Low-temperature hose structure optimization method

Abstract

The invention discloses a low-temperature hose structure optimization method which comprises the following steps: determining structure parameters of a low-temperature hose to be optimized, namely a wave diameter R0, a wall thickness delta0, a wave angle theta0 and a wave height H0, calculating d0 based on the wave diameter R0 and the wave height H0, d0 = H0 / 2-R0, determining four control points P0, P1, P2 and P3 of a Bezier curve and the corresponding Bezier curve by adopting the wave diameter R0, the wave angle theta0 and d0, the waveform of the low-temperature hose is determined according to the Bezier curve; establishing a low-temperature hose numerical calculation model, establishing an optimal Latin square test scheme, performing analog simulation calculation according to the optimal Latin square scheme, and constructing fitting functions Y1, Y2 and Y3 based on an RBF neural network model; and taking Y1 as a constraint function, taking weighted optimization of Y2 and Y3 as a target function, and performing global optimization based on a multi-island genetic algorithm to obtain optimized structure parameters of the low-temperature hose. By optimizing structural parameters, the fatigue performance of the low-temperature hose can be effectively improved.

Description

Technical Field

[0001] The present invention relates to the technical field of cryogenic hoses, and more particularly to a method for optimizing the structure of cryogenic hoses. Background Art

[0002] As an essential component of the cryogenic propellant transport chain at space launch sites, metal hoses compensate for the additional stresses caused by temperature differences and mechanical vibration. However, as the frequency of use increases, fatigue source areas of these hoses are frequently subjected to alternating loads generated by temperature and external bending moments, causing fatigue damage to accumulate. This ultimately leads to fatigue fracture in the bellows, resulting in hose failure and leakage. Therefore, there is an urgent need to provide a cryogenic hose structural optimization method to improve its fatigue resistance. Summary of the Invention

[0003] An object of the present invention is to solve at least the above problems and to provide at least the advantages which will be described hereinafter.

[0004] Another object of the present invention is to provide a method for optimizing the structure of a cryogenic hose, which can effectively improve the fatigue performance of the cryogenic hose by optimizing structural parameters.

[0005] In order to achieve these objects and other advantages according to the present invention, a method for optimizing the structure of a cryogenic hose is provided, comprising the following steps:

[0006] Step 1: Determine the structural parameters of the cryogenic hose to be optimized, namely, wave path R0, wall thickness δ0, wave angle θ0, and wave height H0. Calculate d0 based on the wave path R0 and wave height H0, where d0 = H0 / 2-R0. Use the wave path R0, wave angle θ0, and d0 to determine the four control points P0, P1, P2, and P3 of the Bezier curve and the corresponding Bezier curve. Determine the waveform of the cryogenic hose based on the Bezier curve.

[0007] Step 2. Establish a numerical calculation model for the cryogenic hose, and use the structural parameters of wave diameter R, wall thickness δ, wave angle θ, wave height H, and d as variables to establish an optimal Latin square test scheme. According to the optimal Latin square scheme, apply load to the numerical calculation model of the cryogenic hose, and perform simulation calculations. Based on the simulation calculation results, obtain the stiffness Y1, mass Y2, and life Y3 of each scheme, where the value range of the variable wave diameter is R0-R0*a%≤R≤R0+R0*a%, the value range of the variable wall thickness is δ0-δ0*b%≤δ≤δ0+δ0*b%, the value range of the variable wave angle is θ0-θ0*c%≤θ≤θ0+θ0*c%, the value range of the variable wave height is H0-H0*d%≤H≤H0+H0*d%, and a, b, c, and d are all positive numbers less than 100.

[0008] Step 3: Based on the RBF neural network model, construct the stiffness fitting function Y1=f(R, d, θ, δ), the mass fitting function Y2=g(R, d, θ, δ), and the life fitting function Y3=h(R, d, θ, δ);

[0009] Step 4: Using Y1 as the constraint function and the weighted optimal value of Y2 and Y3 as the objective function, a global optimization is performed based on the multi-island genetic algorithm to obtain the optimized structural parameters of the cryogenic hose.

