Rotor structure and method for an aviation high-speed permanent magnet generator

Abstract

The application discloses an aviation high-speed permanent magnet generator rotor structure and method. The inside of the rotating shaft is provided with an axially extending internal cavity, and a plurality of axially extending weight-reducing cavities are distributed in the circumferential direction outside the radial direction of the internal cavity; a connecting bridge area is arranged between adjacent weight-reducing cavities, and the two ends of the weight-reducing cavities are open; the rotating shaft is further provided with a wind guide opening for connecting the internal cavity and each weight-reducing cavity, and a thin-wall partition for separating and guiding airflow is arranged in the internal cavity and / or the weight-reducing cavity. The rotating induction ventilation structure is formed by combining the weight-reducing cavity, the wind guide opening and the thin-wall partition, so that the external cooling wind can enter the weight-reducing cavity and flow to the air gap heat exchange area under the action of wall surface viscous drag and centrifugal inertia caused by high-speed rotation of the rotor, thereby enhancing the convective heat exchange capacity of the air gap area and the related heat exchange surface, and reducing the local temperature rise.

Description

Technical Field

[0001] This invention belongs to the field of motor technology, and specifically relates to a rotor structure and method for a high-speed permanent magnet generator for aviation. Background Technology

[0002] Aviation power systems are evolving towards higher power density, higher speed, higher efficiency, and higher reliability. High-speed permanent magnet generators, due to their advantages of compact structure, high efficiency, and high power density, are increasingly being used in aviation power, distributed electric propulsion, and airborne auxiliary power systems. For aviation high-speed permanent magnet generators, the rotor is the core component for energy conversion and high-speed rotation; its structural design directly affects the overall weight, strength and safety, critical speed margin, thermal management capabilities, and long-term operational reliability.

[0003] Existing high-speed permanent magnet motors often employ surface-mounted permanent magnet rotor structures, with high-strength sheaths radially constraining the permanent magnets to prevent debonding, cracking, or scattering under high-speed centrifugal loads. Related research and patents indicate that high-speed permanent magnet rotor design typically requires simultaneous consideration of factors such as sheath preload, permanent magnet contact state, rotor stress, critical speed, and temperature rise. For aerospace applications, reducing rotor mass is beneficial for increasing power density, reducing moment of inertia, and improving dynamic response. Therefore, lightweight design of the rotor shaft is of great significance while meeting strength and dynamic performance requirements.

[0004] However, existing shaft weight reduction designs often focus on simple hollow shafts, axial holes, or uniform weight-reduction holes, typically achieving weight reduction by removing material. If such structures lack coordinated consideration of the load-bearing path and magnetic flux return path, they can easily weaken the local load-bearing cross-section of the shaft, increasing the equivalent stress under high-speed centrifugal loads and potentially reducing the overall rotor stiffness, leading to a decrease in the critical speed margin. Furthermore, for surface-mounted permanent magnet rotors without a rotor core or with weak magnetic support, improper local material removal from the shaft can also affect the continuity of magnetic flux return, thereby impacting electromagnetic performance.

[0005] Furthermore, existing high-speed permanent magnet motors mostly employ external air cooling, liquid cooling, axial ventilation holes, radial ventilation holes, or blade-type ventilation structures. While these methods can improve heat dissipation, some solutions require additional fans, blades, or complex flow channels, potentially increasing structural complexity, aerodynamic losses, manufacturing difficulty, and the risk of rotor imbalance. For aerospace high-speed permanent magnet generators, there is still room for improvement in how to utilize the rotor's own high-speed rotational characteristics to create effective airflow organization without significantly increasing additional components, and in reducing the risk of localized temperature rise near the permanent magnets, sheath, and air gap.

[0006] Finally, the design of high-speed permanent magnet rotor structures involves multiple constraints, including lightweighting, strength, stiffness, critical speed, radial deformation, contact pressure, heat dissipation capacity, and manufacturing feasibility. These design variables are significantly coupled. Traditional empirical design or single-objective optimization methods struggle to balance multiple performance indicators. Furthermore, uncertainties such as geometric manufacturing tolerances, temperature fluctuations, and speed disturbances exist in actual manufacturing and operation, potentially leading to insufficient reliability of solutions that meet nominal operating conditions in real-world service. Therefore, it is necessary to propose new structures and methods to achieve lightweight shafts while simultaneously ensuring strength, safety, dynamic performance, heat dissipation capacity, and robust reliability. Summary of the Invention

[0007] To address the problems existing in the background technology, this invention provides a rotor structure and method for a high-speed permanent magnet generator for aviation. It solves the technical problem that high-speed permanent magnet generator rotors in the prior art are difficult to balance lightweight, magnetic circuit continuity, structural load-bearing capacity, dynamic performance and thermal management requirements. It also solves the technical problem that the rotor structure design in the prior art lacks comprehensive consideration of multiple objectives, multiple constraints and uncertainties, which makes it difficult to balance robustness, reliability and engineering feasibility in the optimization results.

[0008] The technical solution adopted in this invention is:

[0009] I. A rotor structure for a high-speed permanent magnet generator for aircraft:

[0010] The device includes a rotating shaft, a surface-mount permanent magnet disposed on the outside of the rotating shaft, and a sheath sleeve disposed on the outside of the surface-mount permanent magnet. The rotating shaft has an internal cavity extending along the axial direction. Multiple weight-reducing cavities are also disposed radially outside the internal cavity and inside the rotating shaft. Each weight-reducing cavity extends along the axial direction, and the multiple weight-reducing cavities are distributed circumferentially at intervals along the internal cavity. Material that has not been removed is retained between two adjacent weight-reducing cavities to form a bridging area, which prevents the two adjacent weight-reducing cavities from communicating. The two ends of the weight-reducing cavities are hollowed out, so that each weight-reducing cavity is connected at its two ends. Both axial ends of each weight-reducing cavity are open, so that each weight-reducing cavity is connected to each other at its two axial ends.

[0011] The rotating shaft is also provided with an air guide and a thin-walled partition. The air guide connects the internal cavity to each weight-reduction cavity. The thin-walled partition is provided in the internal cavity and / or weight-reduction cavity to separate and guide the airflow inside the rotating shaft.

[0012] The rotating shaft is also provided with end hollowing structures at both ends of the axis; in the end hollowing structure, the connecting area between adjacent weight-reducing cavities is removed, so that each weight-reducing cavity is interconnected at the axial end, forming an annular connected area around the internal cavity.

[0013] The air guide is an opening on the radial sidewall of the internal cavity, and the air guide is located at the hollowed-out area at the end of the rotating shaft, so that the internal cavity is connected to the annular connecting area through the air guide, and then connected to each weight reduction cavity.

[0014] The thin-walled partition prevents the cooling airflow from forming ineffective backflow or short-path circulation in local areas, and guides the cooling airflow into the weight reduction cavity through the air guide, so that the cooling airflow flows along a preset path in the weight reduction cavity, and after heat exchange, it flows out from the axial end of the weight reduction cavity, enters the annular connecting area, and then flows to the air gap area on the outer side of the outer surface of the sheath.

[0015] The bridging region includes a local material retention region and a general weight reduction bridging region; the local material retention region is larger than the general weight reduction bridging region in terms of the thickness between two adjacent weight reduction cavities in the circumferential direction and / or in terms of the retention width in the radial direction of the rotating shaft; the local material retention region is located at at least one of the magnetic flux dense region, the key remagnetic region, and the main stress transmission path region, while the other regions are general weight reduction bridging regions.

[0016] The local material retention area extends radially from the side near the surface-mounted permanent magnet toward the center of the rotating shaft, or is locally widened circumferentially between adjacent weight-reducing cavities along the rotating shaft, thereby forming a non-uniform thickness bridge structure.

[0017] The cross-sectional profile of the weight-reducing cavity is formed by five convex arc segments connected end to end along the circumference, and all the convex arc segments have the same radius. The connection between two adjacent convex arc segments naturally forms a concave transition area. Each convex arc segment protrudes away from the center of the five-petal flower shape, and each concave transition area is recessed towards the center of the five-petal flower shape.

[0018] The cross-sectional profile of the weight-reducing cavity is a rotationally symmetric five-petal flower-shaped structure with five axes of symmetry. When arranged in the rotating shaft, one axis of symmetry of the five-petal flower-shaped structure is set radially along the rotating shaft. Along the radially set axis of symmetry, the end away from the center of the rotating shaft corresponds to the highest point of a convex arc segment, and the end closer to the center of the rotating shaft corresponds to the lowest point of the concave transition zone opposite the convex arc segment.

