Thin-wall blade surface strengthening and finishing integrated manufacturing method and system

Abstract

The invention relates to the technical field of aero-engine blade surface strengthening, and discloses a thin-wall blade surface strengthening and finishing integrated manufacturing method and a thin-wall blade surface strengthening and finishing integrated manufacturing system. And acquiring information such as an impact force signal, an impact speed signal, a stroke, an impact width, real-time position information of an impact point, surface roughness after impact, residual stress, impact deformation and the like. And a mapping model between real-time position information of the impact force signal, the impact speed signal, the stroke, the impact width and the impact point and the surface roughness, the residual stress and the impact deformation is constructed, and the surface roughness, the residual stress, the deformation result and the like of the blade are monitored in real time. And finally, the strengthening and finishing impact parameter range value of the blade to be machined is obtained through analysis, high controllable impact in the strengthening and finishing process of the blade is ensured, integrated control over strengthening and finishing of the surface of the thin-wall blade is achieved, and the bottleneck problems that in the traditional blade machining process, controllability is insufficient, and machining quality consistency is poor are solved.

Description

Technical Field

[0001] The invention relates to the technical field of aero-engine blade surface strengthening, and discloses an integrated manufacturing method and system for thin-wall blade surface strengthening and finishing, so as to simultaneously improve the blade strengthening effect and reduce the blade deformation. Background Art

[0002] Surface strengthening of aircraft engine blades is an important process to improve the fatigue life of aircraft engine blades. At present, the main surface strengthening processes include mechanical shot peening, laser shot peening, ultrasonic rolling, high-frequency micro-forging, etc. The above surface strengthening processes are to perform periodic or non-periodic impacts on the blade surface to cause plastic deformation on the surface of the blade parts, thereby introducing residual compressive stress. The introduction of residual compressive stress can improve the fatigue life of aircraft engine blade parts.

[0003] However, no matter which surface strengthening process is used, residual compressive stress will be introduced and the surface roughness of the parts will also change. Especially when the surface impact strength is too high, it will greatly increase the surface roughness, cause surface delamination or cracking, and thus cause deformation of the blade or excessive contour accuracy and other adverse factors, reducing the surface strengthening effect of the blade.

[0004] In addition, blade parts must not only be surface strengthened, but also surface finished. There is a significant mutual influence between the processing effects of surface finishing and surface strengthening processes: if the process route of first finishing and then strengthening is adopted, the surface roughness of the finished parts will be greatly increased, affecting their performance and service life; if the process route of first strengthening and then finishing is adopted, part of the strengthening layer on the surface of the parts will be removed, affecting their fatigue resistance. Summary of the invention

[0005] The purpose of the present invention is to provide an integrated manufacturing method and system for thin-walled blade surface strengthening and finishing, and to realize integrated manufacturing of aircraft engine blade surface finishing and strengthening by using high-frequency micro-forging process. It is intended to monitor the surface integrity and contour accuracy of the aircraft engine blade surface after high-frequency micro-forging process, such as surface roughness and residual stress, and optimize the high-frequency micro-forging parameters according to the monitoring results. By proposing a data-driven method based on the mutual fusion of micro-forging impact force, stroke and machine tool signals to positively monitor the evolution law of the surface integrity and accuracy of the blade and reversely optimize the high-frequency micro-forging process parameters, adaptive control of the high-frequency micro-forging process is realized to improve the surface integrity of the blade while ensuring that the blade contour accuracy does not exceed the tolerance, thereby realizing the integrated control of thin-walled blade surface strengthening and finishing, and solving the bottleneck problems such as insufficient controllability of the traditional blade processing process and poor consistency of processing quality.

