh generation, and the general process is as follows:

[0132] 1. Calculate the total propulsion height H;

[0133] 2. Advance all nodes on the four lines of the wall by a height H along their normal direction to obtain a new line composed of a series of points;

[0134] 3. Intersections may occur in the new lines; intersection handling will be performed.

[0135] 4. Assign the distribution of the four solid lines on the wall to the lines after they have been advanced;

[0136] 5. Calculate and set the support line distribution;

[0137] 6. Generate points inside each grid block;

[0138] The core of this method is cross-intersection processing. There are two possible cross-intersection situations: one is self-intersection within a line; the other is intersection between lines. The case of self-intersection within a line is as follows: when there is a concave point on the line to be propelled (in the state of turbine blades), the propulsion result may produce self-intersection. The method of removing self-intersection in this invention is: start from the starting point of the line after propulsion and calculate the turning angle at each point. When the turning angle at a certain point is greater than 90°, remove this point, and repeat until no point is removed.

[0139] 3) Flow channel mesh generation

[0140] The generation of flow channel meshes mainly involves the generation of mesh line entities. Mesh line entities can approximate the geometric shape to a certain extent, but accurate geometric shapes still require projection of mesh nodes when discretizing mesh lines.

[0141] 1. Entrance and exit grid lines entity

[0142] Based on the intermediate node positions of the inlet and outlet, calculate and extract the corresponding discrete points on the inlet and outlet as the inlet and outlet grid line entities; if the intermediate node positions of the inlet are determined to be 0.3 and 0.7 (i.e., accounting for 30% and 70% of the total inlet), then several points on the segment of the inlet geometric line (between 0.0-0.3 and 0%-30%) can be directly obtained as the grid line entities.

[0143] 2. Periodic boundary grid line entity

[0144] Based on the positions of the upper and lower periodic boundary lines, calculate and determine the four intersection points that intersect with the inlet and outlet. These four intersection points represent the maximum range of the discrete points of the upper and lower periodic boundaries. The discrete points within this range are the discrete point areas of the periodic boundary grid line entity. The discrete points within this range are within the control (accessible) range of the periodic boundary grid line entity (array).

[0145] 3. Entrance and exit horizontal center grid lines solid

[0146] Calculate using the upper and lower periodic solid lines and two endpoints. Given the number of sampling points, calculate a series of vectors within the periodic boundary of the two points. Translate the vectors to the corresponding periodic boundary node positions. The endpoint of the vectors is the sampling point required for the grid line solid.

[0147] 4. Internal vertical grid line entity

[0148] For the geometric model of rotating machinery with excessive blade tilt angle, the vertical grid lines of the generated M-THETA topology leading to the periodic boundary will point to the boundary layer region, and then from the boundary layer region to the periodic boundary. The distribution of the vertical grid line entities is consistent with the distribution of the horizontal intermediate grid lines at the inlet and outlet.

[0149] Furthermore, the specific steps of step S22 are as follows:

[0150] S221, constant partial mesh generation

[0151] In the constant part, one edge of a mesh face directly uses the edge of the original O4H main body, while the other three edges are generated based on the O4H exit and the actual exit position. The mesh distribution of the constant part is as follows: the distribution of the rightmost exit edge is the same as the distribution at the exit of the O4H main body, and the distribution of the upper and lower mesh lines is calculated to transition evenly with the horizontal mesh size at the exit of the O4H main body.

[0152] S222, OHGap Mesh Generation

[0153] The OHGap mesh consists of an O-mesh surrounding the blade wall and an innermost H-mesh. The O-mesh is generated directly using the boundary layer mesh generation method in topology construction, requiring the specification of the initial layer height, layer number, and growth rate. The middle H-mesh needs to distinguish whether the pressure and suction surfaces of the blade have the same number of nodes. If they do, a mesh surface is generated based on the top line of the advanced O-mesh. If they do not, a Connector needs to be generated at the mesh line entity above the O-mesh. The number of nodes in this Connector must match the number of nodes in the mesh line below the O-mesh to form a mesh surface.

[0154] like Figure 6 As shown, when generating the O-grid around the blade wall and the innermost H-grid, if the number of nodes on the pressure surface and the suction surface are the same, node matching is completed. The OHGap grid is directly advanced to generate the feature region, and its corresponding 3D grid has periodicity. After full-cycle replication, the pre-grid boundary is well connected, and the grid quality is high. Conversely, if the 3D grid has periodic node mismatch, the grid quality and grid solution will be affected after full-channel periodic replication.