[0019] The rotor structure of the high-speed permanent magnet generator also includes a magnetic isolation ring and a retaining ring; the magnetic isolation ring is disposed in the axial end region of the surface-mounted permanent magnet; the retaining ring is disposed in the axial end of the surface-mounted permanent magnet and / or the sheath.

[0020] II. An optimization design method for the rotor structure of a high-speed permanent magnet generator for aircraft:

[0021] S1. Based on the overall design specifications and preliminary electromagnetic design results of the high-speed permanent magnet generator for aviation, determine the combination of reference rotor dimensions.

[0022] S2. Based on the combination of reference rotor dimensions, establish a parametric rotor model and use some structural dimensions in the rotor structure as initial design variables.

[0023] S3. Establish a high-fidelity analysis model, calculate and extract multiple performance response data of the rotor structure, and use them as performance evaluation indicators.

[0024] S4. Conduct sensitivity analysis on the initial design variables to identify key design variables that have a significant impact on performance evaluation indicators.

[0025] S5. Design experiments based on key design variables, establish surrogate models, and verify the prediction accuracy of surrogate models.

[0026] S6. Introduce uncertainty factors, conduct uncertainty propagation analysis based on the surrogate model, and evaluate the robustness and reliability of the rotor structure.

[0027] S7. With the optimization objectives of minimizing rotor mass, minimizing maximum equivalent stress, and maximizing critical speed, multi-objective optimization is performed on the basis of satisfying the constraints to obtain the final optimized scheme.

[0028] The parameterized rotor model includes at least one of the following: shaft, surface-mounted permanent magnet, sheath, weight-reducing cavity, air vent, thin-walled partition, ordinary weight-reducing bridge area, local material retention area, and end hollowing structure; the initial design variables include at least one of the structural dimensions of the weight-reducing cavity, thin-walled partition, ordinary weight-reducing bridge area, local material retention area, and end hollowing structure; the performance response data includes at least one of the following: rotor mass, maximum equivalent stress, critical speed, maximum radial deformation, and contact pressure.

[0029] The sensitivity analysis in step S4 uses the Morris method; the surrogate model in step S5 uses the Kriging model, and the verification index of the prediction accuracy includes at least one of the coefficient of determination, root mean square error, and maximum relative error. When the prediction accuracy does not meet the preset requirements, supplementary sampling is performed and the surrogate model is updated; the uncertainty factors in step S6 include at least one of geometric manufacturing tolerance, temperature fluctuation, and rotational speed disturbance, and the uncertainty propagation analysis uses the Monte Carlo sampling method; the multi-objective optimization in step S7 uses the NSGA-II algorithm; the constraints include at least one of strength constraints, critical speed margin constraints, maximum radial deformation constraints, contact pressure constraints, and manufacturing feasibility constraints.

[0030] The beneficial effects of this invention are:

[0031] 1. The present invention provides a weight-reducing cavity inside the rotating shaft, which can effectively reduce the rotor mass and moment of inertia, increase the rotor power density, and help improve the dynamic response performance of the high-speed motor.

[0032] 2. The present invention combines a weight-reducing cavity, an air guide, and a thin-walled partition to form a rotation-induced ventilation structure, which allows external cooling air to enter the weight-reducing cavity and flow to the air gap heat exchange area under the viscous drag of the wall and the centrifugal inertia caused by the high-speed rotation of the rotor. This enhances the convective heat transfer capacity of the air gap area and related heat exchange surfaces, and reduces local temperature rise.

[0033] 3. This invention guides and separates the airflow inside the rotor by using thin-walled partitions, which can limit ineffective airflow circulation and short-circuiting, improve the utilization rate of cooling airflow and effective heat exchange area, and improve the thermal management capability inside the rotor under high-speed operating conditions.

[0034] 4. The present invention provides a differentiated bridge structure between adjacent weight-reducing cavities, including a local material retention area and a normal weight-reducing bridge area, so that the local material retention area is larger than the normal weight-reducing bridge area in terms of circumferential bridge thickness and / or radial retention width, thereby forming an increased magnetic conductive cross section and load-bearing cross section in areas with dense magnetic flux, key areas of magnetic return, and main stress transmission paths.

[0035] 5. This invention avoids excessive weakening of the magnetic circuit and load-bearing path by the uniform hollowing structure, and can maintain the integrity of the rotor magnetic circuit, strength safety margin and critical speed margin while achieving weight reduction.

[0036] 6. The optimization design method proposed in this invention combines sensitivity analysis, sampling, surrogate model, uncertainty propagation, robustness, reliability, lightweight evaluation and multi-objective optimization, which can significantly reduce the number of high-fidelity simulation calls and improve optimization efficiency.

[0037] 7. This invention considers uncertainties such as geometric manufacturing tolerances, temperature fluctuations, and rotational speed disturbances during the optimization process, which can improve the reliability and robustness of the optimized scheme under actual manufacturing and service conditions. Attached Figure Description

[0038] Figure 1 This is a quarter-section sectional axonometric view of the rotor structure of the present invention.

[0039] Figure 2 This is a schematic diagram of the radial cross-section of the rotor structure of the present invention and a partial enlarged view thereof.

[0040] Figure 3 This is a first axial cross-sectional view of the rotor structure of the present invention.

[0041] Figure 4 This is a second axial cross-sectional view of the rotor structure of the present invention.

[0042] Figure 5 A flowchart of the method of the present invention.

[0043] In the figure: 1. Rotating shaft; 2. Surface-mounted permanent magnet; 3. Sheath; 4. Magnetic shielding ring; 5. Retaining ring; 11. Weight reduction cavity; 12. Air vent; 13. Thin-walled partition; 14. Connecting bridge area; 141. Ordinary weight reduction connecting bridge area; 142. Partial material retention area; 15. Internal cavity; 16. End hollowing structure. Detailed Implementation

[0044] The present invention will now be described in more detail with reference to the accompanying drawings and embodiments. However, the present invention is not limited thereto. For those skilled in the art, several improvements and modifications can be made without departing from the principles of the present invention, and these improvements and modifications are also considered to be within the scope of protection of the present invention. Contents not described in detail in this specification are prior art known to those skilled in the art.

[0045] Example 1:

[0046] like Figure 1 As shown, the rotor structure of the high-speed permanent magnet generator in this embodiment includes a shaft 1, a surface-mounted permanent magnet 2 disposed on the outside of the shaft 1, a sheath 3 sleeved on the outside of the surface-mounted permanent magnet 2, a magnetic isolation ring 4 disposed on the axial end region of the surface-mounted permanent magnet 2, and a retaining ring 5 disposed on the axial end of the surface-mounted permanent magnet 2 and / or the sheath 3. The retaining ring 5 is used to axially limit the surface-mounted permanent magnet 2 and / or the sheath 3 to prevent axial movement during assembly or high-speed operation. The rotor structure is a surface-mounted permanent magnet rotor structure, with the shaft 1 serving as the main load-bearing base, the surface-mounted permanent magnet 2 forming the rotor magnetic field, and the sheath 3 providing radial constraint on the surface-mounted permanent magnet 2 to meet the structural safety requirements under high-speed rotation conditions.

[0047] like Figure 2 As shown, the rotating shaft 1 has an internal cylindrical cavity 15 extending axially along the rotating shaft 1. Multiple weight-reducing cavities 11 are also provided radially outside the internal cavity 15 and inside the rotating shaft 1. These cavities are used to reduce the rotor mass and serve as local channels for cooling airflow through the rotor. The weight-reducing cavities 11 are directly connected to the rotor air gap region. Each weight-reducing cavity 11 extends axially, and the multiple weight-reducing cavities 11 are circumferentially spaced along the outer side of the internal cavity 15. Material that has not been removed is retained between adjacent weight-reducing cavities 11 to form a bridging region 14, preventing adjacent weight-reducing cavities 11 from communicating. The two ends of each weight-reducing cavity 11 are hollowed out, allowing each weight-reducing cavity 11 to communicate at both ends. Both axial ends of each weight-reducing cavity 11 are open, allowing each weight-reducing cavity 11 to communicate with each other at both axial ends. Figure 4 As shown, the rotating shaft 1 is also equipped with an air guide 12 and a thin-walled partition 13, the thickness of which is denoted as . The air vent 12 connects the internal cavity 15 to each weight reduction cavity 11; the thin-walled partition 13 is disposed in the internal cavity 15 and / or the weight reduction cavity 11 to separate and guide the airflow inside the rotating shaft 1, so as to reduce the ineffective circulation and short-circuiting of the airflow inside the rotor.