[0006] In order to achieve the above technical effects, the technical solution adopted by the present invention is:

[0007] An integrated manufacturing method for thin-wall blade surface strengthening and finishing, comprising:

[0008] Step 1, fixing the test blades on the CNC machine tool in turn, and using the impact component integrated in the CNC machine tool to periodically impact the surface of each test blade with different impact parameters to cause plastic deformation on the surface of the test blade; real-time monitoring of the impact force signal, impact velocity signal, stroke, impact width, and real-time position information of the impact point during the impact process, and detecting the blade surface roughness, residual stress, and impact deformation information after each test blade is impacted;

[0009] Step 2: Using the impact force signal, impact velocity signal, stroke, impact width, and real-time position information of the impact point obtained in step 1 as input, and the surface roughness, residual stress, and blade profile accuracy of the blade after the impact is completed as label information, a monitoring model is constructed in a multi-task learning manner, and the constructed monitoring model is trained and tested to obtain a monitoring model that can simultaneously output the blade surface roughness, blade residual stress, and blade profile accuracy;

[0010] Step 3, obtaining the designed surface roughness, designed residual stress, and designed contour accuracy data required for the blade to be processed, and obtaining an impact control parameter combination that meets the designed surface roughness, designed residual stress, and designed contour accuracy data according to the monitoring model analysis, wherein the impact control parameter combination includes impact force, impact velocity, stroke, and impact width;

[0011] Step 4: Use the obtained impact control parameter combination to control the impact component of the CNC machine tool to perform strengthening and finishing processing on the blade to be processed.

[0012] Furthermore, in the step one, a high-frequency micro-forging process is used to carry out an enhanced impact test on the blade surface. During the impact process, a full-bridge strain impact force sensor is used to obtain the axial impact force and radial impact force of the punch rod, and a laser displacement sensor is used to obtain the instantaneous displacement of the punch rod. The time difference is performed based on the instantaneous displacement data to obtain the impact velocity of the punch rod; and a laser displacement sensor is used to obtain the real-time position information of each impact point during the impact process.

[0013] Furthermore, the collected impact force signal, impact velocity signal, stroke, impact width, and real-time position information of the impact point are transposed into a matrix form, and multiple serial convolutional neural networks are designed to extract data features from the input matrix information; a multi-task learning method is used to simultaneously output the blade surface roughness, blade residual stress, and blade contour accuracy.

[0014] Furthermore, in step 4, during the impact strengthening process of the blade to be processed, the impact parameters during the impact process are obtained in real time, and the monitoring model is used to output the surface roughness, residual stress and contour accuracy, and compared with the corresponding design surface roughness, design residual stress, and design contour accuracy data. If the data comparison results are equal, the impact is completed, otherwise the impact parameters are iteratively adjusted until they are equal.

[0015] In order to achieve the above technical effects, the present invention also provides an integrated manufacturing system for thin-walled blade surface strengthening and finishing, comprising:

[0016] An impact assembly, which is integrated in a numerically controlled machine tool and is used to periodically impact or collide the surface of a blade fixed on the numerically controlled machine tool to cause plastic deformation of the blade surface;

[0017] Data acquisition module, used for the impact force signal, impact velocity signal, stroke, impact width, and real-time position information of the impact point during the impact process;

[0018] A data acquisition module, used to obtain the surface roughness detection value and residual stress detection value of the blade after the impact is completed;

[0019] A model building module is used to take the impact force signal, impact velocity signal, stroke, impact width, and real-time position information of the impact point as input, and the surface roughness, residual stress, and blade profile accuracy of the blade after the impact is completed as label information, to build a monitoring model in a multi-task learning manner, train and test the built monitoring model, and obtain a monitoring model that can simultaneously output the blade surface roughness, blade residual stress, and blade profile accuracy;

[0020] An analysis module, used for analyzing and obtaining an impact control parameter combination that satisfies the designed surface roughness, designed residual stress, and designed contour accuracy data according to the monitoring model, wherein the impact control parameter combination includes impact force, impact velocity, stroke, and impact width;

[0021] The processing control module is used to use the obtained impact control parameter combination to control the impact component of the numerical control machine tool to perform strengthening and finishing processing on the blade to be processed.

[0022] Furthermore, in the data acquisition module, a high-frequency micro-forging process is used to carry out an enhanced impact test on the blade surface. During the impact process, a full-bridge strain impact force sensor is used to obtain the axial impact force and radial impact force of the punch rod, and a laser displacement sensor is used to obtain the instantaneous displacement of the punch rod. The time difference is performed based on the instantaneous displacement data to obtain the impact velocity of the punch rod; a laser displacement sensor is used to obtain the real-time position information of each impact point during the impact process.