[0155] S3. Generate a 3D mesh by stretching and stacking B2B surfaces, as follows:

[0156] Since all M-THETA planes are the parametric planes of the 3D mesh, the M-THETA plane data for 3D template stacking are actually regular two-dimensional planes. At the same time, the 3D mesh template's spanning slice mesh data and 2D mesh data will be consistent, which can well express the global information of the 3D mesh.

[0157] When constructing a 3D mesh by stretching and stacking M-THETA surfaces, a 3D template needs to be stretched from the 2D M-THETA surface at the hub. The stretching height of the M-THETA surface along the x-direction is consistent with the width of the meridional plane. The height of the 3D mesh is controlled by the width of the meridional plane. Then, according to the specific position of the M-THETA in the meridional plane corresponding to the activation layer (the meridional plane range ratio is [0.0, 1.0]) (for example, the M-THETA in the middle of the meridional plane is the activation layer 0.5), the above-constructed B2B (M-THETA) surfaces are placed layer by layer into this 3D template. Then, the M-THETA surface mesh data stacked in the template is mapped onto the stretched 3D mesh template through surface mapping to form the final 3D stacked mesh.

[0158] In practical simulation applications, the flow surface data is divided into internal flow surface data and upper and lower gap flow surface data. The M-THETA plane with gaps is stacked separately in the stacking template as the top casing (bottom hub), and then stretched and merged with the remaining internal flow channel M-THETA stacked surface (internal flow surface data) to achieve accurate and effective engineering simulation results.

[0159] S4. Complete the generation of the rotating mechanical surface mesh.

[0160] The above description is merely an embodiment of the present invention. It should be reiterated that those skilled in the art can make several improvements to the present invention without departing from the principle of the present invention, and these improvements are also included within the scope of protection of the claims of the present invention.

Claims

1. A method for generating rotating mechanical meshes based on feature region topology construction, characterized in that: Includes the following steps: S1. Import geomTurbo geometry file; S2. Topological construction based on feature region O4H template; S3. Generate a 3D mesh by stretching and stacking B2B surfaces; S4. Complete the generation of the rotating mechanical surface mesh; The specific steps of step S2 are as follows: The topology construction based on the feature region O4H template refers to generating a general O-grid template for the main body of the blade topology and a feature template for the flow channel region that is adaptively generated according to the topology type. By combining the main blade O-grid area and the flow channel feature area, the M-THETA feature topology at the current active layer position is adaptively formed. The topology construction based on the feature region O4H template includes the calculation of key node information and distribution point number of the O4H main body grid and the feature region grid generated by extending the main body of the blade O-grid area. The calculation of key node information and distribution points of the main body mesh of S21 and O4H adopts the topological features of the O mesh of the main body of the blade topology for key node calculation. When generating the topology of the main body mesh, the relative positions of key nodes are pre-calculated. Based on the position information of key nodes on the topological mesh surface, the mesh solid lines are calculated to complete the construction of the O mesh topology of the blade area. S22, the feature region mesh is a mesh generated by extending the O mesh of the blade topology main part. Its structure consists of two parts: the constant part and the OHGap gap layer, which together constitute the feature region of the B2B surface flow channel. The specific method for step S3 is as follows: When constructing a 3D mesh by stretching and stacking M-THETA surfaces, a 3D template needs to be stretched from the 2D M-THETA surface at the hub. The stretching height of the M-THETA surface along the x-direction is consistent with the width of the meridional plane. The height of the 3D mesh is controlled by the width of the meridional plane. Then, according to the specific position of the M-THETA corresponding to the activation layer in the meridional plane, the B2B surfaces constructed above are placed into this 3D template layer by layer. Finally, the M-THETA surface mesh data stacked in the template is mapped onto the stretched 3D mesh template through surface mapping to form the final 3D stacked mesh.

