A method for planning the processing technology of near-net-shape ceramic matrix composite blades
By designing feature planes and reverse modeling on ceramic matrix composite blades, a stable machining coordinate system was established, solving the problems of blade blank allowance fluctuation and unclear datum. This enabled high-precision CNC machining and improved the manufacturing level and quality of the blades.
Patent Information
- Application Number
- CN202511630564.8
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2025-11-10
- Publication Date
- 2026-01-30
- Estimated Expiration
- 2045-11-10
AI Technical Summary
Existing technologies lack systematic and efficient process planning methods for the processing of near-net-shape ceramic matrix composite blades, making it difficult to solve problems such as large fluctuations in blank allowance and unclear benchmarks, resulting in low processing accuracy and yield.
By designing at least three mutually perpendicular feature planes as machining references, and combining line laser measurement and reverse modeling, a stable machining coordinate system is established. CNC precision machining technology is then used to achieve precise positioning of the blade blank and uniform distribution of allowance.
It provides a general process planning method that can be adapted to the manufacturing of ceramic matrix composite blades of different types and specifications, ensuring the accuracy of the machining coordinate system and the consistency of the machining process, and significantly improving machining accuracy and finished product qualification rate.
Smart Images

Figure CN121083406B_ABST
Abstract
Description
Technical Field
[0001] This invention belongs to the field of hard and brittle material processing, specifically relating to a method for planning the processing technology of near-net-shape ceramic matrix composite blades. Background Technology
[0002] Because blades typically have thin walls, irregular shapes, and complex structures, and are often made of difficult-to-machine materials such as high-temperature alloys and titanium alloys, the requirements for machining accuracy, surface integrity, and geometric tolerances are extremely stringent. Therefore, achieving efficient, precise, and high-quality machining of blade-like parts has always been a significant technical challenge in the aerospace manufacturing field.
[0003] To address these challenges, traditional blade manufacturing typically employs advanced technologies such as high-speed cutting, multi-axis CNC machining, efficient cooling and lubrication, and in-machine measurement. High-speed cutting effectively reduces cutting forces, suppresses workpiece deformation, and improves surface quality; efficient cooling technology helps control cutting temperature and extends tool life; and precision measurement ensures that machining results meet design tolerance requirements. Although these technologies have improved the machining quality and efficiency of blades to some extent, problems still exist when dealing with high-temperature alloys, titanium alloys, and other tough materials, including rapid tool wear, low machining efficiency, and difficulty in guaranteeing surface integrity.
[0004] Ceramic matrix composites are inherently hard and brittle, prone to cracking and subsurface damage during processing, placing higher demands on machining techniques. To improve material utilization and reduce subsequent processing costs, ceramic matrix composite blades are often manufactured using near-net-shape processes, resulting in smaller and unevenly distributed machining allowances in the blank, further increasing the difficulty of CNC machining. In particular, the lack of a stable and reliable machining datum makes it difficult to achieve precise positioning and uniform cutting of allowances on CNC machine tools, severely affecting the blade forming accuracy and yield rate.
[0005] In summary, existing technologies still lack systematic and efficient process planning methods for the machining of near-net-shape ceramic matrix composite blades, making it difficult to effectively address key issues such as large fluctuations in blank allowance and unclear benchmarks. Therefore, there is an urgent need to research an efficient and precise machining process suitable for this type of blade to improve the manufacturing level and product quality of aero-engine blades. Summary of the Invention
[0006] To address the problems in the prior art, this invention provides a method for planning the processing technology of near-net-shape ceramic matrix composite blades.
[0007] The technical solution of this invention is:
[0008] A method for planning the processing technology of near-net-shape ceramic matrix composite blades includes the following steps:
[0009] Rough machining of blade blanks: Use a thin diamond grinding wheel to remove excess material from the blade blanks in certain areas, and control the machining allowance of each surface to be within 0.5mm-1mm; if the machining allowance of each surface of the blade blank is <0.5mm, the rough machining step of the blade blanks is omitted.
[0010] Design feature planes: Design at least 3 mutually perpendicular feature planes. The feature planes are selected at positions of existing planar features in the original model, 10 mm away from the edge of the blade blank edge plate, for the establishment of the finishing coordinate system.
