Full-process prediction method based on broaching parameter-surface integrity-fretting wear

Abstract

The invention discloses a full-process prediction method based on broaching parameters, surface integrity and fretting wear, and relates to the field of fretting wear. The method comprises the following steps: S1, acquiring multiple groups of surface integrity parameters and fretting wear parameters of a broached material; s2, the broaching parameters as input and the surface integrity parameters as output are substituted into a first Kriging agent model for training to obtain a first model, and the surface integrity parameters as input and the fretting wear parameters as output are substituted into a second Kriging agent model for training to obtain a second model; and S3, associating the first model with the second model to obtain a whole-process prediction model, and predicting the whole-process prediction model. According to the method, the broaching parameters, the surface integrity and the fretting wear parameters are subjected to full-process prediction, the limitation of traditional single parameter mapping is broken through, and full-process interpretability prediction of machining parameters, the surface state and the wear performance is achieved.

Description

Technical Field

[0001] This invention relates to the field of fretting wear, and specifically to a full-process prediction method based on broaching parameters, surface integrity, and fretting wear. Background Technology

[0002] As a critical load-bearing component of the aircraft propulsion system, the tenon and mortise joint of aero-engine turbine disks operates under extreme conditions such as high temperature and high pressure for extended periods. Their manufacturing often employs a precision broaching process using GH4169 nickel-based superalloy. This machining technology is widely used due to its ability to effectively balance machining efficiency and forming accuracy under mass production conditions. It is noteworthy that the coupling effect of the dynamic temperature field and the plastic strain of the material during broaching significantly alters the surface integrity parameters of GH4169. Surface roughness is influenced by broaching parameters, broaching temperature, and various other parameters. While broaching improves surface precision, the heat generated and its impact on the workpiece surface and subsurface alter the workpiece's microhardness and residual stress. The evolution of these surface state parameters directly affects the fretting wear behavior of the turbine disk tenon joint, leading to accelerated wear and reduced lifespan. Therefore, systematically analyzing the mapping relationship between broaching process parameters and surface integrity indices, and establishing a correlation model between this relationship and fretting wear performance, has significant engineering value for ensuring the service performance and lifespan of superalloy components in aero-engines.

[0003] Current academic research has developed multiple technical approaches to address this problem, primarily focusing on surface integrity parameter optimization and machining process optimization. Regarding residual stress control, Telesman et al. revealed the influence of broaching parameters on residual stress distribution using orthogonal experimental methods. In terms of surface morphology, Thakur's team elucidated the mechanism by which cutting speed regulates surface roughness through thermo-mechanical coupling; Ramana et al. achieved optimal surface roughness in Ti-6Al-4V alloy turning by altering tool geometry parameters, thereby increasing wear resistance. Domestic scholars, such as Zhang Dinghua's team, proposed a surface morphology optimization method based on process chain collaborative optimization in the field of complex component machining, while Feng et al. and Liu Zhanqiang et al. systematically characterized the surface hardness characteristics of nickel-based superalloys, including the evolution of work-hardened and white layers. Furthermore, Busch et al.'s research confirmed that high-pressure cooling technology can effectively improve the surface quality of high-temperature alloy cutting. However, existing studies mostly focus on the independent analysis of a single surface integrity parameter. The optimal process parameter configurations for different surface integrity parameters may vary significantly or even conflict with each other, easily leading to a situation where "one is better than the other." There are still obvious limitations in establishing multi-parameter coupling mechanisms and overall evaluation models, and there is a lack of research that comprehensively considers the synergistic effects of multiple surface integrity parameters.