[0048] Specifically, the thin-walled partition 13 is provided in the internal cavity 15 to prevent airflow short-circuiting, and the thin-walled partition 13 is provided in the weight-reducing cavity 11 to extend the heat exchange path.

[0049] like Figure 2 As shown, the cross-sectional profile of the weight-reducing cavity 11 is formed by five convex arc segments connected end to end along the circumference, and all the convex arc segments have the same arc radius; the connection between two adjacent convex arc segments naturally forms a concave transition area; each convex arc segment protrudes in a direction away from the center of the five-petal flower shape, and each concave transition area is recessed in a direction towards the center of the five-petal flower shape.

[0050] The cross-sectional profile of the weight-reducing cavity 11 is a rotationally symmetric five-petal flower-shaped structure with five axes of symmetry. When arranged in the rotating shaft 1, one axis of symmetry of the five-petal flower-shaped structure is set radially along the rotating shaft 1. Along the radially set axis of symmetry, the end away from the center of the rotating shaft 1 corresponds to the highest point of the convex arc segment, and the end closer to the center of the rotating shaft 1 corresponds to the lowest point of the concave transition zone opposite the convex arc segment.

[0051] Specifically, since the five-petal flower-shaped cross section has rotational symmetry, any one of the five convex arc segments can be selected as the convex segment away from the center of the rotation axis 1; when arranging, select one of the convex arc segments, so that its vertex is located on the side of the rotation axis that is radially away from the center, and so that the lowest point of the concave transition area opposite it is located on the side of the rotation axis that is radially close to the center.

[0052] The five-petal flower-shaped structure of this invention avoids sharp corners through the continuous transition of the arc segment, which helps to reduce local stress concentration under high-speed rotation; the convex arc segment and the concave transition area are arranged alternately and symmetrically along the circumference, so that the material distribution around the weight reduction cavity is more balanced. While achieving weight reduction, it also takes into account the continuity of key magnetic cross section and load-bearing path, thereby improving the strength safety margin and manufacturing feasibility of the rotor.

[0053] Under the action of external cooling air and high-speed rotor rotation, the cooling airflow enters the internal cavity 15 region axially. The viscous drag of the rotating wall and centrifugal inertia create a rotation-induced centrifugal ventilation effect, causing some of the airflow in the internal cavity 15 to enter the weight-reducing cavity 11 through the air guide. Subsequently, it flows through the air gap region and related heat exchange surfaces, thereby enhancing the convective heat transfer capacity of the rotor's air gap region. It should be noted that this airflow transport mechanism differs from the active pressurization effect of traditional bladed centrifugal fans. Essentially, it is a gas transport process induced by the rotating wall, which can improve the heat dissipation conditions inside the rotor and the air gap region without the need for additional independent bladed pressurization components.

[0054] like Figure 3 As shown, the axial ends of the rotating shaft 1 are respectively provided with end hollowing structures 16. The radius of the end hollowing structures 16 is denoted as... Within the end hollowing structure 16, the connecting bridge area 14 between adjacent weight-reducing cavities 11 is removed, so that each weight-reducing cavity 11 is interconnected at its axial end, forming an annular connected area surrounding the internal cavity 15.

[0055] The air vent 12 is an opening on the radial side wall of the internal cavity 15, and the air vent 12 is located at the end of the hollowed-out area of ​​the rotating shaft 1, so that the internal cavity 15 is connected to the annular connecting area through the air vent 12, and then connected to each weight reduction cavity 11.

[0056] Specifically, the air guide includes radial air guides, inclined air guides, or combinations thereof. Radial air guides facilitate the transport of airflow from the inside of the rotor to the outside region under the action of centrifugal inertia; inclined air guides can be arranged according to the rotor rotation direction, the position of the weight reduction cavity 11, and the preset airflow path to improve the directionality and continuity of airflow entering the weight reduction cavity 11.

[0057] The thin-walled partition 13 prevents the cooling airflow from forming ineffective backflow or short-path circulation in a local area, and guides the cooling airflow through the air guide 12 into the weight reduction cavity 11, so that the cooling airflow flows along a preset path in the weight reduction cavity 11, and after heat exchange, it flows out from the axial end of the weight reduction cavity 11, enters the annular connecting area, and then flows to the air gap area on the outer side of the outer surface of the sheath 3.

[0058] In this embodiment, the thin-walled partition 13 is positioned and shaped as follows:

[0059] When multiple thin-walled partitions 13 are provided inside each weight-reducing cavity 11, the thin-walled partitions 13 are thin plate structures extending from one side wall of the weight-reducing cavity 11 into the interior of the weight-reducing cavity 11, and a gap is left between the extended end of the thin-walled partition 13 and the other side wall of the weight-reducing cavity 11 for the cooling airflow to pass through; the multiple thin-walled partitions 13 are arranged at intervals along the axial direction of the weight-reducing cavity 11, so that the cooling airflow is guided by each thin-walled partition 13 and forced to change its flow direction when flowing in the weight-reducing cavity 11, thereby prolonging the flow path and heat exchange time of the cooling airflow in the weight-reducing cavity 11, thereby improving the utilization rate of cooling airflow and the convective heat transfer effect.

[0060] When multiple thin-walled partitions 13 are provided inside the internal cavity 15, the thin-walled partitions 13 are thin plate structures extending from one side wall of the internal cavity 15 into the interior of the internal cavity 15, and a gap is left between the extended end of the thin-walled partition 13 and the other side wall of the internal cavity 15 for the cooling airflow to pass through, thereby dividing the internal cavity 15 into multiple flow areas along the axial direction, preventing the cooling airflow from directly penetrating along the axial direction of the internal cavity 15, limiting the airflow from forming ineffective backflow or short-path circulation in local areas, and guiding the cooling airflow to the inlet position of each air guide 12.

[0061] like Figure 2 As shown in the enlarged partial view, the bridging region 14 between adjacent weight-reducing cavities 11 is not completely of uniform thickness, but includes a local material retention region 142 and a regular weight-reducing bridging region 141. The local material retention region 142 is larger than the regular weight-reducing bridging region 141 in terms of thickness between two adjacent circumferential weight-reducing cavities 11 and / or in terms of radial retention width of the rotating shaft 1, thereby forming an increased effective magnetic conduction section and load-bearing section in critical areas. The local material retention region 142 is located at at least one of the corresponding locations in the magnetic flux density region, the key recirculation region, and the main stress transmission path region, while the other regions are the regular weight-reducing bridging region 141. By arranging the local material retention region 142 in the above-mentioned critical areas, the magnetic conduction continuity and structural load-bearing capacity of the bridging region 141 under high-speed rotation conditions can be improved, the risk of local stress concentration can be reduced, and the critical speed margin of the rotor can be increased.

[0062] like Figure 2 and Figure 3 As shown, a connecting bridge region 14 is provided between adjacent weight-reducing cavities 11. The connecting bridge region 14 is used to connect the shaft solid parts on both sides of the weight-reducing cavity 11, and has a predetermined extension length in the rotor axial direction. The extension length of the connecting bridge region 14 in the rotor axial direction is denoted as . .in, The axial length of the ordinary weight-reducing bridge region 141 is measured along the rotor rotation axis, with its two ends corresponding to the axial boundaries of the end hollowing structure 16.

[0063] The local material retention area 142 extends radially from the side near the surface-mounted permanent magnet 2 towards the center of the shaft 1, or is locally widened circumferentially between adjacent weight-reduction cavities 11, thereby forming a non-uniform thickness bridge structure. This structure can selectively retain material according to the magnetic flux distribution and stress transmission path, avoiding excessive material retention in non-critical areas, thus balancing weight reduction and performance maintenance. The local material retention area is used to increase the local cross-sectional area in areas with dense magnetic flux, critical remagnetic return areas, and / or areas corresponding to the main stress transmission paths, so that the shaft 1 maintains predetermined magnetic circuit integrity, strength safety margin, and critical speed margin while achieving weight reduction.

[0064] The ordinary weight-reducing bridge area 141 and the weight-reducing cavity 11 are used to reduce the rotor mass while meeting the minimum mechanical strength requirements; the local material retention area 142 is used to form an increased effective magnetic conduction section and load-bearing section in the magnetic flux dense area, the key area of ​​magnetic return and / or the area corresponding to the main stress transmission path.

[0065] Specifically, the local material retention area 142 is set up in at least one of the following two ways:

[0066] Firstly, in the circumferential direction of the connecting bridge region 14, corresponding to the magnetic flux density region, the key region of magnetic return and / or the main stress transmission path region, the thickness of the local material retention region 142 between two adjacent weight reduction cavities 11 in the circumferential direction is greater than that of other regions on the connecting bridge region 14, thereby forming an increased effective magnetic cross section.