[0011] Measuring blade blanks: After removing the blade blanks from the machine tool, high-precision measurement model data is obtained through a line laser measurement platform, including the surface contour and feature plane data of the shaped ceramic matrix composite blade blanks to be processed. The measurement accuracy is more than 5 times higher than the blade processing accuracy.
[0012] Reverse modeling and registration: A reverse blade blank model is established based on the measurement model data, and the blade theoretical model is imported for registration to achieve uniform distribution of the margin except for the feature plane. The ideal feature plane is then fitted using a plane fitting algorithm based on the feature plane data, and the model is then exported.
[0013] Precision machining of blade blanks: Based on the feature plane, the machining coordinate system is set, the exported model is imported into the CNC software, NC code is generated, and CNC precision machining is performed;
[0014] For blade blanks that cannot be machined in one clamping, they need to be machined in parts. Three mutually perpendicular feature planes are set for each part of the blade blank, the finishing coordinate system of each part is determined, and the blade blank is measured, reverse modeled and registered, and the blade blank is finished.
[0015] Furthermore, the above-mentioned near-net-shape ceramic matrix composite blade processing planning method derives a model that includes a blade blank model, a blade theoretical model, and a fitted ideal feature plane in the same coordinate system.
[0016] Furthermore, in the above-mentioned near-net-shape ceramic matrix composite blade processing planning method, the feature plane design includes three mutually perpendicular reference feature planes, and the remaining feature planes need to be parallel to the reference feature planes to assist in verifying tool setting accuracy or confirming tool setting points.
[0017] Furthermore, in the aforementioned near-net-shape ceramic matrix composite blade processing planning method, the three mutually perpendicular reference feature planes in the processing coordinate system based on the feature planes are the XOY plane, the XOZ plane, and the YOZ plane.
[0018] Furthermore, in the above-mentioned near-net-shape ceramic matrix composite blade processing planning method, the blade blank model containing the processing allowance of each part is imported into the CNC software after registration. The scanned blade blank model is set as the blade blank before processing, the blade theoretical model is set as the blade model after processing, and the three ideal feature planes are set as XOY plane, XOZ plane and YOZ plane respectively.
[0019] Advantages and beneficial effects of the present invention:
[0020] 1. The method described in this invention is highly versatile and does not rely on specific blade shapes or blade blanks. Whether it is a blade blank with a large allowance that requires rough machining first, a near-net-shape blade blank that can be directly finished, or a complex blade that requires multi-stage processing, this method can adapt by adjusting the layout and registration strategy of the feature planes. This high degree of flexibility enables it to be widely applied to the manufacture of ceramic matrix composite blades of different types and specifications.
[0021] 2. This invention provides a stable and reliable physical reference for subsequent finishing processes by actively designing and machining at least three mutually perpendicular feature planes on the blade blank. This method fundamentally solves the industry problem of near-net-shape blade blanks being difficult to accurately position and align on CNC machine tools due to the lack of regular geometric features, ensuring the accurate establishment of the machining coordinate system and laying a solid foundation for the final realization of high-precision surface machining.
[0022] 3. The reverse modeling process of this invention can be performed manually or by writing a registration program for precise registration, depending on the needs. Through the technical path of "measuring the blade blank - reverse modeling and registration," the actual three-dimensional state of the blade blank is accurately obtained. The measured blade blank model and the theoretical blade model are intelligently registered in the software, automatically planning the optimal tool path. This ensures a uniform and reasonable distribution of cutting allowance even when the blade blank has a small and uneven allowance. This effectively avoids quality problems such as tool overload and breakage due to excessive local allowance, or incomplete machining due to insufficient allowance, significantly improving the consistency of the machining process and the yield rate of the finished product. Attached Figure Description
[0023] Figure 1 This is a schematic diagram of the non-machined surface during the rough machining of the blade blank in Embodiment 1 of the present invention;
[0024] Figure 2 This is a schematic diagram of the machined surface during the rough machining of the blade blank in Embodiment 1 of the present invention;
[0025] Figure 3 This is a schematic diagram of a set of blade blank features in Embodiment 1 of the present invention;
[0026] Figure 4 This is a schematic diagram of the blade blank measurement in Embodiment 1 of the present invention;
