Aero-engine casing bolted joint contact stress prediction method based on PCK model

Abstract

The present application relates to the field of aero-engine casing mounting edge contact stress prediction, in particular to an aero-engine casing bolt connection contact stress prediction method based on PCK model. The method can efficiently construct complex contact stress distribution and provide higher prediction accuracy than traditional methods; through systematic modeling and optimization of input variables, the method can accurately capture the influence of different loads, structures and material parameters on contact stress and achieve accurate prediction within the global range; compared with traditional methods, the prediction method based on PCK model can effectively avoid oversimplification and handle complex nonlinear relationships in high-dimensional parameter space; the method can provide reliable data support and theoretical basis for engineering design. The method calculates a large amount of contact stress data under the combination of different structures, loads and materials, and uses the contact stress prediction method of PCK model to realize accurate prediction of contact stress at any point of the casing contact surface.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of contact stress prediction of the mounting edge of an aero-engine casing, and in particular to a method for predicting the contact stress of the bolted connection of the mounting edge of an aero-engine casing based on a PCK model. BACKGROUND

[0002] The casing is one of the core components of an aero-engine, and plays an important role in supporting the rotor, transmitting loads and forming an airflow passage. The structural design and performance of the casing directly affect the overall efficiency and safety of the aero-engine. The bolted connection structure of the mounting edge of the casing not only serves as the main interface connecting the casing and other components, but also bears various loads such as gas pressure, temperature changes, casing tension and vibration. Therefore, the design of the bolted connection structure of the mounting edge of the casing requires a detailed analysis of the loads under various operating conditions of the engine to ensure that the mounting edge structure reliably transmits loads and provides good sealing during engine operation. Research has shown that if the contact stress distribution at the bolted connection is uneven or exceeds the allowable range, it may cause the bolts to loosen, gas to leak and the connection structure to be damaged, and even lead to serious accidents. Currently, there are still few quantitative studies on the bolted connection structure of the mounting edge of the casing at home and abroad, and there is a lack of methods for quantitatively predicting the contact stress of the mounting edge of the casing. In view of this, we propose a method for predicting the contact stress of the bolted connection of the mounting edge of an aero-engine casing based on a PCK model. SUMMARY

[0003] The purpose of the present application is to solve the problems existing in the prior art, and to propose a method for predicting the contact stress of the bolted connection of the mounting edge of an aero-engine casing based on a PCK model.

[0004] In order to achieve the above-mentioned purpose, the present application adopts the following technical solutions:

[0005] The method for predicting the contact stress of the bolted connection of the mounting edge of an aero-engine casing based on a PCK model comprises the following steps:

[0006] Step S1, determine the main structural parameters, load parameters and material parameters affecting the contact stress of the mounting edge of the casing and their value ranges;

[0007] Step S2, select a combination of data suitable for training the PCK model by Latin hypercube sampling;

[0008] Step S3, establish a bolted connection model of the mounting edge of the casing according to the combination selected by Latin hypercube sampling, and obtain the contact stress cloud map of the contact surface of each model by finite element calculation;

[0009] Step S4, extract the contact stress at each node of the contact surface of all models;

[0010] Step S5, establish the relationship between the point to be solved and the stress cloud map;

[0011] On the contact stress contour map obtained by finite element calculation, four straight lines are introduced, which intersect with six contour lines on the stress contour map, so as to divide the stress contour map into six ellipses, so as to better fit the stress distribution.

[0012] Determine the position of the point in the stress distribution;

[0013] Based on the equation formula (1) of the ellipse:

[0014] ax 2 +bxy 2 +cy 2 +dx+ey+f=0(1)

[0015] Where a, b, c, d, e, f are the coefficients of the equation.

[0016] Convert the point to be solved and the ellipse into the same coordinate system, and the converted coordinates (u", v") can be obtained from formula (2).

[0017]

[0018] Determine whether the point (u, v) to be solved is inside the ellipse, and calculate the T value in formula (3) by substituting u and v:

[0019] T=au 2 +buv+cu 2 +du+ev+f (3)

[0020] If T < 0, the point is inside the ellipse, otherwise, it is outside the ellipse, and when T = 0, the point is on the ellipse.