[0023] Furthermore, in the model building module, the collected impact force signal, impact velocity signal, stroke, impact width, and real-time position information of the impact point are transposed into a matrix form, and multiple serial convolutional neural networks are designed to extract data features from the input matrix information; a multi-task learning method is used to simultaneously output the blade surface roughness, blade residual stress, and blade contour accuracy.

[0024] Compared with the prior art, the beneficial effects of the present invention are as follows: the present invention uses the impact force signal, impact velocity signal and real-time position information of blade parts collected in real time during the high-frequency micro-forging process as input and transposes the input data to make it a matrix form, and uses multiple serial convolutional neural networks to extract data features, so that the model can learn all the input information, greatly improving the monitoring accuracy of the model. Secondly, based on the multi-task learning method, multiple parallel fully connected neural networks are designed to simultaneously output the surface roughness, residual stress and contour accuracy results of the blade after impact, so that the blade surface roughness, blade residual stress, blade contour accuracy and other results can be output simultaneously under one model, greatly improving the calculation time of the model. The model, such as Figure 3 shown.

[0025] At the same time, the blade surface roughness, blade residual stress and blade contour accuracy obtained by the model are compared with the values ​​required by the design. If they are not equal, the impact parameters are adjusted iteratively until the values ​​obtained in step two are equal to or close to the values ​​required by the design. In this way, the process parameters are used as the impact parameters for the final blade surface strengthening. This ensures that the surface roughness and residual stress are improved at the same time, realizes the integrated manufacturing of blade surface strengthening and finishing, realizes the integrated control of thin-walled blade surface strengthening and finishing, and solves the bottleneck problems such as insufficient controllability of traditional blade processing and poor consistency of processing quality. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] Figure 1 It is a flow chart of the integrated manufacturing method of thin-wall blade surface strengthening and finishing in the embodiment;

[0027] Figure 2 It is a block diagram of the integrated manufacturing system for thin-walled blade surface strengthening and finishing in the embodiment;

[0028] Figure 3 Schematic diagram of the monitoring model in the embodiment;

[0029] Figure 4 It is a schematic diagram of the depth of the residual stress field after the optimization of the strengthening and finishing of the blade to be processed in the embodiment;

[0030] Figure 5Schematic diagram of surface roughness measurement results before and after optimization of strengthening and finishing adaptive control of the blade to be processed in the embodiment;

[0031] Figure 6 It is a schematic diagram of the measurement results of the profiles of the four sections of the blade to be processed before and after the adaptive control optimization of strengthening and finishing in the embodiment;

[0032] Among them, 1. impact component; 2. data collection module; 3. data acquisition module; 4. model building module; 5. analysis module; 6. processing control module. DETAILED DESCRIPTION

[0033] The present invention is further described in detail below in conjunction with the embodiments and drawings. However, it should not be understood that the scope of the above subject matter of the present invention is limited to the following embodiments, and all technologies realized based on the content of the present invention belong to the scope of the present invention.

[0034] The present invention is further described below in conjunction with specific embodiments.

[0035] Example

[0036] See also Figure 1-Figure 6 , according to Figure 1 The surface strengthening and finishing process of the aircraft engine blade shown in the figure is implemented. The high-temperature alloy blade of a certain aircraft engine compressor is used as the verification object to carry out high-frequency micro-forging process tests. The surface roughness and blade accuracy indicators in the surface integrity of the high-temperature alloy blade are taken as the target to carry out specific case implementation. The specific steps include:

[0037] Step 1, fixing the test blades on the CNC machine tool in turn, and using the impact component integrated in the CNC machine tool to periodically impact the surface of each test blade with different impact parameters to cause plastic deformation on the surface of the test blade; real-time monitoring of the impact force signal, impact velocity signal, stroke, impact width, and real-time position information of the impact point during the impact process, and detecting the blade surface roughness, residual stress, and impact deformation information after each test blade is impacted;

[0038] Step 2: Using the impact force signal, impact velocity signal, stroke, impact width, and real-time position information of the impact point obtained in step 1 as input, and the surface roughness, residual stress, and blade profile accuracy of the blade after the impact is completed as label information, a monitoring model is constructed in a multi-task learning manner, and the constructed monitoring model is trained and tested to obtain a monitoring model that can simultaneously output the blade surface roughness, blade residual stress, and blade profile accuracy;

[0039] Step 3, obtaining the designed surface roughness, designed residual stress, and designed contour accuracy data required for the blade to be processed, and obtaining an impact control parameter combination that meets the designed surface roughness, designed residual stress, and designed contour accuracy data according to the monitoring model analysis, wherein the impact control parameter combination includes impact force, impact velocity, stroke, and impact width;

[0040] Step 4: Use the obtained impact control parameter combination to control the impact component of the CNC machine tool to perform strengthening and finishing processing on the blade to be processed.