[0027] Figure 5 This is a registration diagram of the blade blank model and the blade theoretical model displayed by the CNC software after the first reverse modeling and registration in Embodiment 2 of the present invention;
[0028] Figure 6 This is a schematic diagram of the first group of blade blank features in Embodiment 2 of the present invention;
[0029] Figure 7 This is a schematic diagram of the second group of blade blank features in Embodiment 2 of the present invention;
[0030] Figure 8 This is a schematic diagram of the blade blank measurement in Embodiment 2 of the present invention;
[0031] Figure 9 This is a schematic diagram of the first finishing process in Embodiment 2 of the present invention;
[0032] Figure 10 This is a schematic diagram of the second finishing process in Embodiment 2 of the present invention;
[0033] In the figure, 1-roughing fixture, 2-upper edge plate of blade blank, 3-thin diamond grinding wheel, 4-lower edge plate of blade blank, 5-feature plane A, 6-feature plane B, 7-feature plane C, 8-line laser sensor, 9-theoretical model B of blade, 10-blade blank model B, 11-feature plane D, 12-feature plane E, 13-feature plane F, 14-feature plane G, 15-finishing fixture A, 16-finishing fixture B. Detailed Implementation
[0034] In a specific embodiment of the present invention, the process mainly includes rough machining of the blade blank, designing feature planes, measuring the blade blank, reverse modeling and registration, and finish machining of the blade blank. Next, the specific embodiments of the present invention will be described in further detail with reference to the accompanying drawings and examples. The following examples are for illustrative purposes only and should not be used to limit the scope of the present invention.
[0035] Example 1
[0036] This embodiment presents a near-net-shape ceramic matrix composite blade processing planning method, including the following steps:
[0037] Rough-machined blades: Rough-machined blade blanks, such as... Figure 1 and Figure 2As shown, the blade blank is placed on the roughing fixture 1, and the upper edge plate 2 and lower edge plate 4 of the blade blank are roughed by the thin diamond grinding wheel 3. This step is mainly to remove the local excess material of the near-net-shape blade blank, with a margin of 1mm, in preparation for subsequent precision machining.
[0038] Design feature planes: such as Figure 3 As shown, based on the characteristics of the blade theoretical model, three mutually perpendicular feature planes, A5, B6, and C7, are designed. These are mainly used to determine the machining coordinate system for finishing. In this embodiment, these three feature planes are located at the corner of the part, 10mm away from the edge of the blade blank edge plate, and at the positions where planar features already exist in the original model. This facilitates tool setting and leveling using machine tool probes, dial indicators, etc. Feature planes can also be added at other positions on the blade blank as needed to improve tool setting accuracy or to confirm the accuracy of the tool setting point.
[0039] Measuring blade blanks: such as Figure 4 As shown, the blade blank is removed from the machining tool, and the blade blank in its current state is measured using a line laser sensor 8 to obtain the blade blank measurement model data. The measurement accuracy should be more than 5 times higher than the blade machining accuracy. In this embodiment, the measurement accuracy is 0.01 mm, and the part accuracy requirement is 0.05 mm, which includes the surface contour data and feature plane data of the blade blank.
[0040] Reverse modeling and registration: Based on the blade measurement model data, a reverse blade blank model A is established and imported into the blade theoretical model A for registration, so as to achieve uniform distribution of the allowance except for the feature plane.
[0041] This embodiment employs the Iterated Closest Points (ICP) algorithm. First, coarse registration is performed by manually selecting corresponding points: three or more pairs of feature points (such as corner points and center points) are manually selected on both the reverse blade blank model A and the theoretical blade model A. An initial transformation is calculated, and then fine registration is performed, specifically including:
[0042] Data Association: For each point in the transformed reverse blade blank model A, find the nearest point on the model surface. This "nearest point" can be a vertex on the model mesh or any point on the model surface. Find the point where the normal line intersects the theoretical model by passing through points on the surface of the blade blank model A.
[0043] Calculate the optimal transformation: After obtaining this set of correspondences between "measurement points and the nearest points of the model", calculate a rigid body transformation (R and t), where R refers to the rotation matrix and t refers to the translation vector, which minimizes the sum of squared distances between all corresponding points.
[0044] Apply the transformation: Apply the calculated transformation to the reverse blank model.
[0045] Iteration and Convergence Determination: Repeat the above steps. After each iteration, calculate the average distance or mean square error of all corresponding points. The algorithm stops when the change in this error is less than a certain threshold (0.04 mm in this embodiment) or when the maximum number of iterations is reached.