[0021] Step S6, a linear interpolation formula is established by using the minimum distance between the point to be solved and the stress values of the inner and outer ellipses.

[0022] For a given point (u, v), find the minimum distance between the point and the point (u", v") on the ellipse.

[0023]

[0024] Based on this parameter, the interpolation formula (5) can be obtained

[0025]

[0026] d i and d o are the distances from the inner ellipse and the outer ellipse respectively, p i and p o are the pressure values of the inner ellipse and the outer ellipse respectively.

[0027] Step S7, the contact stress data obtained by finite element calculation is divided into a training group and a verification group;

[0028] Step S8, the training group data is used to train the PCK model, and then the verification group data is input into the model for prediction, so as to obtain the contact stress prediction value of the 24 feature points;

[0029] Step S9, the 24 point data predicted by the PCK model is input into the formula of step S5 and step S6, the contact stress of the point to be solved is solved, and is verified with the verification group data.

[0030] Compared with the prior art, the method can efficiently construct complex contact stress distribution and provide higher prediction accuracy than traditional methods; through systematic modeling and optimization of input variables, the method can accurately capture the influence of different loads, structures and material parameters on contact stress and realize accurate prediction in the whole range; compared with traditional methods, the prediction method based on the PCK model can effectively avoid over-simplification and handle complex nonlinear relationships in high-dimensional parameter space; the method provides a powerful tool for evaluation and optimization of the contact stress distribution of the casing flange of the aero-engine, and provides reliable data support and theoretical basis for engineering design.

[0031] The method, by calculating a large amount of contact stress data under different structures, loads and materials, uses the contact stress prediction method of the PCK model to realize accurate prediction of the contact stress of any point on the casing contact surface.

[0032] Preferably, the specific method of step S4 is:

[0033] First step, extract the coordinates and contact stress of all nodes on the contact surface, and extract the coordinates of the 24 feature points;

[0034] Second step, use MATLAB command to fit the contact stress corresponding to the 24 feature points.

[0035] Further, the accuracy of the detection of the contact stress corresponding to the feature points is fully ensured, and the accuracy of the detection is ensured by the 24 points.

[0036] Preferably, the specific method of step S7 is:

[0037] First step, divide the finite element calculation results of 120 groups of models into a training group and a verification group, use the determination coefficient (R 2 ) as an evaluation index to test the prediction accuracy of the PCK model under different data sets;

[0038] Second step, through this process, the best training and verification data set grouping is determined, and finally 90 groups of training groups and 30 groups of verification groups are obtained.

[0039] Further, by setting multiple groups of models, training and verification can be effectively performed to improve the prediction accuracy of the PCK model.

[0040] Preferably, the step S1 specifically includes that the structure parameters of the casing mounting edge include the mounting edge thickness, bolt diameter and number of bolts.

[0041] The load parameters include bolt pre-tightening force and temperature.

[0042] The material parameters include TC-6, GH4169 and T700G.

[0043] Further, the method is more suitable for the installation and use of the aero-engine casing, fully ensures the firmness of the parts and improves the firmness of the connection.

[0044] Preferably, the step S2 specifically includes:

[0045] Firstly, all possible combinations are generated by arranging and combining the given structure, load and parameters.

[0046] Secondly, 120 uniformly distributed data sets are selected from the combinations by using the Latin hypercube sampling method to ensure comprehensive coverage of various conditions in the multi-dimensional space.

[0047] Further, by setting multiple groups of data, comprehensive coverage of various conditions in the multi-dimensional space is effectively ensured, and various bolt connection conditions are fully adapted.

[0048] Preferably, the step S3 specifically includes that the bolt connection model of the casing mounting edge is established according to the combinations selected by the Latin hypercube sampling, then the contact stress of the contact surface of each model is calculated by using the finite element analysis method, and the corresponding stress nephogram is generated to visualize the contact stress distribution under different combinations.

[0049] Further, the bolt connection model of the casing mounting edge is fully established, and the visual data graph is effectively generated to understand the bolt connection condition.