[0067] Secondly, in the radial direction of the connecting bridge area 14, at the location corresponding to the magnetic flux density area, the key area of ​​magnetic return and / or the main stress transmission path area, the retention width of the local material retention area 142 in the radial direction of the rotating shaft 1 is greater than that of other areas on the connecting bridge area 14, thereby forming an increased load-bearing cross section.

[0068] The above configuration allows the local material retention area 142 to form a cross-sectional dimension larger than that of the ordinary weight-reducing bridge area 141 in the circumferential and / or radial directions, thereby forming an increased effective magnetic conduction cross section and load-bearing cross section in the magnetic flux dense area, the key area of ​​magnetic return and / or the area corresponding to the main stress transmission path.

[0069] Therefore, the interior of shaft 1 forms a non-uniform, non-uniform thickness differential material distribution hollow structure, rather than a simple uniform circumferential hollow structure. This differential material distribution hollow structure avoids the excessive weakening of the magnetic circuit and load-bearing path by the traditional uniform hollow structure, enabling the rotor to combine lightweight, magnetic circuit continuity, and structural load-bearing capacity.

[0070] During operation, external cooling air first enters the internal cavity 15 axially. When the rotor rotates at high speed, the inner wall of the shaft 1, the wall of the weight-reducing cavity 11, and the wall of the internal cavity 15 exert a viscous drag effect on the gas inside the cavity, giving the gas a circumferential velocity. Under the action of centrifugal inertia, the gas further forms a radial pressure gradient from the inside of the rotor to the outside. This radial pressure gradient causes part of the cooling airflow to enter the weight-reducing cavity 11 through the air guide 12, and then flows through the weight-reducing cavity 11 to the air gap region and related heat exchange surfaces, thereby forming a centrifugal ventilation effect induced by the rotating wall.

[0071] It should be noted that the centrifugal ventilation effect in this embodiment is not achieved through active pressurization generated by an independent blade centrifugal fan, but rather through the airflow transport process formed by the viscous drag of the gas on the high-speed rotating wall and the centrifugal inertia of the gas. Therefore, this rotor structure does not require additional blade pressurization components to improve the heat dissipation conditions of the internal area and air gap area of ​​the rotor shaft 1 to a certain extent.

[0072] With the above structure, the rotor in this embodiment forms circumferentially distributed weight-reducing cavities 11 and connecting bridge regions 14 in its cross-section, and weight-reducing cavities 11, air guides 12, and thin-walled partitions 13 with predetermined extension lengths in its axial direction. This structure can reduce rotor mass and rotational inertia while maintaining the continuity of critical magnetic circuits and main load-bearing paths through local material retention areas 142, and improving cooling airflow utilization through air guides 12 and thin-walled partitions 13, thereby achieving a balance between lightweight design, strength and safety, critical speed margin, and thermal management capabilities.

[0073] like Figure 5 As shown, this embodiment also provides an optimized manufacturing method for the rotor structure of an aviation high-speed permanent magnet generator, which is implemented according to the following steps:

[0074] S1. Based on the overall design specifications and preliminary electromagnetic design results of the high-speed permanent magnet generator for aviation, determine the combination of reference rotor dimensions.

[0075] Specifically, based on the overall design specifications and preliminary electromagnetic design results of the high-speed permanent magnet generator for aviation, the rotor outer diameter, effective axial length, thickness of the surface-mounted permanent magnet 2, thickness of the sheath 3, quality targets, strength limits, critical speed margin, maximum radial deformation limit, contact pressure requirements, and manufacturing constraints are determined. Different rotor size combinations are screened to obtain a benchmark rotor size combination that meets the requirements of electromagnetic performance, structural strength, dynamic performance, and quality, which serves as the initial design object for subsequent hollow structure optimization. Among them, the effective axial length refers to the length of the effective working area of ​​the rotor that actually participates in electromagnetic energy conversion along the rotation axis.

[0076] S2. Based on the combination of reference rotor dimensions, establish a parametric rotor model and set some structural dimensions in the rotor structure as initial design variables.

[0077] Based on the baseline rotor size combination, a parametric rotor model is established, including the shaft 1, surface-mount permanent magnet 2, sheath 3, weight-reducing cavity 11, air vent 12, thin-walled partition 13, connecting bridge area 14, ordinary weight-reducing connecting bridge area 141, local material retention area 142, and end hollowing structure 16, etc., and the relevant structural lightweighting dimensions (the relevant structural lightweighting dimensions in the embodiment include the structural dimensions of the weight-reducing cavity 11, thin-walled partition 13, end hollowing structure 16, ordinary weight-reducing connecting bridge area 141 and local material retention area 142 in the connecting bridge area 14) are set as initial design variables.

[0078] S3. Establish a high-fidelity analysis model, calculate and extract rotor mass, maximum equivalent stress, critical speed, maximum radial deformation and contact pressure as performance evaluation indicators.

[0079] A high-fidelity finite element analysis model, a rotor dynamics analysis model, and an analysis model under additional temperature conditions are established to describe the parameterized rotor model. The rotor mass, maximum equivalent stress, critical speed, maximum radial deformation, and contact pressure between the sheath 3, the surface-mounted permanent magnet 2, and the shaft 1 are calculated and extracted. The response is used as the performance evaluation index for subsequent sensitivity analysis, proxy modeling, and optimization design.

[0080] S4. Conduct sensitivity analysis on the initial design variables to identify key design variables that have a significant impact on performance evaluation indicators.

[0081] A global sensitivity analysis method was used to evaluate the influence of initial design variables on rotor mass, maximum equivalent stress, critical speed, maximum radial deformation, and contact pressure. The results of the influence of each initial design variable on various performance parameters under different values ​​were obtained. Based on this, key design variables that have a significant impact on lightweight performance, structural strength, and dynamic performance were selected to reduce the design space dimension of subsequent experimental design, surrogate modeling, and optimization solutions.

[0082] S5. Design experiments based on key design variables, establish surrogate models, and verify the prediction accuracy of surrogate models.

[0083] Based on the reduced-dimensional design space composed of key design variables, experimental design is carried out and sample points are generated. Then, a high-fidelity finite element analysis model and the corresponding thermal analysis model are called for each sample point to obtain sample data of rotor mass, maximum equivalent stress, critical speed, maximum radial deformation and contact pressure.

[0084] A surrogate model is established based on sample points and their corresponding response sample data. The surrogate model is used to quickly predict rotor mass, maximum equivalent stress, critical speed, maximum radial deformation and contact pressure. The prediction accuracy of the surrogate model is verified by cross-validation, coefficient of determination, root mean square error and / or maximum relative error. When the prediction accuracy does not meet the preset requirements, supplementary sampling is performed in the error area and the surrogate model is retrained until the preset accuracy requirements are met.

[0085] S6. Introduce uncertainty factors, conduct uncertainty propagation analysis based on the surrogate model, and evaluate the robustness and reliability of the rotor structure.

[0086] By introducing geometric manufacturing tolerances, temperature fluctuations, and speed disturbances as uncertainty factors, a probabilistic model or interval model of the uncertainty factors is established. Based on the surrogate model, uncertainty propagation analysis is performed to obtain the mean, standard deviation, distribution range, and / or failure probability of rotor mass, maximum equivalent stress, critical speed, maximum radial deformation, and contact pressure under uncertainty conditions, so as to evaluate the robustness and reliability of the rotor structure.

[0087] S7. With the optimization objectives of minimizing rotor mass, minimizing maximum equivalent stress, and maximizing critical speed, multi-objective robust reliability optimization is performed on the basis of satisfying the constraints to obtain the final optimized scheme.

[0088] Based on satisfying at least one of the following constraints: ultimate strength, critical speed margin, maximum radial deformation, sheath contact pressure, and manufacturing feasibility, robustness and reliability constraints (maximum equivalent stress, critical speed, maximum radial deformation, and temperature rise response) are introduced to establish a multi-objective robust and reliable optimization model. The model is solved using a multi-objective optimization algorithm with the optimization objectives of minimizing rotor mass, minimizing maximum equivalent stress, and maximizing critical speed, obtaining the Pareto optimal solution set. Candidate schemes are selected from the Pareto optimal solution set according to engineering design requirements, and these schemes are then substituted back into the high-fidelity finite element analysis model to verify their strength safety, critical speed margin, maximum radial deformation, contact state, thermal safety, and manufacturing feasibility. Finally, an optimized rotor structure scheme for aerospace high-speed permanent magnet generators that meets the requirements of lightweight, high reliability, and robustness is determined.