[0004] The process inheritance characteristics of surface integrity give it a dual nature in the process of processing-service performance transfer: it is both the response output of broaching process parameters and a key input affecting fretting wear performance. Li et al. revealed the strengthening mechanism of the wear resistance performance of materials by residual stress amplitude and distribution characteristics through finite element simulation; Ke Li et al. [] explained the influence mechanism of a single surface integrity parameter on fretting wear from the perspective of microstructure evolution; Wen et al. verified the correlation between surface microstructure and wear performance in the study of Ti-45Al alloy. These studies show that constructing a multi-parameter synergistic effect model of surface integrity is the core link in establishing the process-performance mapping relationship. However, existing studies have failed to fully establish a closed-loop correlation of "process parameters-surface integrity-service performance". The processing process lacks a comprehensive evaluation of multi-parameter coupling. When constructing a correlation model between surface integrity and its fretting wear performance, the model's predictability and universality are insufficient due to the incomplete consideration of surface integrity factors. Due to the lack of the above correlation, it is difficult to achieve reverse optimization and precise planning of broaching process parameters with the goal of improving fretting wear performance. Therefore, constructing a correlation model that comprehensively considers the coupling effects of multiple parameters is a key issue that urgently needs in-depth research. Summary of the Invention

[0005] To address at least one of the aforementioned problems, this invention provides a comprehensive prediction method based on broaching parameters, surface integrity, and fretting wear.

[0006] The technical solution of the present invention to solve the above problems is as follows: A comprehensive prediction method based on broaching parameters, surface integrity, and fretting wear includes the following steps: S1. Obtain multiple sets of surface integrity parameters and fretting wear parameters of the broached material; S2. The first model is obtained by substituting the broaching parameters as input and the surface integrity parameters as output into the first Kriging surrogate model for training. The second model is obtained by substituting the surface integrity parameters as input and the fretting wear parameters as output into the second Kriging surrogate model for training. S3. Link the first model and the second model to obtain the full-process prediction model, which can then be used to make predictions.

[0007] One embodiment of the present invention is that the surface integrity parameters include hardness, residual stress and surface roughness, the fretting wear parameters include wear volume, and the broaching parameters include broaching times, measurement position, broaching speed and broaching depth.

[0008] One embodiment of the present invention is that the broaching parameters and surface integrity parameters are obtained based on measured values ​​and / or finite element simulation experiments, and the fretting wear parameters are obtained based on measured values.

[0009] One embodiment of the present invention involves using the Matern Cubic function as the correlation function and the coefficient of determination R during the training process of the first and second Kriging surrogate models. 2 The maximum error (Maximum) is used to evaluate the accuracy of the trained model.

[0010] One embodiment of the present invention includes, in step S2, the following step: performing response surface analysis on the first model, the second model, and the full-process prediction model.

[0011] The beneficial effects of this invention are as follows: This invention performs full-process prediction of broaching parameters, surface integrity, and fretting wear parameters, breaking through the limitations of traditional single-parameter mapping. By introducing surface integrity as a latent variable, it achieves interpretable prediction of the entire process from machining parameters to surface condition to wear performance, providing a reference for process optimization of complex geometric components. Furthermore, this invention achieves high prediction accuracy with only a small sample size. Attached Figure Description