[0046] Based on the feature plane data, the least squares algorithm (OLS) is used to fit the ideal feature plane, and then the model is derived. The model includes the blade blank model A, the blade theoretical model A, and the fitted ideal feature plane in the same coordinate system.
[0047] Finishing the blade blank: Install the blade blank back into the CNC machine tool. In the machine tool, determine the machining coordinate system based on the feature planes A5, B6, and C7 of the current blade blank. Feature plane A5 is the XOY plane, feature plane B6 is the XOZ plane, and feature plane C7 is the YOZ plane. In the CNC software, import the model generated in the previous step, set the blade blank model A as the blade blank before machining, and set the theoretical blade model A as the blade after machining. Set the three fitted ideal feature planes as the XOY plane, XOZ plane, and YOZ plane, respectively. Generate the machining program and NC code in the CNC software, and import them into the CNC machine tool for finishing the blade blank.
[0048] Example 2
[0049] In this embodiment, the relative distance between the blade blank model and the theoretical blade model is small, and the maximum machining allowance is only 0.4mm. Therefore, rough machining is not required, and the machining process planning method includes the following steps:
[0050] Design feature planes: such as Figure 6 and 7 As shown, based on the characteristics of the blade theoretical model, four feature planes are designed, including feature plane D11, feature plane E12, feature plane F13, and feature plane G14. Feature plane G14 is parallel to feature plane E12, while feature planes D11, E12, and F13 are perpendicular to each other and are used to determine the machining coordinate system for the first finishing process. Feature planes D11, F13, and G14 are perpendicular to each other and are used to determine the machining coordinate system for the second finishing process. These feature planes are all located on the side of the end face of the part, not in the curved area of the part, 10mm away from the edge of the blade blank edge plate, and located at the position of the planar features that already exist in the original model.
[0051] Measuring blade blanks: such as Figure 8As shown, the blade blank is removed from the machining tool, and the blade in its current state is measured using a line laser sensor 8 to obtain the blade blank measurement model data. The measurement accuracy should be more than 5 times higher than the blade machining accuracy. In this embodiment, the part machining accuracy is 0.1 mm and the measurement accuracy is 0.02 mm, which includes the surface contour data of the blade blank and the data of 4 feature planes.
[0052] First reverse modeling and registration: Figure 9 The first finishing process is demonstrated. Due to the small allowance of the blade blank, 1 / 4 of the blank is used for clamping and embedding in the finishing fixture A15, while the remaining 3 / 4 is finished. A reverse blade blank model B10 is established based on the blade blank measurement model data and imported into the blade theoretical model B9 for registration, ensuring a uniform distribution of the machining allowance except for the feature planes. Figure 5 The image shows the registration diagram of the blade blank model B10 and the blade theoretical model B9 in the CNC software after the first reverse modeling and registration.
[0053] This embodiment employs the Iterated Closest Points (ICP) algorithm. First, coarse registration is performed by manually selecting corresponding points: three or more pairs of feature points (such as corner points and center points) are manually selected on the reverse blade blank model B10 and the blade theoretical model B9. An initial transformation is calculated, and then fine registration is performed, specifically including:
[0054] Data Association: For each point in the transformed reverse blade blank model B10, find the nearest point on the model surface. This "nearest point" can be a vertex on the model mesh or any point on the model surface. It is necessary to find the point where the normal line intersects the blade theoretical model B9 by using the points on the surface of the blade blank model B10 to find the point where the normal line intersects the normal line.
[0055] Calculate the optimal transformation: After obtaining this set of correspondences between "measurement points and the nearest points in the model", calculate a rigid body transformation that minimizes the sum of squared distances between all corresponding points.
[0056] Application of transformation: The calculated transformation is applied to the reverse blade blank model B10.
[0057] Iteration and Convergence Determination: Repeat the above steps. After each iteration, calculate the average distance or mean square error of all corresponding points. The algorithm stops when the change in this error is less than a certain threshold (0.04 mm in this embodiment) or when the maximum number of iterations is reached.
[0058] An ideal feature plane is fitted based on the feature plane data obtained from the scan. Then, a model is exported, which includes the blade blank model B10, the blade theoretical model B9, and the fitted ideal feature plane in the same coordinate system.