[0089] Furthermore, in order to demonstrate the beneficial effects of the present invention, this embodiment is further compared with the technical solutions in the prior art:

[0090] Patent document CN104079134A discloses a lightweight, high-power permanent magnet brushless DC motor rotor for airborne equipment, comprising permanent magnets, a rotor core, and a shaft. The shaft is a solid shaft and is thermally interference-fitted with the rotor core. Multiple circular slots are evenly arranged on the outer and inner diameter centers of the rotor yoke, with the number of slots corresponding to the number of permanent magnets. The centers of the slots are located on the central axis of each magnet, reducing the mass of the rotor yoke. Simultaneously, multiple semi-circular air slots are arranged on the circumference of the area where the shaft connects to the rotor core, forming ventilation channels with the slots in the rotor yoke for airflow, achieving rotor weight reduction and ventilation / heat dissipation during operation.

[0091] However, the patent document mainly sets circular slots in the rotor yoke and semi-circular air slots in the area where the shaft 1 connects with the rotor core. The structure is scattered and does not form the continuous axial hollowing and airflow organization structure for the inside of the shaft 1 as described in this invention. Furthermore, it does not disclose the technical solution in this invention where the internal cavity 15 and multiple weight-reducing cavities 11 are connected by air guides 12 and the airflow is separated and guided by thin-walled partitions 13. Its ventilation and heat dissipation mainly rely on the general ventilation channels formed by the slots, and the controllability of the airflow path and flow distribution is limited. This invention, on the other hand, takes the hollowing out of the inside of the shaft 1 as the optimization object. It forms a rotation-induced ventilation structure through the internal cavity 15, the axial weight-reducing cavities 11, the air guides 12, and the thin-walled partitions 13. This achieves the weight reduction of the shaft 1 while enhancing the heat exchange in the air gap area, and also takes into account the high-speed operating performance such as rotor strength, critical speed, radial deformation, and contact pressure.

[0092] Patent document CN113515843A discloses a method for optimizing the strength design of a permanent magnet motor rotor based on multidimensional visualization. The method includes: using analytical methods to solve the equations for the equivalent Mises stress of the non-magnetic metal sheath and the pressure of the permanent magnet, constructing the relationship between the equivalent Mises stress of the non-magnetic metal sheath, the pressure of the permanent magnet, the sheath thickness, and the interference between the sheath and the permanent magnet, and obtaining the design range of the sheath thickness and interference through multidimensional visualization, which is used to achieve the optimization design of the strength of the permanent magnet motor rotor.

[0093] However, this patent document did not establish a parametric optimization model for the hollowed-out structure inside the shaft 1, nor did it consider the dimensions and arrangement of the weight-reduction cavity 11, air guide 12, thin-walled partition 13, connecting bridge area 14, and end hollowed-out structure 16 as design variables. It also did not address the synergistic optimization of maximum equivalent stress, critical speed, maximum radial deformation, contact pressure, and ventilation and heat transfer performance based on rotor mass reduction. In contrast, this invention takes the hollowed-out structure inside the shaft 1 as the optimization object. Under the condition that the external structural parameters such as the surface-mounted permanent magnet 2 and the sheath 3 remain unchanged, it constructs the correlation between the internal structural parameters of the shaft 1 and lightweighting, structural strength, rotor dynamics, and rotation-induced ventilation performance, achieving optimized design of the hollowed-out structure of the shaft 1 under multiple performance constraints.

[0094] Example 2:

[0095] This embodiment uses the same rotor structure as Embodiment 1. The optimized design method of this embodiment is implemented according to the following steps:

[0096] S1. Based on the overall design specifications and preliminary electromagnetic design results of the high-speed permanent magnet generator for aviation, determine the combination of reference rotor dimensions.

[0097] Based on the overall design specifications of the high-speed permanent magnet generator, including power rating, rated speed, maximum speed, electromagnetic load, cooling conditions, and installation space, and combined with the preliminary electromagnetic design results, the basic dimensions such as rotor outer diameter, effective axial length, thickness of surface-mounted permanent magnet 2, thickness of sheath 3, and outer diameter of shaft 1 were determined. Further considering quality objectives, strength limits, critical speed margin, maximum radial deformation limit, contact pressure requirements, and manufacturing constraints, multiple candidate size combinations were screened to obtain a baseline rotor size combination that meets the requirements for electromagnetic performance, structural strength, dynamic performance, and quality.

[0098] S2. Based on the combination of reference rotor dimensions, establish a parametric rotor model and set some structural dimensions in the rotor structure as initial design variables.

[0099] Based on the combination of reference rotor dimensions, a parameterized rotor model is established, including a weight reduction cavity 11, a thin-walled partition 13, a common weight reduction connecting bridge area 141, a local material retention area 142, an end hollowing structure 16, and the effect of temperature.

[0100] The design variable vector is denoted as:

[0101]

[0102] in, To design a variable vector; The distance from the radial inner boundary of the weight reduction cavity 11 to the center of the rotating shaft 1 is the distance from the radial inner boundary to the center of the rotating shaft 1. The radial inner boundary is the location of the lowest point of the concave transition zone on the side closer to the center of the rotating shaft 1 in the direction of the axis of symmetry set radially along the rotating shaft 1. The distance from the radial outer boundary of the weight reduction cavity 11 to the center of the rotor is the distance from the radial outer boundary to the center of the rotor. The radial outer boundary is the location of the highest point of the convex arc segment on the side away from the center of the rotor shaft 1, along the symmetry axis direction set radially along the shaft 1. For rib thickness, the minimum radial thickness of the ordinary weight-reducing connecting bridge zone 141; The axial length of the rib is the axial length of the ordinary weight-reducing bridge section 141 after subtracting the depth of the end hollow structure 16. The radius of the end hollow structure 16 is the distance from the radial outer edge of the end hollow structure 16 to the center of the rotating shaft 1. The radius of the convex arc segment in the cross-sectional profile of cavity 11 is used to reduce weight. The wall thickness is 13 mm. This refers to the temperature effect. These seven parameters are used as design variable inputs, and are also set... The temperature effect term was added to the subsequent high-fidelity analysis model.

[0103] Each design variable vector satisfies manufacturing and structural constraints:

[0104]

[0105] in, and These are the lower and upper bound vectors of the design variables, respectively.

[0106] S3. Establish a high-fidelity analysis model, calculate and extract rotor mass, maximum equivalent stress, critical speed, maximum radial deformation and contact pressure as performance evaluation indicators.

[0107] A high-fidelity finite element analysis model, a rotor dynamics analysis model, and an analysis model under temperature rise conditions were established to calculate the rotor mass, maximum equivalent stress, critical speed, maximum radial deformation, and contact pressure.

[0108] The rotor mass can be expressed as:

[0109]

[0110] in, To design variable vectors, and The first The material density and volume of each constituent region.

[0111] The rotor weight reduction rate can be expressed as:

[0112]

[0113] in, For the mass of the un-hollowed or reference rotor, This represents the rotor mass under the current design.

[0114] The maximum equivalent stress constraint is:

[0115]

[0116] in, Design variable vector The maximum equivalent stress of the determined rotor structure, The allowable stress of a material is generally determined based on its yield strength or fatigue strength in conjunction with a safety factor.

[0117] The critical speed margin constraint is:

[0118]

[0119] in, For design variable vector The first critical speed of the rotor is determined. This is the actual maximum operating speed of the rotor. It is the critical speed margin coefficient. For rigid rotors, it is usually taken as 0.75, which means that the rotor operating speed is required to be less than 75% of the first-order bending critical speed.

[0120] The maximum radial deformation constraint is:

[0121]

[0122] in, For design variable vector The determined maximum radial displacement of the rotor, The maximum permissible radial deformation is usually determined by the air gap margin or mechanical safety requirements.

[0123] The contact pressure constraint is:

[0124]

[0125] in, The contact pressure between the sheath 3, the surface-mount permanent magnet 2, and the rotating shaft 1; Minimum contact pressure limit to ensure that loosening or detachment does not occur; Maximum contact pressure limit is used to prevent material crushing or excessive contact stress.

[0126] S4. Conduct sensitivity analysis on the initial design variables to identify key design variables that have a significant impact on performance evaluation indicators.

[0127] The Morris global sensitivity analysis method was used to evaluate the influence of each initial design variable on rotor mass, maximum equivalent stress, critical speed, maximum radial deformation and contact pressure.