[0010] Preferably, a=30, b=30, c=20, and d=20.

[0011] Preferably, the four control points P, P1, P2, and P3 are determined by first determining the horizontal center line of the cryogenic hose, and based on d, determining the upper and lower parallel lines parallel to the horizontal center line, determining the vertical center line of the wave crest and the wave trough, and drawing a circle with the wave diameter R as the center point of the intersection of the vertical center line and the upper and lower horizontal lines, determining the control point P0 at the wave crest by the wave angle θ, drawing a first extension line from P0 to the center point of the circle, and drawing a first tangent line at the control point P0;

[0012] Draw a circle with the wave diameter R at the trough adjacent to the wave crest, determine the control point P3 at the trough by the wave angle θ, draw a second extension line along P3 and the center of the circle, and draw a second tangent line at the control point P3. The second extension line is parallel to the first extension line, and the first tangent line is parallel to the second tangent line.

[0013] The intersection point of the first extension line and the second tangent line is the control point P1, and the intersection point of the second extension line and the first tangent line is the control point P2.

[0014] The present invention further provides an electronic device comprising: at least one processor, and a memory communicatively connected to the at least one processor, wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to perform the low-temperature hose structure optimization method.

[0015] The present invention further provides a storage medium on which a computer program is stored. When the program is executed by a processor, the method for optimizing the structure of the cryogenic hose is implemented.

[0016] The present invention has at least the following beneficial effects:

[0017] First, the present invention proposes a numerical calculation model for low-temperature metal hoses, which can calculate the fatigue performance under mechanical loads.

[0018] Second, the present invention applies a parametric design method, which is more efficient and accurate than traditional design methods.

[0019] Third, the present invention proposes a method for applying the multi-island genetic algorithm to the optimization design of low-temperature metal hoses, which can avoid falling into local optimal solutions and achieve efficient and high-performance global optimization of product structure.

[0020] Fourth, from the results of Example 1, it can be seen that under the premise of meeting the compensation capacity, the optimized model of the present invention has a 55.6% increase in stiffness compared with the original model, a lifespan increased to 9.45 times the original, and a weight reduction of 5%. The optimization effect is good, proving that the design method is reliable.

[0021] Other advantages, objectives and features of the present invention will be reflected in part through the following description, and in part will be understood by those skilled in the art through study and practice of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] Figure 1 This is a flow chart of a method for optimizing the structure of a cryogenic hose according to one of the technical solutions of the present invention;

[0023] Figure 2 A waveform diagram of a cryogenic hose according to one of the technical solutions of the present invention;

[0024] Figure 3 A life cloud diagram of a cryogenic hose to be optimized according to one of the technical solutions of the present invention;

[0025] Figure 4 This is a corrugated cross-sectional view of a cryogenic hose to be optimized according to one of the technical solutions of the present invention;

[0026] Figure 5 This is a life cloud diagram of the optimized cryogenic hose according to one of the technical solutions of the present invention;

[0027] Figure 6 This is a corrugated cross-sectional view of an optimized cryogenic hose according to one of the technical solutions of the present invention;

[0028] Figure 7 A waveform diagram of the cryogenic hose to be optimized according to one of the technical solutions of the present invention;

[0029] Figure 8 A grid division diagram of a numerical calculation model of a cryogenic hose according to one of the technical solutions of the present invention;

[0030] Figure 9 A graph showing a change in mean square error of a neural network according to one of the technical solutions of the present invention;

[0031] Figure 10 This is a schematic diagram of the iterative process described in one of the technical solutions of the present invention. DETAILED DESCRIPTION

[0032] The present invention is further described in detail below with reference to the embodiments so that those skilled in the art can implement the invention with reference to the description.