[0012] Figure 1 This is a graph showing the model error. Figure 2 The spatial distribution characteristics of residual stress with broaching speed and depth are shown when the broaching position is 1 mm and the broaching number is 1. Figure 3 The spatial distribution characteristics of residual stress with broaching speed and depth are shown when the broaching position is 1 mm and the broaching number is 2. Figure 4 The spatial distribution characteristics of residual stress with broaching speed and depth are shown when the broaching position is 1 mm and the broaching number is 3. Figure 5 The spatial distribution characteristics of residual stress with broaching speed and depth are shown when the broaching position is 6mm and the broaching number is 1. Figure 6 The spatial distribution characteristics of residual stress with broaching speed and depth are shown when the broaching position is 6mm and the broaching number is 2. Figure 7 The spatial distribution characteristics of residual stress with broaching speed and depth are shown when the broaching position is 1 mm and the broaching number is 3. Figure 8 The spatial distribution characteristics of surface roughness with broaching speed and depth are shown when the broaching position is 1 mm and the broaching number is 1. Figure 9 The spatial distribution characteristics of surface roughness with broaching speed and depth are shown when the broaching position is 1 mm and the broaching number is 2. Figure 10The spatial distribution characteristics of surface roughness with broaching speed and depth are shown when the broaching position is 1 mm and the broaching number is 3. Figure 11 The spatial distribution characteristics of surface roughness with broaching speed and depth are shown when the broaching position is 6 mm and the broaching number is 1. Figure 12 The spatial distribution characteristics of surface roughness with broaching speed and depth are shown when the broaching position is 6 mm and the broaching number is 2. Figure 13 The spatial distribution characteristics of surface roughness with broaching speed and depth are shown when the broaching position is 6 mm and the broaching number is 3. Figure 14 The spatial distribution characteristics of hardness with broaching speed and depth are shown when the broaching position is 1 mm and the broaching number is 1. Figure 15 The spatial distribution characteristics of hardness with broaching speed and depth are shown when the broaching position is 1 mm and the broaching number is 2. Figure 16 The spatial distribution characteristics of hardness with broaching speed and depth are shown when the broaching position is 1 mm and the broaching number is 3. Figure 17 The spatial distribution characteristics of hardness with broaching speed and depth are shown when the broaching position is 6mm and the broaching number is 1. Figure 18 The spatial distribution characteristics of hardness with broaching speed and depth are shown when the broaching position is 6mm and the broaching number is 2. Figure 19 The spatial distribution characteristics of hardness with broaching speed and depth are shown when the broaching position is 6mm and the broaching number is 3. Figure 20 The spatial distribution characteristics of fretting wear as a function of residual stress and surface roughness under a hardness of 560HV are shown. Figure 21 The spatial distribution characteristics of fretting wear as a function of residual stress and surface roughness under a hardness of 640HV are shown. Figure 22 The spatial distribution characteristics of fretting wear as a function of residual stress and hardness under a surface roughness of 0.45 μm are shown. Figure 23 The spatial distribution characteristics of fretting wear as a function of residual stress and hardness under a surface roughness of 0.6 μm are shown. Figure 24 The spatial distribution characteristics of fretting wear as a function of residual stress and hardness under a residual stress of 400 MPa; Figure 25 The spatial distribution characteristics of fretting wear as a function of residual stress and hardness under a residual stress of 700 MPa; Figure 26 This is a characteristic diagram showing the change of wear volume with broaching speed and broaching depth when the broaching position is 1 mm and the broaching number is 1. Figure 27 This is a characteristic diagram showing the change of wear volume with broaching speed and broaching depth when the broaching position is 1 mm and the number of broaching cycles is 2. Figure 28 This is a characteristic diagram showing the change of wear volume with broaching speed and broaching depth when the broaching position is 1 mm and the number of broaching cycles is 3. Figure 29 This is a characteristic diagram showing the change of wear volume with broaching speed and broaching depth when the broaching position is 6mm and the broaching number is 1. Figure 30 This is a characteristic diagram showing the change of wear volume with broaching speed and broaching depth when the broaching position is 6mm and the number of broaching cycles is 2. Figure 31 This is a characteristic diagram showing the change in wear volume with broaching speed and broaching depth when the broaching position is 6mm and the number of broaching cycles is 3. Detailed Implementation

[0013] The specific embodiments of the present invention will now be clearly and completely described with reference to examples and accompanying drawings. Obviously, the described examples are only some embodiments of the present invention, and not all embodiments.

[0014] A comprehensive prediction method based on broaching parameters, surface integrity, and fretting wear includes the following steps: S1. Obtain multiple sets of surface integrity parameters and fretting wear parameters of the broached material; In this embodiment, the surface integrity parameters include hardness, residual stress, and surface roughness; the fretting wear parameters include wear volume; and the broaching parameters include the number of broaching passes, measurement position, broaching speed, and broaching depth. The broaching parameters can be obtained directly from the broaching equipment. For the surface integrity parameters, in this embodiment, the hardness is measured using a KELITI-000ZB ultra-micro Vickers hardness tester (test force 0.49N, holding time 15s), the residual stress is measured using X-ray diffraction (Cr-Kα radiation, diffraction plane {311}), and the surface roughness is measured using a Chotest SuperView W1 white light interferometer. For the fretting wear parameters, in this embodiment, they are also obtained using a Chotest SuperView W1 white light interferometer. Those skilled in the art will understand that this embodiment only provides one method for obtaining the above parameters; those skilled in the art can also use other equipment and methods in the art to obtain the above parameters.