[0128] Specifically, the design variable vector involved in the Morris global sensitivity analysis is:

[0129]

[0130] in, They are respectively number 1 to number 2. There are 10 design variables, all of which are input variables; This is the output for the model. Morris basic effect of design variables It can be represented as:

[0131]

[0132] in, The perturbation step size for normalized design variables is used to represent the perturbation step size for the th design variable while keeping other design variables constant. Design variables The applied standardization increment.

[0133] After calculating multiple basic effects from multiple Morris trajectories, the absolute value of the mean and the standard deviation can be further obtained:

[0134]

[0135]

[0136] in:

[0137]

[0138] in, For indexing; This represents the number of Morris sampling trajectories. Reflecting the The degree of influence of each design variable on the overall response. It reflects the degree of interaction or nonlinear influence between this variable and other variables. According to... and Key design variables can be filtered, reducing the dimensionality of optimization. It is generally believed that... The larger it is, the more likely it is to be the first The more significant the effect of a variable on the overall response, the greater its influence; conversely, if... A smaller value indicates that the variable has a weaker impact on the output within a given range of variation, and therefore has the potential to be filtered out. If... A smaller value indicates that the variable's effect is relatively stable across different sampling regions, and its response pattern is closer to linear or approximately additive; if... A larger value usually means that the effect of the variable changes significantly with the design point, or that there is a strong interaction between it and other variables.

[0139] S5. Design experiments based on key design variables, establish surrogate models, and verify the prediction accuracy of surrogate models.

[0140] Based on the dimensionality-reduced design space, the optimal Latin hypercube sampling method is used to generate sample points:

[0141]

[0142] in, The design variable sample matrix is ​​composed of all sample points; They are respectively the 1st, 2nd to the 3rd The vector of dimensionality-reduced design variables corresponding to each sample point.

[0143] The corresponding response was obtained by calling the high-fidelity analysis model. :

[0144]

[0145] in:

[0146]

[0147] They are respectively the 1st, 2nd to the 3rd The performance response vector obtained by high-fidelity analysis model for each sample point; This represents the number of sample points. Rotor mass; This represents the maximum equivalent stress of the rotor. This is the critical speed of the rotor; This represents the maximum radial deformation of the rotor; This represents the rotor contact pressure.

[0148] Establishing the Kriging proxy model:

[0149] Before establishing the Kriging surrogate model, the design variables and response values ​​are normalized. The Kriging surrogate model is typically represented as follows:

[0150]

[0151] in, To design variable vectors Predicted response at the location; This is a global trend term used to describe the overall trend of the response as the design variables change; The mean is zero and the variance is The stochastic process is used to describe local deviations and nonlinear variation characteristics.

[0152] In engineering applications, a constant trend term can be used:

[0153]

[0154] in, This indicates the average response or the overall trend level.

[0155] Related function expressions:

[0156] The Kriging model describes the correlation between different sample points using a spatial correlation function. For any two sample points... and Its correlation function can be expressed as :

[0157]

[0158] in; For indexing, To determine the number of design variables, For the first The correlation parameters in the direction of each design variable are determined by maximum likelihood estimation and are used to characterize the degree of influence of the variable on the response change. and The first The and the first The corresponding sample point of the th sample point The values ​​of the design variables; The smoothing parameter of the correlation function usually satisfies When using the Gaussian correlation function, we can let... Then we have:

[0159]

[0160] The correlation matrix can be constructed from the correlation function among all sample points. :

[0161]

[0162] For the point to be predicted The correlation vector between it and existing sample points is :

[0163]

[0164] in, This indicates the number of existing sample points. They represent the points to be predicted. With the 1st, 2nd to the 3rd The correlation function values ​​between existing sample points.

[0165] Kriging prediction formula:

[0166] Under the assumption of a constant trend term, the global trend coefficient It can be obtained by generalized least squares estimation:

[0167]

[0168] in, Global trend coefficient The generalized least squares estimate, for A vector of all 1s.

[0169] Points to be predicted The Kriging predicted response at point is:

[0170]

[0171] This formula shows that the predicted value of the Kriging model consists of two parts: one part is the global trend term. The other part consists of local correction terms based on sample correlation. This model allows for rapid prediction of rotor mass, maximum equivalent stress, critical speed, maximum radial deformation, and contact pressure under different hollowed-out structural parameters without repeatedly calling high-fidelity finite element models and rotor dynamics models.

[0172] The mean squared error of the Kriging model's predictions can be expressed as:

[0173]

[0174] in, Used to characterize the uncertainty of the surrogate model prediction at the point to be predicted; The variance estimate of a stochastic process is expressed as follows:

[0175]

[0176] Regions with large mean square errors in prediction usually indicate insufficient sample information or strong nonlinear response in those regions, and can be considered key areas for subsequent supplementary sampling.

[0177] Proxy model accuracy verification metrics:

[0178] To ensure that the Kriging surrogate model can reliably replace the high-fidelity analysis model, its prediction accuracy needs to be verified. This can be achieved through a comprehensive evaluation using metrics such as cross-validation, coefficient of determination, root mean square error, and maximum relative error.

[0179] 1. Cross-validation:

[0180] When using leave-one-out cross-validation, one sample point is removed from the sample set each time, and the remaining sample points are used for cross-validation. A Kriging surrogate model is built for each sample point, and predictions are made for the removed sample points. Cross-validation error of one sample point for:

[0181]

[0182] in, For the first High-fidelity response values ​​for each sample point; Not including the first The surrogate model's prediction for a given sample point.

[0183] 2. Coefficient of determination:

[0184] Coefficient of determination The expression used to evaluate the explanatory power of a surrogate model for changes in sample responses is:

[0185]

[0186] in, For the proxy model to the first Predicted values ​​for each sample point; This represents the mean of the sample responses. The closer the value is to 1, the better the overall fit of the surrogate model.

[0187] 3. Root mean square error:

[0188] The root mean square error (RMSE) measures the average deviation between the surrogate model's predicted values ​​and the high-fidelity response values. Its expression is:

[0189]

[0190] To facilitate comparison between different responses, the normalized root mean square error can also be used. :

[0191]

[0192] in, and These are the maximum and minimum values ​​of the sample response, respectively.

[0193] 4. Maximum relative error:

[0194] The maximum relative error is used to evaluate the prediction bias of the surrogate model at the most unfavorable sample points, and its expression is:

[0195]

[0196] This indicator reflects whether the surrogate model has significant prediction errors in local areas. For safety-sensitive responses such as maximum equivalent stress, critical speed, and contact pressure, the maximum relative error should be given special attention.

[0197] S6. Introduce uncertainty factors, conduct uncertainty propagation analysis based on the surrogate model, and evaluate the robustness and reliability of the rotor structure.

[0198] S61. Definition of the Uncertainty Factor Vector:

[0199] In the actual manufacturing, assembly, and operation of the rotor structure of a high-speed permanent magnet generator for aircraft, the structural response is affected not only by design variables but also by uncertainties such as manufacturing errors, material dispersion, assembly deviations, and fluctuations in operating conditions. Therefore, an uncertainty factor vector is introduced. :

[0200]

[0201] in, They represent the 1st, 2nd and... One uncertainty factor, The number of uncertain factors is denoted as . In this embodiment, geometric manufacturing tolerances, temperature fluctuations, and rotational speed disturbances are mainly selected as uncertain factors, and they are used as input variables of the surrogate model or input variable disturbances for uncertainty propagation analysis.

[0202] After taking uncertainty into account, the actual input variables can be expressed as a combination of design variables and uncertainty disturbances:

[0203]

[0204] in, To design variable vectors, For actual input variables that take uncertainty into account, The variable disturbance is caused by uncertainty disturbance.

[0205] S62. Probability Models and Interval Models:

[0206] For uncertainties with statistical samples or engineering experience data, a probabilistic model can be established:

[0207]

[0208] in, For the first One uncertainty factor, Indicates the first The probability distribution function of the uncertain factors. and These represent the mean and standard deviation of the uncertainty factor, respectively.

[0209] Furthermore, the set of uncertainties can be represented as:

[0210]

[0211] in, It is the uncertainty space composed of the probability distributions or intervals of various uncertainty factors.

[0212] S63. Uncertainty propagation based on the Kriging surrogate model:

[0213] In uncertainty propagation analysis, the perturbed input variables will be considered. Substituting these values ​​into the established Kriging proxy model, we can obtain rapid predictions of each performance response:

[0214]

[0215] in, Indicates the first A fast prediction of the performance response. Indicates the first The Kriging proxy model corresponding to each performance response.