[0033] like Figure 1 As shown, the low-temperature hose structure optimization method includes the following steps:

[0034] Step 1: Determine the structural parameters of the cryogenic hose to be optimized, including wave diameter R0, wall thickness δ0, wave angle θ0, and wave height H0 (the wave diameter, wall thickness, wave angle, and wave height of the cryogenic hose are all reference values ​​for the cryogenic hose to be optimized, such as Figure 8 As shown in the figure, d0 is calculated based on the wave path R0 and wave height H0, d0=H0 / 2-R0, and the coordinates of the four control points P0, P1, P2, and P3 of the Bezier curve and the corresponding Bezier curve are determined using the wave path R0, wave angle θ0, and d0. The waveform of the cryogenic hose is determined according to the Bezier curve (the waveform is converted from U-type to S-type by controlling the value range of the wave angle θ). The formula of the Bezier curve is:

[0035] B(t)=(1-t) 3 P0+3t(1-t) 2 P1+3t 2 (1-t)P2+t 3 P3,t∈[0,1]

[0036] Where B(t) is a point on the Bezier curve, P0 is the starting point, P1 and P2 are control points, and P3 is the end point;

[0037] Step 2: Based on the waveform and wall thickness, a numerical calculation model of the cryogenic hose is established (a three-dimensional model is first established based on the waveform and wall thickness, and the three-dimensional model is meshed to obtain the numerical calculation model of the cryogenic hose). The optimal Latin square test scheme is established with the structural parameters of wave path R, wall thickness δ, wave angle θ, wave height H, and d as variables. According to the optimal Latin square scheme, load is applied to the numerical calculation model of the cryogenic hose, and simulation calculation is performed (in Ansys The FatigueTools module in the WorkBench simulation platform analyzes the fatigue life of the metal hose based on the S-N curve method (strain-life curve) to obtain the life of each scheme, and obtains the support reaction force and displacement of each scheme based on the simulation calculation results. The ratio of the support reaction force and displacement is the stiffness; the volume of each scheme is read based on the three-dimensional modeling software or simulation software, and the product of the volume and the material density is the mass). Based on the simulation calculation results, the stiffness Y1, mass Y2, and life Y3 of each scheme are obtained, among which the value range of the variable wave path is R0-R0*a%≤R≤R0+R0*a%, the value range of the variable wall thickness is δ0-δ0*b%≤δ≤δ0+δ0*b%, the value range of the variable wave angle is θ0-θ0*c%≤θ≤θ0+θ0*c%, and the value range of the variable wave height is H0-H0*d%≤H≤H0+H0*d%, and a, b, c, and d are all positive numbers less than 100;

[0038] Step 3: Based on the RBF neural network model (the stiffness, mass, and life obtained from the Latin square scheme and each scheme are used as sample data, and the sample data are divided into a training set, and the training set is used for training to obtain the fitting results, and then verified by the test set. The composite correlation number is greater than 0.9, that is, the relevant fitting function is obtained), a stiffness fitting function Y1 = f(R, d, θ, δ), a mass fitting function Y2 = g(R, d, θ, δ), and a life fitting function Y3 = h(R, d, θ, δ) are constructed;

[0039] Step 4: Using Y1 as the constraint function and the weighted optimal value of Y2 and Y3 as the objective function, a global optimization is performed based on the multi-island genetic algorithm to obtain the optimized structural parameters of the cryogenic hose.

[0040] In another technical solution, a=30, b=30, c=20, d=20, the value range of the variable wave diameter is R0-R0*30%≤R≤R0+R0*30%, the value range of the variable wall thickness is δ0-δ0*30%≤δ≤δ0+δ0*30%, the value range of the variable wave angle is θ0-θ0*20%≤θ≤θ0+θ0*20%, and the value range of the variable wave height is H0-H0*20%≤H≤H0+H0*20%, further narrowing the value range of the variables.

[0041] In another technical solution, Figure 2As shown, the method for confirming the four control points P0, P1, P2, and P3 is as follows: first determine the horizontal center line 1 of the cryogenic hose (the horizontal center line is located at 1 / 2 of the wave height), and based on d, determine the upper parallel line 2 and the lower parallel line 8 parallel to the horizontal center line (the distance between the horizontal center line and the upper horizontal line, and the distance between the horizontal center line and the lower horizontal line are both d), determine the vertical center line 3 of the wave crest and the wave trough (there is a vertical center line at the wave crest and a vertical center line at the wave trough), use the intersection of the vertical center line 3 with the upper horizontal line 2 and the lower horizontal line 8 as the center of the circle, and draw a circle with the wave diameter R. At the wave crest, determine the control point P0 by the wave angle θ, draw a first extension line 5 from P0 to the center of the circle, and draw a first tangent line 4 at the control point P0;