[0050] Preferably, the step S8 specifically includes that the PCK model is trained using 90 training data sets, then 30 verification data sets are input into the trained model to verify the effectiveness and accuracy thereof, and the structure parameters, load parameters and material parameters of the model are input into the PCK model to output the contact stress and coordinates of 24 feature points of the model.

[0051] Further, the accuracy of the PCK model is effectively ensured, and the contact stress of 24 feature points and the coordinates of each feature point are improved.

[0052] Preferably, the step S9 is specifically that the 24 feature points predicted by the PCK model are input into the step S5 and the step S6, so that the contact stress of the point is obtained.

[0053] Further, the contact stress of the corresponding point is effectively obtained.

[0054] The beneficial effects of the present application are:

[0055] 1. The present application can efficiently construct complex contact stress distribution and provide higher prediction accuracy than traditional methods; through systematic modeling and optimization of input variables, the method can accurately capture the influence of different loads, structures and material parameters on contact stress and realize accurate prediction in the global range; compared with traditional methods, the prediction method based on the PCK model can effectively avoid oversimplification and handle complex nonlinear relationships in high-dimensional parameter space; the method provides a powerful tool for evaluation and optimization of the contact stress distribution of the shell flange of the aero-engine, and provides reliable data support and theoretical basis for engineering design.

[0056] 2. The present application calculates a large amount of contact stress data under the combination of different structures, loads and materials, and uses the contact stress prediction method of the PCK model to realize accurate prediction of the contact stress of any point on the casing contact surface. BRIEF DESCRIPTION OF DRAWINGS

[0057] Figure 1 The flowchart of the present application;

[0058] Figure 2 The bolt connection structure schematic diagram of the present application;

[0059] Figure 3 The installation edge structure parameter and load parameter value range of the casing in the present application

[0060] Figure 4 The casing installation edge and bolt material parameters in the present application

[0061] Figure 5 The three-dimensional geometric model of the bolt connection structure of the present application;

[0062] Figure 6 The stress nephogram construction schematic diagram of the 24 feature points in the present application;

[0063] Figure 7 The contact surface all node error diagram obtained by the contact stress quantitative prediction based on the PCK model. DETAILED DESCRIPTION

[0064] Clearly, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments.

[0065] With reference to Figures 1-7 , the PCK model-based contact stress prediction method for the bolt connection of an aero-engine casing includes the following steps:

[0066] Step S1, determining the main structure parameters, load parameters and material parameters affecting the contact stress of the casing mounting edge and their value ranges;

[0067] The structure parameters of the casing mounting edge include the mounting edge thickness, bolt diameter and number of bolts; the load parameters include the bolt pre-tightening force and temperature; and the material parameters include TC-6, GH4169 and T700G; Figure 2 The bolt connection structure schematic diagram provided for the embodiments of the present application is shown in the figure;

[0068] Step S2, screening out data combinations suitable for training the PCK model by Latin hypercube sampling;

[0069] All possible combinations are generated by arranging and combining the given structures, loads and parameters. Figure 3 and Figure 4 Then, 120 uniformly distributed data sets are screened out from these combinations by using the Latin hypercube sampling method, so as to ensure comprehensive coverage of each condition in the multi-dimensional space.

[0070] Step S3, establishing the bolt connection model of the casing mounting edge according to the combinations screened out by the Latin hypercube sampling, and obtaining the contact stress nephogram of each model by finite element calculation;

[0071] The bolt connection model of the casing mounting edge is established according to the combinations screened out by the Latin hypercube sampling. Figure 5 The three-dimensional geometric model of the bolt connection structure provided for the embodiments of the present application is shown in the figure. Then, the contact stress of each model is calculated by using the finite element analysis method, and the corresponding stress nephogram is generated, so as to visualize the contact stress distribution under different combinations.

[0072] Step S4, extracting the contact stress of each node of the contact surface of all models;

[0073] The coordinates and contact stress of all nodes of the contact surface are extracted, and the coordinates of 24 feature points are extracted. Then, the contact stress corresponding to the 24 feature points is fitted by using the MATLAB command.