[0216] Through Monte Carlo sampling, Latin hypercube sampling, or other uncertainty propagation methods, from the uncertainty space Generated in Group of uncertain samples:

[0217]

[0218] in, They represent the 1st, 2nd, and 3rd respectively. and Group of uncertain samples.

[0219] The corresponding number One response sample for:

[0220]

[0221] Based on the above response samples, the mean of the response under uncertainty conditions can be calculated. :

[0222]

[0223] Standard deviation :

[0224]

[0225] Minimum value of the distribution range and maximum value :

[0226] ;

[0227] Mean, standard deviation, and distribution range can be used to describe the response fluctuation characteristics of rotor structures under manufacturing and operational uncertainties. For responses with smaller maximum equivalent stress and maximum radial deformation, the smaller the standard deviation, the less sensitive the structural performance is to uncertainties and the better its robustness. For responses with larger critical speed, the smaller the standard deviation and the higher the mean, the more stable the dynamic performance.

[0228] S64. Robustness evaluation indicators:

[0229] To quantify the robustness of rotor structure designs, the mean and standard deviation of the response can be used as robustness evaluation indices. For responses that need to be minimized, such as maximum equivalent stress, maximum radial deformation, and temperature rise response, the robustness index can be defined as:

[0230]

[0231] in, For the first The robustness evaluation index corresponding to the response that needs to be minimized is such that the smaller the value, the better the overall performance of the response. This is a robustness weighting factor or safety factor used to adjust the degree of penalty for response fluctuations.

[0232] For responses that need to be maximized, such as critical speeds, a conservative evaluation form can be defined:

[0233]

[0234] in, This is the robustness evaluation index corresponding to the critical speed. and These are the mean and standard deviation of the critical speed under uncertainty conditions, respectively. This represents the corresponding robustness weighting coefficient. This formula indicates that a more conservative critical speed evaluation value is adopted after considering uncertain fluctuations.

[0235] For contact pressure, the lower and upper limits can be considered according to the constraint type. When sufficient contact pressure is required between the sheath 3 and the surface-mounted permanent magnet 2 to prevent separation, the following can be adopted:

[0236]

[0237] When excessive contact pressure is required to avoid localized damage to the permanent magnet or sheath, the following can be used:

[0238]

[0239] in, and These are the lower and upper limits of the robustness evaluation index for contact pressure, respectively. Let be the average contact pressure under uncertain conditions. The robustness weighting factor or safety factor corresponding to the contact pressure. Let the standard deviation of contact pressure under uncertain conditions be denoted as . and These are the lower and upper limits of the contact pressure, respectively.

[0240] S65. Reliability Evaluation and Failure Probability:

[0241] Reliability evaluation can be achieved by establishing functional functions. For the first... Given several constraints, the function is defined as follows:

[0242]

[0243] in, For the first The functional groups corresponding to each constraint condition For permitted limits, This is the corresponding predicted response value. When When, it indicates that the constraint is satisfied; when When this occurs, it indicates that a failure has occurred.

[0244] For responses that are preferred when smaller and are not allowed to exceed the upper limit, such as maximum equivalent stress and maximum radial deformation, they can be defined separately:

[0245]

[0246]

[0247] in, and These are the maximum equivalent stress constraint function and the maximum radial deformation constraint function, respectively. For allowable stress, To design the variable vector as Uncertainty factors are Predicted maximum equivalent stress of the rotor. This is the allowable value for maximum radial deformation. To design the variable vector as Uncertainty factors are The predicted value of the maximum radial deformation of the rotor.

[0248] For critical speed constraints, the following can be defined:

[0249]

[0250] in, This is the function for constraining critical speed. To design the variable vector as Uncertainty factors are The predicted critical speed of the rotor. For the maximum operating speed, This is the safety margin factor for the critical speed. When When the critical speed meets the safety margin requirements, it indicates that the critical speed meets the safety margin requirements.

[0251] For contact pressure constraints, a lower limit function can be defined:

[0252]

[0253] And the upper limit function:

[0254]

[0255] in, and These are the lower limit constraint function and the upper limit constraint function for contact pressure, respectively. and These are the lower and upper limits of the contact pressure, respectively. To design the variable vector as Uncertainty factors are Predicted contact pressure value.

[0256] The failure probability can be expressed as:

[0257]

[0258] in, For the first The failure probability corresponding to each constraint condition The probability of an event occurring. For the first The functional functions corresponding to each constraint condition.

[0259] based on For a set of uncertain samples, sample statistics can be used to estimate the probability of failure:

[0260]

[0261] in, To design the variable vector as , No. Group of uncertain samples Time The function value corresponding to each constraint condition. This is an indicator function that takes the value 1 if the condition within the parentheses is true, and 0 otherwise.

[0262] The corresponding reliability can be expressed as:

[0263]

[0264] in, For the first The reliability corresponding to each constraint condition; For the first The failure probability corresponding to each constraint condition.

[0265] S7. With the optimization objectives of minimizing rotor mass, minimizing maximum equivalent stress, and maximizing critical speed, multi-objective robust reliability optimization is performed on the basis of satisfying the constraints to obtain the final optimized scheme.

[0266] After completing uncertainty propagation and robust reliability evaluation, the key design variables selected through sensitivity analysis are used as optimization variables. The optimization variable vector can be represented as:

[0267]

[0268] Its value range is:

[0269]

[0270] in, For the 1st to the 1st One optimization variable, and These represent the lower and upper limits of the optimization variables, respectively. For the rotor hollowing structure illustrated in this example, the optimization variables may include the radius of the hollowing arc. The distance from the radial outer boundary of the weight-reducing cavity 11 to the rotor center , ribs thick axial length of the reinforcing bar The radius of the end hollow structure 16 The radius of the fillet in the weight-reducing cavity 11 and thin-walled partition 13 thickness One or more of them.

[0271] A multi-objective robust and reliable optimization model is established based on constraints of strength, critical speed, radial deformation, contact pressure, thermal safety, and manufacturing feasibility.

[0272] The optimization objective can be expressed as:

[0273]

[0274] in:

[0275]

[0276]

[0277]

[0278] In the formula, This represents the rotor quality target after taking fluctuations into account. This represents the target of maximum equivalent stress after considering fluctuations. By taking the negative sign, the problem of maximizing the critical speed is transformed into a minimization problem. , and All are robust weighting coefficients; and These are the mean and standard deviation of the rotor mass under uncertainty conditions, respectively. and These are the mean and standard deviation of the maximum equivalent stress under uncertainty conditions, respectively. and These are the mean and standard deviation of the critical speed under uncertainty conditions, respectively.

[0279] The constraints can be expressed as:

[0280]

[0281] in, The probability of the event within the parentheses occurring; and These are the predicted maximum equivalent stress and allowable stress of the rotor, respectively; This is the rotor's maximum operating speed; This is the safety margin factor for the critical speed; This is a predicted value for the critical speed. This is the predicted value of the rotor's maximum radial deformation; This is the allowable value for maximum radial deformation; This is the predicted contact pressure value; and These are the lower and upper limits of the contact pressure, respectively. This is the optimized variable vector after sensitivity analysis. and These are the lower bound vector and upper bound vector of the optimization variables, respectively; , , and The requirements are respectively for strength, critical speed, radial deformation, and contact pressure reliability, and the influence of temperature effect is taken into account.

[0282] The multi-objective robust and reliable optimization model can be comprehensively expressed as:

[0283]

[0284] in, , , , These are the reliability values ​​corresponding to the maximum equivalent stress, critical speed, maximum radial deformation, and contact pressure constraint, respectively. , , and These represent the rotor's maximum equivalent stress, critical speed, maximum radial deformation, and contact pressure, respectively.

[0285] The above model was solved using the NSGA-II multi-objective optimization algorithm to obtain the Pareto optimal solution set:

[0286]

[0287] in, This represents the Pareto optimal solution set obtained using the NSGA-II multi-objective optimization algorithm. These are the 1st, 2nd, and 3rd solutions in the Pareto optimal solution set, respectively. A non-dominated design scheme; The number of Pareto optimal design schemes.

[0288] For any two design schemes and If the following conditions are met:

[0289]

[0290] in, and Each represents a design scheme. and The corresponding number The objective function value is to be optimized.

[0291] And at least one objective satisfies:

[0292]

[0293] in, The index of the objective function is j∈{1,2,3}; and Each represents a design scheme. and The corresponding number The objective function value is to be optimized.

[0294] Then it is called Dominate In the Pareto optimal solution set, no solution is completely dominated by other feasible solutions.