[0042] Draw a circle with the wave diameter R at the trough adjacent to the wave crest. Determine the control point P3 at the trough using the wave angle θ. Draw a second extension line 7 along P3 and the center of the circle, and draw a second tangent line 6 at the control point P3. The second extension line 7 is parallel to the first extension line 5, and the first tangent line 4 is parallel to the second tangent line 6.

[0043] Among them, the intersection of the first extension line 5 and the second tangent line 6 is the control point P1, and the intersection of the second extension line 7 and the first tangent line 4 is the control point P. Accurately determining the positions of the four control points P0, P1, P2, and P3 facilitates determining the waveform of the low-temperature hose to be optimized.

[0044] An electronic device includes: at least one processor, and a memory communicatively connected to the at least one processor, wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to perform the method for optimizing the structure of a cryogenic hose.

[0045] A storage medium stores a computer program, which, when executed by a processor, implements the method for optimizing the structure of a cryogenic hose.

[0046] <Example 1>

[0047] Take DN350 as the low-temperature hose to be optimized (U-shaped corrugation, such as Figure 7 As shown), the life cloud diagram of DN350 before optimization is as follows Figure 3 As shown, the waveform cross-section of the cryogenic hose to be optimized is as follows Figure 4 As shown, the U-shaped bellows with a design pressure of 1.6 MPa is the target optimization model;

[0048] 1. The structural parameters of the cryogenic hose to be optimized are shown in Table 1. In step 1, the wave path R0, wall thickness δ0, wave angle θ0, and wave height H0 of the cryogenic hose to be optimized are determined to correspond to the reference values ​​in Table 1. The waveform diagram of the cryogenic hose to be optimized is shown in Figure 2 As shown, the red line is the waveform;

[0049] Table 1 shows the structural parameters of the cryogenic hose to be optimized

[0050]

[0051] 2. Establish a numerical calculation model for cryogenic hose

[0052] The double-wave model is selected for the three-dimensional model. The three-dimensional model is established using three-dimensional modeling software based on the waveform and wall thickness. The three-dimensional model is then meshed with 61867 grids and 30231 nodes. A single side of the model is set as a fixed constraint, and a displacement constraint of 1mm is applied to the other side. The mesh division of the numerical calculation model of the cryogenic hose is as follows: Figure 8 As shown;

[0053] 3. Latin square test scheme and simulation calculation

[0054] The stiffness, mass, and life obtained from the Latin square test scheme and simulation calculations are shown in Table 2;

[0055] Table 2 shows the Latin square scheme and simulation results

[0056]

[0057]

[0058] 4. Constructing the fitting function

[0059] 4.1 Training Model

[0060] Using an RBF neural network model, a mapping relationship is constructed between structural parameters (i.e., the four optimized parameters: wave path R, wall thickness δ, wave angle θ, and wave height H) and output parameters. Since the optimal structural parameters are wave path R, wave height H, wave angle θ, and wall thickness δ, the input layer is set to have u = 4 neurons; the output layer is set to have v = 3 neurons, representing stiffness, mass, and lifespan. According to Kolmogorov's theorem, the number of neurons k in the hidden layer is determined by the following general empirical formula:

[0061] k=2u+1 (3)

[0062] In Table 2, data sets 1-22 are training sets, and data sets 23-27 are test sets;

[0063] In the Isight RBF neural network tool flow, set up the following:

[0064] Smoothing Filter: Select the default value of 0.

[0065] Type of Basis Function: Select RBF.

[0066] Maximum Iterations to fit: Set to 500.

[0067] Figure 9 The figure shows the change curve of the mean square error of the RBF neural network. Figure 9 As can be seen, the RBF neural network converges to a learning error accuracy of 10 at the 300th step. -3 , the neural network training is completed.