2. The method for generating rotating mechanical meshes based on feature region topology construction as described in claim 1, characterized in that: The specific steps of step S21 are as follows: S211 Calculation of the positions of the leading and trailing edges of the mesh The blade width at the leading and trailing edge points can be approximated as the diameter of the inscribed circle of the blade at the leading and trailing edge points. Multiple sets of sampling points near the leading and trailing edge points are used to calculate the center of the circle. The smallest circle is used as the final inscribed circle. At the same time, the maximum blade width is calculated as the maximum limit of the inscribed circle diameter. The positions of the leading and trailing edge points of the grid are obtained according to the ratio of the relative control distance to the length of the blade pressure surface and suction surface. S212. Calculation of blade position in the flow channel The blade has four positions in the flow channel, specifically: blade to upper cycle region AREA2, blade to lower cycle region AREA3, blade leading edge to inlet region AREA1, and blade trailing edge to outlet region AREA4. The distance from the blade to the upper / lower cycle is calculated by taking the geometric discrete points of the upper / lower cycle boundary and the discrete points of the blade suction / pressure surface, finding the two discrete points that are closest to each other, and then calculating the distance between the upper / lower cycle points by using the coordinate discrete points. At the same time, the point that is closest to the endpoint of the upper / lower cycle boundary from the blade inlet position is determined. The distance from the blade to the inlet is calculated by obtaining the midpoint at the inlet, the distance from the midpoint to the endpoint of the blade leading edge, and the horizontal distance from the inlet to the blade leading edge is obtained based on the length of the midpoint and the endpoint of the leading edge. This gives us the location and size of the nodes on the blade. S213, Calculation of Periodic Boundary Node Positions When generating the mesh, the positions of the periodic boundary nodes are pre-calculated, and the periodic boundary mesh line entities are subsequently calculated using these positions. In the main mesh of O4H, there are 4 periodic boundary point positions that need to be calculated for each periodic boundary. When matching periodicity, the positions may be related to each other. The 4 periodic boundary point positions include position 0 as the inlet end face, position 1 as the boundary surface corresponding to the upper and lower periodic boundaries from the blade leading edge position, position 2 as the boundary surface corresponding to the upper and lower periodic boundaries from the blade trailing edge position, and position 3 as the outlet end face. 1) For positions 0 and 3, the presence or absence of a constant component needs to be considered: If there is no constant at the entrance, position 0 is 0.0; if there is a constant at the entrance, position 0 is the intersection of the last constant line at the entrance and the periodic boundary line. If there is no constant at the exit, position 3 is 1.0; if there is a constant at the exit, position 1 is the intersection of the foremost constant line at the exit and the periodic boundary line. 2) For positions 1 and 2, directly use the method of calculating the nearest point from the midline endpoint to the periodic boundary, and take the nearest point as the positions of positions 1 and 2. If the obtained positions 1 and 2 intersect with positions 0 and 3 or are too close, then positions 1 and 2 need to be optimized as follows:

1. Cross-check: Determine if position 2 is greater than position 3; 2. Distance Judgment: Calculate the horizontal distance from the trailing edge to position 3, and determine whether the ratio of the distance from position 2 to position 3 to this horizontal distance is less than a given threshold.

3. If either of the above two conditions is met, for position 2, take the position of the corresponding upper cycle boundary line based on the midpoint of the M coordinates of the trailing edge point and position 3, and use it as the new optimized position; S214. Calculation of intermediate node positions at entrance and exit boundaries. When generating the mesh, the positions of the intermediate nodes at the inlet and outlet boundaries are pre-calculated. These positions are then used to calculate the mesh line entities at the inlet and outlet boundaries. The methods for determining the two intermediate node positions at the inlet and outlet boundaries are as follows: 1) Uniform distribution of entrances and exits: The location is calculated based on the ratio of the number of points on the three lines of the entrance and exit boundaries; 2) The overall entrance and exit are dense in the middle and gradually sparse towards both ends. The thickness of the top layer of the boundary layer mesh is based on the distance between the two nodes at the front and rear edges after advancement. This thickness is used as the width of the dividing point where the entrance and exit are divided into 3 segments. The width of the remaining two ends is allocated according to the number of points. S215, O4H main body point count calculation The edge lines of the O4H initial type topology template are divided into 7 groups. The number and distribution of points in each group are the same. The determination of the number of points is divided into two methods: a given number of points and a given minimum number of points followed by calculation based on the actual model. The given number of points method is used for the relatively fixed geometry of HorizonMiddle, VerticalMiddle, and OutO edge lines in the O4H initial type topology template. The given minimum number of points followed by calculation based on the actual model is used for the drastic changes in the O4H initial type topology template edge lines, such as HorizonFront and HorizonBack, where the number of points needs to be calculated based on the length of the channel in the flow direction, or VerticalUp and VerticalDown, where the corresponding geometric scale is affected by the number of periodic blades. A reference length is set based on the geometric model. If the calculated geometric length of a certain part is too long relative to the reference length, the number of points is increased; if it is too short, the number of points is decreased. At the same time, the number of topological points is calculated using a logarithmic function. S216, Topology Construction Methods Topology construction refers to the computation of line entities. These line entities need to represent the shape of ideal mesh lines. Lines mapped to geometric edges are obtained by sampling a small number of point coordinates on the geometric edges as mesh line entities. During mesh generation, a projection function is used to project these points onto the geometric edges. The shape of mesh lines within the mesh is calculated using boundary lines, thus generating the initial mesh. The specific steps are as follows: 1) Initial topology template generation involves 4 steps:

1. Create all points in the initial topology state; 2. Create all lines in the initial topology state; 3. Modify the grid lines to match the specific topology type; 4. Based on the specific topology type, create all faces but do not generate interior points on the faces; 2) Boundary layer mesh generation requires calculating the distribution of the four lines on the wall and advancing the points on the blade outwards. Boundary layer mesh generation typically uses a normal vector-guided advancement method. A relatively simple method for advancing the frame lines is chosen for boundary layer mesh generation, and the process is as follows:

1. Calculate the total propulsion height H; 2. Advance all nodes on the four lines of the wall by a height H along their normal direction to obtain a new line composed of a series of points; 3. Intersections may occur in the new lines; intersection handling will be performed.

4. Assign the distribution of the four solid lines on the wall to the lines after they have been advanced; 5. Calculate and set the support line distribution; 6. Generate points inside each grid block; 3) Flow channel mesh generation 1. Entrance and exit grid lines entity Based on the intermediate node positions of the inlet and outlet, calculate and extract the corresponding discrete points on the inlet and outlet as the inlet and outlet grid line entities; 2. Periodic boundary grid line entity Based on the positions of the upper and lower periodic boundary lines, calculate and determine the four intersection points that intersect with the inlet and outlet. These four intersection points are the maximum range of the left and right discrete points of the upper and lower periodic boundary lines. The discrete points within this range are the discrete point areas of the periodic boundary grid line entities.

3. Entrance and exit horizontal center grid lines solid Calculate using the upper and lower periodic boundary lines and two endpoints. Given the number of sampling points, calculate a series of vectors within the periodic boundary of the two points. Translate the vectors to the corresponding periodic boundary node positions. The endpoint of the vectors is the sampling point required for the grid line entity.

4. Internal vertical grid line entity For the geometric model of rotating machinery with excessive blade tilt angle, the vertical grid lines of the generated M-THETA topology leading to the periodic boundary will point to the boundary layer region, and then from the boundary layer region to the periodic boundary. The distribution of the vertical grid line entities is consistent with the distribution of the horizontal intermediate grid lines at the inlet and outlet.

3. The method for generating rotating mechanical meshes based on feature region topology construction as described in claim 1, characterized in that: The specific steps of step S22 are as follows: S221, constant partial mesh generation In the constant part, one edge of a mesh face directly uses the edge of the original O4H main body, while the other three edges are generated based on the O4H exit and the actual exit position. The mesh distribution of the constant part is as follows: the distribution of the rightmost exit edge is the same as the distribution at the exit of the O4H main body, and the distribution of the upper and lower mesh lines is calculated to transition evenly with the horizontal mesh size at the exit of the O4H main body. S222, OHGap Mesh Generation The OHGap mesh consists of an O-mesh surrounding the blade wall and an innermost H-mesh. The O-mesh is generated directly using the boundary layer mesh generation method in the topology construction method. The middle H-mesh needs to distinguish whether the number of nodes on the pressure surface and suction surface of the blade are the same. If they are the same, a mesh surface is generated based on the top line of the O-mesh after the push. If they are different, a Connector needs to be generated at the mesh line entity above the O-mesh. The number of nodes of this Connector is the same as the number of nodes of the mesh line below the O-mesh, forming a mesh surface. When generating the O-grid around the blade wall and the innermost H-grid, if the number of nodes on the pressure surface and the suction surface are the same, node matching is completed, and the OHGap grid is directly advanced to generate the feature region, and its corresponding 3D grid has periodicity; otherwise, its 3D grid has periodic node mismatch.

Images