[0015] Specifically, in this embodiment, the measurement position in the broaching parameters refers to the distance from the starting end of the tool feed along the broaching direction. This parameter is considered because during the broaching process, when the tool performs broaching operations along a path, factors such as changes in tool load, increased heat accumulation, and decreased cooling efficiency will change, causing different positions on the broaching path to have a significant impact on the surface integrity of the material.

[0016] In actual implementation, the broaching parameters need to be paired with the surface integrity parameters, and the surface integrity parameters need to be paired with the fretting wear parameters. For ease of description, the broaching parameters and surface integrity parameters are referred to as the first dataset, and the surface integrity parameters and fretting wear parameters are referred to as the second dataset.

[0017] In practice, the number of samples in the first dataset and the second dataset do not need to be the same. The number of samples in the first dataset can be more, less, or equal. According to the inventor's experience, the number of samples in the first dataset is usually set to 100-180, and the number of samples in the second dataset is usually no less than 40. Of course, theoretically, it would be better if the number of samples in the second dataset and the first dataset were equal, but in practice, the second dataset is obtained through experiments, which is relatively costly, so it is only necessary to have more than 40 samples.

[0018] Meanwhile, the methods for obtaining samples in the first dataset and the second dataset can also be different. At the most basic level, both can be obtained through experimental methods. For example, after broaching the material to be tested, such as GH4169 alloy, according to certain parameters using a broaching device, its surface integrity parameters can be directly obtained through the above method. Subsequently, using the broached material to be tested, a tangential fretting wear test device is used to perform a fretting wear test to obtain its fretting wear parameters.

[0019] However, in practice, the cost of GH4169 alloy is high, and the costs of subsequent broaching and fretting wear tests are also high. Therefore, if only measured parameters are used, the cost will rise rapidly. In order to reduce costs, this embodiment also uses finite element simulation experiments to obtain more samples of the first dataset.

[0020] In this embodiment, a finite element simulation model is established in ABAQUS software, such as... Figure 4As shown, the working parameters of the YG8 tool are: rake angle 3°, clearance angle 15°, cutting edge radius 0.05mm, elastic modulus 640GPa, Poisson's ratio 0.22. The parameters of the GH4169 workpiece are: dimensions 6×10×3mm, elastic modulus 206GPa, Poisson's ratio 0.3. The mesh uses C3D8RT elements, and the mesh is refined to 0.01×0.03×0.1mm in the cut area to enhance the accuracy of the cutting analysis. The mesh size is increased along the tool cutting edge direction to reduce the total mesh size and increase the calculation speed. A master-slave setting is used for the contact interface, with the tool surface defined as the master surface and the cut surface of the GH4169 workpiece defined as the slave surface. The JC constitutive model is used, and its parameters are shown in Table 1. In finite element analysis, it is crucial to set a chip separation criterion for the mesh in order to simulate the separation of chips from the workpiece. The JC dynamic failure criterion is adopted as the physical separation criterion, and the equivalent plastic strain value of the overall mesh nodes is used as the judgment criterion. When the failure parameter ω≥1, the mesh element fails. The JC failure criterion parameters are shown in Table 2.

[0021] Table 1 JC Constitutive Model Parameter Table Table 2 JC Damage Criterion Parameter Table After completing the simulation calculations, the surface integrity parameters can be directly obtained from the above model.

[0022] The surface integrity parameters obtained from the simulation showed minimal error compared to the actual results. The inventors selected five sets of broaching parameters for actual broaching operations, and calculated the final measured and simulated surface integrity parameters. The accuracy was expressed as the average relative error. The average relative error for residual stress was 4.61%, for surface roughness 15.44%, and for hardness 7.94%, with an overall accuracy of 90.67%, indicating relatively high accuracy. Furthermore, a t-test was used to analyze the statistical differences between the measured and simulated surface integrity parameters. The results showed that the critical value P-value for the t-test was 0.987, indicating that the simulation results were relatively reliable.