[0059] First finishing of the blade blank: The blade blank is installed back into the CNC machine tool. In the machine tool, the machining coordinate system is determined according to the feature planes D11, E12, and F13 of the current blade blank. Feature plane D11 is the XOY plane, feature plane E12 is the XOZ plane, and feature plane F13 is the YOZ plane. In the CNC software, the model is imported. The blade blank model B10 is set as the blade blank before machining, and the blade theoretical model B9 is set as the blade after machining. The three fitted ideal feature planes are set as the XOY plane, XOZ plane, and YOZ plane, respectively. The machining program and NC code are generated in the CNC software and imported into the CNC machine tool for finishing of the blade blank.
[0060] Second reverse modeling and registration: Figure 10 The second finishing process is demonstrated, where the unfinished 1 / 4 from the previous step is finished, with 3 / 4 used for clamping and embedded in the finishing fixture B16. The same registration method as the first step is used for re-registration, and an ideal feature plane is fitted based on the feature plane data obtained from the scan. Then, the model is exported, which includes the blade blank model C, the theoretical blade model C, and the fitted ideal feature plane in the same coordinate system.
[0061] Second finishing of the blade blank: The blade blank is installed back into the CNC machine tool. In the machine tool, the machining coordinate system is determined according to the feature planes D11, F13, and G14 of the current blade blank. Feature plane D11 is the XOY plane, feature plane F13 is the XOZ plane, and feature plane G14 is the YOZ plane. In the CNC software, the model is imported, the blade blank model C is set as the blade blank before machining, and the blade theoretical model C is set as the blade after machining. The three ideal feature planes are set as the XOY plane, XOZ plane, and YOZ plane respectively. The machining program and NC code are generated in the CNC software and imported into the CNC machine tool for finishing of the blade blank.
Claims
1. A method for near-net-shape ceramic matrix composite blade process planning, characterized in that, The method comprises the following steps: Coarse machining of the blade blank: thin sheet diamond grinding wheel is used to remove the local excess material of the blade blank, and the excess amount of each surface of the blade blank is controlled to be within 0.5-1 mm; if the machining allowance is less than 0.5 mm, the step of coarse machining of the blade blank is omitted; Design of feature planes: at least three mutually perpendicular feature planes are designed, which are selected at a distance of 10 mm from the edge of the blade blank rim plate and at the positions of the original model existing plane features, and are used for establishing the finishing coordinate system; The feature plane design comprises three mutually perpendicular reference feature planes, and the remaining feature planes need to be parallel to the reference feature planes; Measurement of the blade blank: after the blade blank is removed from the machine tool, high-precision measurement model data are obtained through a line laser measurement platform, which include the surface profile of the near-net-shape ceramic matrix composite blade blank to be machined and feature plane data, and the measurement accuracy is more than 5 times higher than the blade machining accuracy; Reverse modeling and registration: a reverse blade blank model is established according to the measurement model data, and is imported into the blade theoretical model for registration, so as to realize uniform distribution of the allowance except the feature planes, and an ideal feature plane is fitted by using a plane fitting algorithm according to the feature plane data, and then the model is exported; The exported model comprises the blade blank model, the blade theoretical model and the fitted ideal feature plane in the same coordinate system; Finishing machining of the blade blank: the machining coordinate system is set based on the feature planes, the exported model is imported into the numerical control software, NC code is generated, and numerical control finishing machining is performed; In the machining coordinate system set based on the feature planes, the three mutually perpendicular reference feature planes are XOY plane, XOZ plane and YOZ plane; The blade blank model comprising the machining allowance of each part after registration is imported into the numerical control software, the scanned blade blank model before machining is set in the numerical control software, the blade theoretical model is set as the blade after machining, and the three ideal feature planes are set as XOY plane, XOZ plane and YOZ plane; For the blade blank that cannot be clamped and machined in one time, the blade blank needs to be machined in parts, three mutually perpendicular feature planes are set for each part of the blade blank, the finishing machining coordinate system of each part is determined, and the measurement of the blade blank, reverse modeling and registration and finishing machining of the blade blank are completed.
Citation Information
Patent Citations
Adaptive machining method of wide-chord hollow fan blade
CN105739440A
Complex structure blank processing and positioning method based on three-dimensional inverse solution
CN108917600A