[0074] Step S5, establishing the relationship between the sought point and the stress nephogram;

[0075] On the contact stress contour map obtained by finite element calculation, four straight lines are introduced, which intersect with six contour lines on the stress contour map, so as to divide the stress contour map into six ellipses, as shown in Figure 6 .

[0076] Determine the position of the point in the stress distribution;

[0077] Based on the equation formula (1) of the ellipse:

[0078] ax 2 +bxy 2 +cy 2 +dx+ey+f=0(1)

[0079] Where a, b, c, d, e, f are the coefficients of the equation.

[0080] Convert the point to be solved and the ellipse into the same coordinate system, and the converted coordinates (u", v") can be obtained from formula (2).

[0081]

[0082] Determine whether the point (u, v) to be solved is inside the ellipse, and calculate the T value in formula (3) by substituting u and v:

[0083] T=au 2 +buv+cu 2 +du+ev+f (3)

[0084] If T<0, the point is inside the ellipse, otherwise, it is outside the ellipse, and when T=0, the point is on the ellipse.

[0085] Step S6, a linear interpolation formula is established by using the minimum distance between the stress value of the point to be solved and the inner and outer ellipses.

[0086] For a given point (u, v), the minimum distance between the point and the point (u", v") on the ellipse is solved.

[0087]

[0088] Based on this parameter, the interpolation formula (5) can be obtained

[0089]

[0090] d i and d o are the distances from the inner ellipse and the outer ellipse respectively, p i and p o are the pressure values of the inner ellipse and the outer ellipse.

[0091] Step S7, the contact stress data obtained by the finite element calculation is divided into a training group and a verification group;

[0092] The finite element calculation results of 120 groups of models are divided into a training group and a verification group, the coefficient of determination (R 2 ) is used as an evaluation index to test the prediction accuracy of the PCK model under different data sets. Through this process, the best training and verification data set grouping is determined, and finally 90 groups of training groups and 30 groups of verification groups are obtained.

[0093] Step S8, the training group data is used to train the PCK model, and then the verification group data is input into the model for prediction, so as to obtain the contact stress prediction value of the 24 characteristic points;

[0094] The PCK model is trained using 90 groups of training data sets, and then 30 groups of verification data sets are input into the trained model to verify its effectiveness and accuracy. Finally, the structure parameters, load parameters and material parameters of the model are input into the PCK model, and the contact stress and coordinates of the 24 characteristic points of the model are output.

[0095] Step S9, the 24 point data predicted by the PCK model is input into the formulas in steps 5 and 6 to solve the contact stress of the point, and is verified with the verification group data.

[0096] The 24 characteristic points predicted by the PCK model are input into steps 5 and 6, so that the contact stress of the point can be obtained.

[0097] In this embodiment, the predicted example parameters are 4mm for the installation edge thickness, 8mm for the bolt diameter, 96 for the number of bolts, 8750N for the pretightening force, 300K for the temperature, TC-6 for the installation edge material, and GH4169 for the bolt material. As shown in Figure 5 The maximum absolute error of the contact stress obtained by the global contact stress quantitative prediction method based on the PCK model is only 17.45MPa. The contact surface error is shown in Figure 7 The accuracy and reliability of the prediction method are verified, which can provide a reference for the contact stress prediction method of the bolt connection of the installation edge of the casing.

[0098] The above is only a preferred specific embodiment of the present application, but the protection scope of the present application is not limited thereto, any skilled person in the art can make equivalent replacement or change according to the technical solution and the inventive concept of the present application within the technical range disclosed by the present application, which should be covered within the protection scope of the present application.