[0023] The above operations have expanded the parameter samples before and after the pin pull. For fretting wear, the reliability of existing conventional simulation experiments is poor. Therefore, this embodiment only uses a small sample of the second dataset.

[0024] Meanwhile, from a practical standpoint, the fretting wear test parameters have a significant impact on the final wear volume. However, in this embodiment, to focus on and quantify the influence of the broaching machining parameters themselves on surface properties, the fretting wear test parameters are kept consistent. This isolates the influence of machining effects from numerous variables, avoiding interference from changes in wear parameters on the final results, thereby enabling the accurate establishment of a 'machining parameter-surface property' response model.

[0025] In this embodiment, the first dataset contains 150 samples. Due to the large amount of data, only the upper and lower limits of the broaching parameters are given here: broaching times of 1 to 3 times, measurement positions of 1 mm and 6 mm, broaching speed of 25 to 125 mm / s, and broaching depth of 0.01 mm to 0.08 mm. Within this range, simulation experiments are conducted based on Latin hypercube sampling.

[0026] The second dataset contains 50 samples. Due to the large amount of data, only the upper and lower limits of surface integrity and wear volume are given here: hardness 457~652.52, residual stress 261.3~970MPa, and surface roughness 244.65×10⁻⁶. -3 ~794×10 -3 The wear volume ranged from 31003 to 45520.662 mm. This dataset was obtained by randomly selecting 55 samples from the first dataset and performing broaching operations, surface parameter measurements, and fretting wear tests. Of these, 49 samples were used for model training, and the remaining 6 samples were used for model validation.

[0027] S2. The first model is obtained by substituting the broaching parameters as input and the surface integrity parameters as output into the first Kriging surrogate model for training. The second model is obtained by substituting the surface integrity parameters as input and the fretting wear parameters as output into the second Kriging surrogate model for training. In this step, the biggest difference between the first and second Kriging surrogate models lies in their training samples; the rest of the model architecture and parameters can be set to be the same.

[0028] During training, the Matern Cubic function was used as the correlation function, and the coefficient of determination R was used. 2 The maximum error (Maximum) is used to evaluate the accuracy of the trained model.

[0029] In this embodiment, the first dataset and the second dataset are respectively substituted into the first Kriging surrogate model and the second Kriging surrogate model for training, and finally the first model and the second model are obtained. The first model and the second model are validated using a validation set, and the final results are shown below: For the first model, such as Figure 1 As shown, its hardness R 2 The residual stress R is 0.932. 2 The surface roughness R is 0.922. 2 The values ​​are all greater than 90%, with a maximum error of 15.88% for the three parameters. This indicates that the first model has higher accuracy.

[0030] For the second model, the wear volume R 2 The value is 0.88, and the maximum error is 7.02%, indicating that its accuracy is high and it can be applied in this field.

[0031] In particular, considering the interpretability advantage of the method in this embodiment, this embodiment also includes the following step: performing response surface analysis on the first model and the second model.

[0032] like Figures 2-7 As shown, this illustrates the spatial distribution characteristics of residual stress with varying broaching speed and depth under the coupled effect of the number of broaching passes and the measurement location. Among these, Figures 2-4 The spatial distribution characteristics of residual stress with broaching speed and depth are shown for different broaching cycles (1 to 3 times) when the broaching position is 1 mm. Figures 5-7 The figures show the spatial distribution characteristics of residual stress with broaching speed and depth at different broaching cycles (1 to 3 times) when the broaching position is 1 mm.