[0041] In this embodiment, the micro-forging impact stroke is 0.2mm, the step distance of the micro-forging impact is set to 0.1m, the impact frequency is set to 200Hz, the feed speed of the machine tool is 1200mm / min, and high-frequency micro-forging is performed in the direction perpendicular to the tangent line of the outer contour of the blade. After multiple adaptive control optimizations using the monitoring model, the surface roughness of the blade and the accuracy of the blade have been significantly improved. The residual stress field depth after the adaptive control optimization of strengthening and finishing of the blade to be processed, the surface roughness measurement results before and after the adaptive control optimization of strengthening and finishing, and the measurement results of the profiles of the four sections are shown in the figure. Figure 4-Figure 6 As shown in the figure, the accuracy of the blade profiles has been improved from about 0.5 μm to about 0.2 μm, and the blade profile accuracy deviation has also been significantly reduced. Figure 4 It can be seen from the figure that the present invention is very effective and can effectively improve the surface roughness and profile accuracy during the high-frequency micro-forging impact process.

[0042] Based on the same inventive concept, this embodiment also provides an integrated manufacturing system for thin-walled blade surface strengthening and finishing, including:

[0043] An impact assembly 1, which is integrated in a CNC machine tool and is used to periodically impact or collide the surface of a blade fixed on the CNC machine tool to cause plastic deformation of the blade surface;

[0044] Data acquisition module 2, used for the impact force signal, impact velocity signal, stroke, impact width, and real-time position information of the impact point during the impact process;

[0045] Data acquisition module 3, used to obtain the surface roughness detection value and residual stress detection value of the blade after the impact is completed;

[0046] Model building module 4 is used to take the impact force signal, impact velocity signal, stroke, impact width, and real-time position information of the impact point as input, and the surface roughness, residual stress, and blade profile accuracy of the blade after the impact is completed as label information, to build a monitoring model in a multi-task learning manner, train and test the built monitoring model, and obtain a monitoring model that can simultaneously output the blade surface roughness, blade residual stress, and blade profile accuracy;

[0047] An analysis module 5 is used to analyze and obtain an impact control parameter combination that satisfies the designed surface roughness, designed residual stress, and designed contour accuracy data according to the monitoring model, wherein the impact control parameter combination includes impact force, impact velocity, stroke, and impact width;

[0048] The processing control module 6 is used to use the obtained impact control parameter combination to control the impact component of the numerical control machine tool to perform strengthening and finishing processing on the blade to be processed.

[0049] The above-mentioned specific implementation can be partially adjusted in different ways by those skilled in the art without departing from the principle and purpose of the present invention. The protection scope of the present invention shall be based on the claims and shall not be limited by the above-mentioned specific implementation. Each implementation scheme within its scope shall be subject to the constraints of the present invention.

Claims

1. An integrated manufacturing method for thin-walled blade surface strengthening and finishing, characterized in that: include: Step 1, fixing the test blades on the CNC machine tool in turn, and using the impact component integrated in the CNC machine tool to periodically impact the surface of each test blade with different impact parameters to cause plastic deformation on the surface of the test blade; real-time monitoring of the impact force signal, impact velocity signal, stroke, impact width, and real-time position information of the impact point during the impact process, and detecting the blade surface roughness, residual stress, and impact deformation information after each test blade is impacted; Step 2: Using the impact force signal, impact velocity signal, stroke, impact width, and real-time position information of the impact point obtained in step 1 as input, and the surface roughness, residual stress, and blade profile accuracy of the blade after the impact is completed as label information, a monitoring model is constructed in a multi-task learning manner, and the constructed monitoring model is trained and tested to obtain a monitoring model that can simultaneously output the blade surface roughness, blade residual stress, and blade profile accuracy; Step 3, obtaining the designed surface roughness, designed residual stress, and designed contour accuracy data required for the blade to be processed, and obtaining an impact control parameter combination that meets the designed surface roughness, designed residual stress, and designed contour accuracy data according to the monitoring model analysis, wherein the impact control parameter combination includes impact force, impact velocity, stroke, and impact width; Step 4: Use the obtained impact control parameter combination to control the impact component of the CNC machine tool to perform strengthening and finishing processing on the blade to be processed.