Claims

1. A method for predicting contact stress of an aeroengine case bolted joint based on PCK model, characterized in that, The method comprises the following steps: Step S1, determining the main structure parameters, load parameters and material parameters affecting the contact stress of the mounting edge of the casing and their value ranges; Step S2, screening out data combinations suitable for training the PCK model by Latin hypercube sampling; Step S3, establishing a bolt connection model of the mounting edge of the casing according to the combinations screened out by the Latin hypercube sampling, and obtaining the contact stress nephogram of each model by finite element calculation; Step S4, extracting the contact stresses of each node of the contact surface of all models; Step S5, establishing the relationship between the point to be solved and the stress nephogram; On the contact stress nephogram obtained by finite element calculation, four straight lines are introduced, which intersect with six contour lines on the stress nephogram, so that the stress nephogram is divided into six ellipses, so as to better fit the stress distribution; Judging the position of the point to be solved in the stress distribution; Based on the equation formula (1) of the ellipse: ax 2 +bxy 2 +cy 2 +dx+ey+f = 0 (1) Wherein, a, b, c, d, e, f are the coefficients of the equation; The point to be solved and the ellipse are converted to the same coordinate system, and the converted coordinates (u", v") can be obtained from formula (2); Judging whether the point to be solved (u, v) is inside the ellipse, substituting u and v into formula (3) to calculate T value: T = au 2 + buv + cu 2 + du + ev + f (3) If T < 0, the point is inside the ellipse, otherwise, it is outside the ellipse, and when T = 0, the point is on the ellipse; Step S6, establishing a linear interpolation formula using the stress values and minimum distances of the point to be solved and the inner and outer ellipses; For a given point (u, v), the minimum distance between the point and the point (u", v") on the ellipse is solved; Based on this parameter, the interpolation formula (5) can be obtained d i and d o distance from the inner and outer ellipses, p i and p o pressure values for the inner and outer ellipses Step S7, dividing the contact stress data obtained by finite element calculation into training groups and verification groups; Step S8, using the training group data to train the PCK model, and then inputting the verification group data into the model for prediction, so as to obtain the contact stress prediction values of the 24 feature points; Step S9, inputting the 24 point data predicted by the PCK model into the formulas of step S5 and step S6 to solve the contact stress of the point to be solved, and verifying it with the verification group data.

2. The PCK model based aero-engine case bolted joint contact stress prediction method of claim 1, wherein: The specific method of step S4 is: First step, extracting the coordinates and contact stresses of all nodes of the contact surface, and extracting the coordinates of the 24 feature points; Second step, using MATLAB commands to fit the contact stresses corresponding to the 24 feature points.

3. The PCK model based aero-engine case bolted joint contact stress prediction method of claim 1, wherein: The specific method of step S7 is: Firstly, the results of 120 groups of models were divided into training group and validation group, and the coefficient of determination (R 2 ) was used as the evaluation index to test the prediction accuracy of PCK model under different data sets. Second step, through this process, the best training and verification data set grouping is determined, and finally 90 training groups and 30 verification groups are obtained.

4. The PCK model based aero-engine case bolted joint contact stress prediction method of claim 1, wherein: The specific method of step S1 is that the structure parameters of the mounting edge of the casing include the thickness of the mounting edge, the diameter of the bolt and the number of bolts: The load parameters include the bolt pretightening force and the temperature; The material parameters include TC-6, GH4169 and T700G.

5. The PCK model based aero-engine case bolted joint contact stress prediction method of claim 1, wherein: The specific method of step S2 is: First step, arranging and combining the given structure, load and parameters to generate all possible combinations; Second step, using the Latin hypercube sampling method to screen out 120 evenly distributed data sets from these combinations to ensure comprehensive coverage of each condition in the multi-dimensional space.

6. The PCK model based aero-engine case bolting contact stress prediction method of claim 1, wherein: The step S3 specifically comprises: establishing a bolt connection model of the installation edge of the case according to the combination screened out by the Latin hypercube sampling; then, calculating the contact stress of the contact surface of each model by using a finite element analysis method, and generating a corresponding stress nephogram to visualize the contact stress distribution under different combinations.

7. The PCK model based aero-engine case bolting contact stress prediction method of claim 1, wherein: The step S8 specifically comprises: training the PCK model using 90 sets of training data sets, then inputting 30 sets of verification data sets into the trained model to verify the effectiveness and accuracy thereof; inputting the structural parameters, load parameters and material parameters of the model into the PCK model, and outputting the contact stress and coordinates of 24 feature points of the model.

8. The PCK model based aero-engine case bolting contact stress prediction method of claim 1, wherein: The step S9 specifically comprises: inputting the 24 feature points predicted by the PCK model into the steps S5 and S6, so as to obtain the contact stress of the points.

Images