[0033] As shown in the figure, the evolution of surface roughness distribution during broaching reflects the complex coupling effect between process parameters. The monotonically decreasing trend of roughness in the high-speed region during a single broaching pass indicates that higher cutting speeds help reduce friction and uneven plastic deformation, thus improving surface quality. The roughness peaks and nonlinear fluctuations appearing in the medium- and low-speed regions may be related to unstable cutting, uneven chip breaking, and material re-adhesion. With the increase of broaching passes, the roughness distribution exhibits more local extremes and low-value areas, indicating that repeated machining induces the accumulation of plastic deformation and the release of residual stress. The shift in roughness peak caused by changes in measurement position may originate from factors such as changes in tool load, increased heat accumulation, and decreased cooling efficiency, indicating that the broaching path position also has a significant impact on surface morphology.

[0034] like Figures 8-13 As shown, this illustrates the spatial distribution characteristics of surface roughness with varying broaching speed and depth under the coupled effect of the number of broaching passes and the measurement location. Among these, Figures 8-10 The spatial distribution characteristics of surface roughness as a function of broaching speed and depth are shown for different broaching cycles (1 to 3 times) when the broaching position is 1 mm. Figures 11-13The spatial distribution characteristics of surface roughness with broaching speed and depth are shown for different broaching cycles (1 to 3 times) when the broaching position is 6 mm.

[0035] As can be seen from the figure, the monotonic decreasing trend of roughness in the high-speed region of a single broaching operation indicates that higher cutting speeds help reduce friction and uneven plastic deformation, thus improving surface quality. The roughness peaks and nonlinear fluctuations appearing in the low-speed region may be related to unstable cutting, uneven chip breaking, and material re-adhesion. With the increase of broaching cycles, the roughness distribution shows more local extremes and low-value areas, indicating that repeated machining induces the accumulation of plastic deformation and the release of residual stress. The shift in roughness peak caused by changes in measurement position may be due to factors such as changes in tool load, increased heat accumulation, and decreased cooling efficiency, indicating that the broaching path position also has a significant impact on surface morphology.

[0036] like Figures 14-19 As shown, this illustrates the spatial distribution characteristics of hardness with broaching speed and depth under the coupled effect of broaching cycles and measurement location. Among them, Figures 14-16 The spatial distribution characteristics of hardness as a function of broaching speed and depth are shown for different broaching cycles (1 to 3 times) when the broaching position is 1 mm. Figures 17-19 The images show the spatial distribution characteristics of hardness as a function of broaching speed and depth at different broaching cycles (1 to 3 times) when the broaching position is 6 mm.

[0037] As can be seen from the figure, single broaching is mainly characterized by work hardening. With the increase of cutting speed, the plastic strain rate increases, leading to an increase in surface hardness. However, hardness troughs appear in the shallow region, which may be related to dynamic recovery or stress relaxation caused by local heat accumulation. After multiple broaching operations, the hardness distribution shows an asymmetrical or reversed trend, indicating that the material exhibits strain saturation during repeated processing. Multiple processing operations can easily induce residual stress redistribution, which may lead to a decrease in local hardness. The further away from the starting point, the slower the gradient of thermal-mechanical influence, resulting in a more gradual hardness distribution.

[0038] like Figures 20-25 As shown, this illustrates the impact of different surface integrity levels on fretting wear performance. Among them, Figures 20-21 This study demonstrates the spatial distribution characteristics of fretting wear as a function of residual stress and roughness at a specific hardness. Figures 22-23 This study demonstrates the spatial distribution characteristics of fretting wear as a function of residual stress and hardness under specific surface roughness conditions. Figures 24-25 This study demonstrates the spatial distribution characteristics of fretting wear as a function of hardness and surface roughness under specific residual stress conditions.

[0039] As shown in the figure, increased hardness generally improves the material's resistance to plastic deformation, thereby inhibiting wear. However, under high residual stress conditions, wear is exacerbated, possibly due to stress-induced microcrack propagation mechanisms. The coupling effect of roughness and residual stress is equally crucial. Lower roughness helps homogenize contact stress and reduce the wear rate, but an overly smooth surface may also lead to instability in adhesion and local oxide layers, thus triggering a new round of wear.

[0040] S3. Link the first model and the second model to obtain the full-process prediction model, which can then be used to make predictions.

[0041] In this step, the association refers to using the output of the first model as the input of the second model.