[0295] Finally, candidate solutions are selected from the Pareto optimal solution set according to the engineering design requirements, and then substituted back into the high-fidelity finite element analysis model for verification.

[0296]

[0297] in, The rotor performance response results obtained by high-fidelity finite element analysis model calculation of the final candidate design scheme include rotor mass, maximum equivalent stress, critical speed, maximum radial deformation, contact pressure and thermal safety related responses; This represents a high-fidelity analysis model. This represents the final candidate design scheme. If the back-substitution results meet the requirements for strength safety, critical speed margin, maximum radial deformation, contact state, thermal safety, and manufacturing feasibility, then this scheme is determined as the final optimized rotor structure scheme for the high-speed permanent magnet generator for aviation.

[0298] The above embodiments are merely preferred embodiments provided to fully illustrate the present invention, and the scope of protection of the present invention is not limited thereto. Equivalent substitutions or modifications made by those skilled in the art based on the present invention are all within the scope of protection of the present invention. The scope of protection of the present invention is defined by the claims.

Claims

1. A rotor structure for a high-speed permanent magnet generator for aircraft, comprising a shaft (1), a surface-mounted permanent magnet (2) disposed on the outside of the shaft (1), and a sheath (3) sleeved on the outside of the surface-mounted permanent magnet (2), characterized in that: The rotating shaft (1) has an axially extending internal cavity (15) inside; multiple weight-reducing cavities (11) are also provided on the radially outer side of the internal cavity (15) and inside the rotating shaft (1); each weight-reducing cavity (11) extends axially, and the multiple weight-reducing cavities (11) are distributed circumferentially around the internal cavity (15); between two adjacent weight-reducing cavities (11), there is a bridging area (14) formed by the material that has not been removed, and the bridging area (14) prevents the two adjacent weight-reducing cavities (11) from communicating; the two ends of the weight-reducing cavities (11) are hollowed out so that each weight-reducing cavity (11) is connected at the two ends respectively; both ends of each weight-reducing cavity (11) are open in the axial direction, so that each weight-reducing cavity (11) is connected to each other at the two ends in the axial direction; The rotating shaft (1) is also provided with an air guide (12) and a thin-walled partition (13). The air guide (12) connects the internal cavity (15) to each weight-reducing cavity (11). The thin-walled partition (13) is provided in the internal cavity (15) and / or the weight-reducing cavity (11) to separate and guide the airflow inside the rotating shaft (1).

2. The rotor structure of a high-speed permanent magnet generator for aircraft according to claim 1, characterized in that: The rotating shaft (1) is also provided with end hollowing structures (16) at both ends of the axial direction. In the end hollowing structure (16), the connecting area (14) between adjacent weight-reducing cavities (11) is removed, so that each weight-reducing cavity (11) is connected to each other at the axial end, forming an annular connected area around the inner cavity (15).

3. The rotor structure of a high-speed permanent magnet generator for aviation as described in claim 2, characterized in that: The air guide (12) is an opening on the radial side wall of the internal cavity (15), and the air guide (12) is located at the end of the hollowed-out area of ​​the rotating shaft (1), so that the internal cavity (15) is connected to the annular connecting area through the air guide (12), and then connected to each weight reduction cavity (11).

4. The rotor structure of a high-speed permanent magnet generator for aviation as described in claim 2, characterized in that: The thin-walled partition (13) prevents the cooling airflow from forming an ineffective backflow or short-path circulation in the local area, and guides the cooling airflow into the weight reduction cavity (11) through the air guide (12), so that the cooling airflow flows along a preset path in the weight reduction cavity (11), and after heat exchange, it flows out from the axial end of the weight reduction cavity (11), enters the annular connecting area, and then flows to the air gap area on the outer side of the outer surface of the sheath (3).

5. The rotor structure of a high-speed permanent magnet generator for aviation according to claim 1, characterized in that: The bridging region (14) includes a local material retention region (142) and a general weight reduction bridging region (141); the local material retention region (142) is larger than the general weight reduction bridging region (141) in terms of the thickness between two adjacent weight reduction cavities (11) in the circumferential direction and / or in terms of the retention width in the radial direction of the rotating shaft (1); the local material retention region (142) is located at at least one of the corresponding positions in the magnetic flux dense region, the key region of magnetic return and the main stress transmission path region, and the other regions are general weight reduction bridging regions (141); The local material retention area (142) extends radially from the side near the surface-mounted permanent magnet (2) toward the center of the rotating shaft (1), or is locally widened between adjacent weight-reducing cavities (11) along the circumference of the rotating shaft (1), thereby forming a non-uniform thickness bridge structure.

6. The rotor structure of a high-speed permanent magnet generator for aviation as described in claim 1, characterized in that: The cross-sectional profile of the weight-reducing cavity (11) is formed by five convex arc segments connected end to end in the circumferential direction. All the convex arc segments have the same arc radius. The connection between two adjacent convex arc segments naturally forms a concave transition area. Each convex arc segment protrudes in a direction away from the center of the five-petal flower shape, and each concave transition area is concave in a direction towards the center of the five-petal flower shape. The cross-sectional profile of the weight-reducing cavity (11) is a rotationally symmetrical five-petal flower-shaped structure. When arranged in the rotating shaft (1), one axis of symmetry of the five-petal flower-shaped structure is set along the radial direction of the rotating shaft (1). Along the radially set axis of symmetry, the end away from the center of the rotating shaft (1) corresponds to the highest point of the convex arc segment, and the end close to the center of the rotating shaft (1) corresponds to the lowest point of the concave transition zone opposite to the convex arc segment.

7. The rotor structure of the high-speed permanent magnet generator for aviation according to claim 1, characterized in that, It also includes a magnetic shielding ring (4) and a retaining ring (5); the magnetic shielding ring (4) is disposed in the axial end region of the surface-mount permanent magnet (2); the retaining ring (5) is disposed in the axial end of the surface-mount permanent magnet (2) and / or the sheath (3).

8. An optimized design method for the rotor structure of an aerospace high-speed permanent magnet generator as described in any one of claims 1-7, characterized in that: S1. Based on the overall design specifications and preliminary electromagnetic design results of the high-speed permanent magnet generator for aviation, determine the combination of reference rotor dimensions; S2. Based on the combination of reference rotor dimensions, establish a parametric rotor model and use some structural dimensions in the rotor structure as initial design variables; S3. Establish a high-fidelity analysis model, calculate and extract multiple performance response data of the rotor structure as performance evaluation indicators; S4. Conduct sensitivity analysis on the initial design variables to identify key design variables that have a significant impact on performance evaluation indicators; S5. Design experiments based on key design variables, establish surrogate models, and verify the prediction accuracy of surrogate models; S6. Introduce uncertainty factors, conduct uncertainty propagation analysis based on the surrogate model, and evaluate the robustness and reliability of the rotor structure; S7. With the optimization objectives of minimizing rotor mass, minimizing maximum equivalent stress, and maximizing critical speed, multi-objective optimization is performed on the basis of satisfying the constraints to obtain the final optimized scheme.

9. The optimized design method for the rotor structure of a high-speed permanent magnet generator for aviation according to claim 8, characterized in that: The parameterized rotor model includes at least one of the following: shaft (1), surface-mounted permanent magnet (2), sheath (3), weight-reducing cavity (11), air vent (12), thin-walled partition (13), ordinary weight-reducing bridge area (141), local material retention area (142), and end hollowing structure (16); the initial design variables include at least one of the structural dimensions of the weight-reducing cavity (11), thin-walled partition (13), ordinary weight-reducing bridge area (141), local material retention area (142), and end hollowing structure (16); the performance response data includes at least one of the following: rotor mass, maximum equivalent stress, critical speed, maximum radial deformation, and contact pressure.

10. The optimized design method for the rotor structure of a high-speed permanent magnet generator for aviation according to claim 8, characterized in that: The sensitivity analysis in step S4 uses the Morris method; the surrogate model in step S5 uses the Kriging model, and the verification index of the prediction accuracy includes at least one of the coefficient of determination, root mean square error, and maximum relative error. When the prediction accuracy does not meet the preset requirements, supplementary sampling is performed and the surrogate model is updated; the uncertainty factors in step S6 include at least one of geometric manufacturing tolerance, temperature fluctuation, and rotational speed disturbance, and the uncertainty propagation analysis uses the Monte Carlo sampling method; the multi-objective optimization in step S7 uses the NSGA-II algorithm; the constraints include at least one of strength constraints, critical speed margin constraints, maximum radial deformation constraints, contact pressure constraints, and manufacturing feasibility constraints.

Images