2. The integrated manufacturing method for thin-walled blade surface strengthening and finishing according to claim 1, characterized in that: In the step one, a high-frequency micro-forging process is used to carry out an enhanced impact test on the blade surface. During the impact process, a full-bridge strain impact force sensor is used to obtain the axial impact force and radial impact force of the punch rod, and a laser displacement sensor is used to obtain the instantaneous displacement of the punch rod. The time difference is performed based on the instantaneous displacement data to obtain the impact velocity of the punch rod; and a laser displacement sensor is used to obtain the real-time position information of each impact point during the impact process.

3. The integrated manufacturing method for thin-walled blade surface strengthening and finishing according to claim 1, characterized in that: The collected impact force signal, impact velocity signal, stroke, impact width, and real-time position information of the impact point are transposed to form a matrix, and multiple serial convolutional neural networks are designed to extract data features from the input matrix information; The blade surface roughness, blade residual stress and blade profile accuracy are output simultaneously using a multi-task learning approach.

4. The integrated manufacturing method for thin-walled blade surface strengthening and finishing according to claim 1, characterized in that: In step 4, during the impact strengthening process of the blade to be processed, the impact parameters during the impact process are obtained in real time, and the monitoring model is used to output the surface roughness, residual stress and contour accuracy, and compared with the corresponding design surface roughness, design residual stress, and design contour accuracy data. If the data comparison results are equal, the impact is completed, otherwise the impact parameters are iteratively adjusted until they are equal.

5. An integrated manufacturing system for thin-walled blade surface strengthening and finishing, characterized in that: include: An impact assembly, which is integrated in a numerically controlled machine tool and is used to periodically impact or collide the surface of a blade fixed on the numerically controlled machine tool to cause plastic deformation of the blade surface; Data acquisition module, used for the impact force signal, impact velocity signal, stroke, impact width, and real-time position information of the impact point during the impact process; A data acquisition module, used to obtain the surface roughness detection value and residual stress detection value of the blade after the impact is completed; A model building module is used to take the impact force signal, impact velocity signal, stroke, impact width, and real-time position information of the impact point as input, and the surface roughness, residual stress, and blade profile accuracy of the blade after the impact is completed as label information, to build a monitoring model in a multi-task learning manner, train and test the built monitoring model, and obtain a monitoring model that can simultaneously output the blade surface roughness, blade residual stress, and blade profile accuracy; An analysis module, used for analyzing and obtaining an impact control parameter combination that satisfies the designed surface roughness, designed residual stress, and designed contour accuracy data according to the monitoring model, wherein the impact control parameter combination includes impact force, impact velocity, stroke, and impact width; The processing control module is used to use the obtained impact control parameter combination to control the impact component of the numerical control machine tool to perform strengthening and finishing processing on the blade to be processed.

6. The integrated manufacturing system for thin-walled blade surface strengthening and finishing according to claim 5, characterized in that: In the data acquisition module, a high-frequency micro-forging process is used to carry out an enhanced impact test on the blade surface. During the impact process, a full-bridge strain impact force sensor is used to obtain the axial impact force and radial impact force of the punch rod, and a laser displacement sensor is used to obtain the instantaneous displacement of the punch rod. The time difference is performed based on the instantaneous displacement data to obtain the impact velocity of the punch rod; and a laser displacement sensor is used to obtain the real-time position information of each impact point during the impact process.

7. The integrated manufacturing system for thin-walled blade surface strengthening and finishing according to claim 5, characterized in that: In the model building module, the collected impact force signal, impact velocity signal, stroke, impact width, and real-time position information of the impact point are transposed to form a matrix, and multiple serial convolutional neural networks are designed to extract data features from the input matrix information; The blade surface roughness, blade residual stress and blade profile accuracy are output simultaneously using a multi-task learning approach.

Images