[0042] The prediction method in this embodiment achieves interpretable prediction of the entire process from machining parameters to surface condition to wear performance. Specifically, it analyzes the impact of broaching on surface integrity and the impact of surface integrity on fretting wear, and the model in this embodiment can analyze the correlation between different parameters.

[0043] In engineering, the prediction method in this embodiment can trace the source of wear acceleration under specific conditions, enabling technicians to identify which parameter(s) are causing the wear acceleration, thus making it easier to adjust the parameters.

[0044] In particular, considering the interpretability advantage of the method in this embodiment, this step also includes the following step: performing response surface analysis on the full-process prediction model.

[0045] like Figures 26-31 As shown, it illustrates the effect of broaching parameters on the fretting wear volume. Among them, Figures 26-28 The diagram shows the characteristic of wear volume variation with broaching speed and broaching depth when the broaching position is 1mm and the number of broaching cycles (1 to 3 times in sequence); Figures 29-31 The diagram shows the characteristic changes of wear volume with broaching speed and broaching depth when the broaching position is 6mm and the number of broaching cycles (1 to 3 times in sequence).

[0046] As shown in the figure, GH4169 exhibits a significant nonlinear wear response under different broaching depths and speeds. This nonlinearity stems not only from variations in a single parameter but also from the coupling effects between parameters. During the machining of high-temperature alloys, the interaction of thermo-mechanical loads significantly influences the wear mechanism, especially under high-speed cutting conditions, where the material surface is prone to thermal softening and plastic deformation, thus exacerbating wear. Multiple parameters jointly influence the mechanism. The expansion of the high wear band caused by repeated broaching also indicates that the accumulation of residual stress and the evolution of the material surface microstructure have a continuous impact on wear behavior. Furthermore, the differences in wear distribution corresponding to different broaching lengths indicate that the distribution of different surface integrity parameters of the workpiece leads to uneven wear response space. In summary, wear behavior is nonlinearly coupled and controlled by multiple parameters. Introducing a hybrid method of multivariables and intermediate parameters can accurately model the wear and improve the prediction accuracy of the complex wear law of GH4169.

[0047] The present invention has been disclosed above with preferred embodiments. However, those skilled in the art should understand that these embodiments are for illustrative purposes only and should not be construed as limiting the scope of the invention. Further improvements can be made without departing from the principles of the invention, and these improvements should also be considered as protections of the present invention.

Claims

1. A full-process prediction method based on broaching parameters, surface integrity, and fretting wear, characterized in that, Includes the following steps: S1. Obtain multiple sets of surface integrity parameters and fretting wear parameters of the broached material; S2. The first model is obtained by substituting the broaching parameters as input and the surface integrity parameters as output into the first Kriging surrogate model for training. The second model is obtained by substituting the surface integrity parameters as input and the fretting wear parameters as output into the second Kriging surrogate model for training. S3. Link the first model and the second model to obtain the full-process prediction model, which can then be used to make predictions.

2. The full-process prediction method based on broaching parameters, surface integrity, and fretting wear according to claim 1, characterized in that, The surface integrity parameters include hardness, residual stress, and surface roughness; the fretting wear parameters include wear volume; and the broaching parameters include broaching cycles, measurement location, broaching speed, and broaching depth.

3. The full-process prediction method based on broaching parameters, surface integrity, and fretting wear according to claim 1, characterized in that, The broaching parameters and surface integrity parameters are obtained based on measured values ​​and / or finite element simulation experiments, and the fretting wear parameters are obtained based on measured values.

4. The full-process prediction method based on broaching parameters, surface integrity, and fretting wear according to claim 1, characterized in that, During the training of the first and second Kriging surrogate models, the Matern Cubic function was used as the correlation function, and the coefficient of determination R was used. 2 The maximum error (Maximum) is used to evaluate the accuracy of the trained model.

5. The full-process prediction method based on broaching parameters, surface integrity, and fretting wear according to claim 1, characterized in that, S2 also includes the following steps: performing response surface analysis on the first model, the second model, and the full-process prediction model.

Images