[0068] 4.2 Accuracy test

[0069] The value of the complex correlation coefficient is used to test the fitting results of the RBF neural network model. The complex correlation coefficient R 2 is defined as follows:

[0070]

[0071] Where: n is the number of solution samples for the calculation model; y is the mean of the true value; y i is the actual value of the model; is the estimated value obtained by the approximate model. 2 The closer the value of is to 1, the higher the accuracy of the constructed performance prediction model.

[0072] After error analysis, the correlation coefficients based on the RBF neural network model are all above 0.9, which proves that the fitting results of the RBF model are good, that is, the stiffness fitting function Y1 = f(R, d, θ, δ), the mass fitting function Y2 = g(R, d, θ, δ), and the life fitting function Y3 = h(R, d, θ, δ) are obtained.

[0073] 5. Global training and excellence

[0074] The purpose of training with the multi-island genetic algorithm is to meet the requirements of optimizing the flexibility of the cryogenic hose while improving its fatigue life. Therefore, the optimization target is maxN, and the constraints are: 3000N / mm≤K≤6000N / mm; the optimization variables x1, x2, x3, and x4 represent R, δ, H, and θ, respectively, and the corresponding value ranges are: 4.375≤x1≤8.125, 0.84≤x2≤1.56, 25.6≤x3≤38.4, and 144≤x4≤216.

[0075] The number of iterations of the optimization process is as follows Figure 10As shown in the figure, this optimization process iterated for 1002 steps, with obvious local convergence appearing from the 400th generation. As the optimization calculation progressed, the number of local convergences gradually increased, indicating good convergence. Finally, all iterative calculation data were arranged to obtain the global search results. The blue dots represent feasible solutions in the optimization process, that is, solutions that meet the constraints; the red dots represent infeasible solutions, that is, solutions that do not meet the constraints; and the green dots represent the optimal solution. The optimal solution obtained in this optimization process is: K = 5678 N / mm, N = 158990 times.

[0076] The final optimization results are shown in Table 3, and the optimized life cloud diagram is as follows: Figure 5 As shown, the cross-sectional view (three-dimensional) of the optimized cryogenic hose is as follows Figure 6 As shown;

[0077] Table 3 Optimized structural parameters

[0078]

[0079] By comparison, it can be seen that, under the premise of meeting the compensation capacity, the stiffness of the optimized model is increased by 55.6% compared with the original model, the life is increased to 9.45 times of the original, and the mass is reduced by 5%. The optimization effect is good, proving that the design method is reliable.

[0080] Although the embodiments of the present invention have been disclosed above, they are not limited to the applications listed in the description and implementation methods. They can be fully applied to various fields suitable for the present invention. For those familiar with the art, additional modifications can be easily implemented. Therefore, without departing from the general concept defined by the claims and the scope of equivalents, the present invention is not limited to the specific details and illustrations shown and described herein.

Claims

1. A method for optimizing the structure of a cryogenic hose, characterized in that: The following steps are involved: Step 1, determine the structural parameters of the cryogenic hose to be optimized, namely, wave diameter R0, wall thickness δ0, wave angle θ0, and wave height H0, calculate d0 based on the wave diameter R0 and wave height H0, d0=H0 / 2-R0, use the wave diameter R0, wave angle θ0, and d0 to determine the four control points P0, P1, P2, and P3 of the Bezier curve and the corresponding Bezier curve, and determine the waveform of the cryogenic hose according to the Bezier curve; Step 2, establish a numerical calculation model of the cryogenic hose, and use the structural parameters of wave diameter R, wall thickness δ, wave angle θ, wave height H, d as variables, establish an optimal Latin square test scheme, apply load to the numerical calculation model of the cryogenic hose according to the optimal Latin square scheme, and perform simulation calculations. Based on the simulation calculation results, obtain the stiffness Y1, mass Y2, and life Y3 of each scheme, where the value range of the variable wave diameter is R0-R0*a%≤R≤R0+R0*a%, the value range of the variable wall thickness is δ0-δ0*b%≤δ≤δ0+δ0*b%, the value range of the variable wave angle is θ0-θ0*c%≤θ≤θ0+θ0*c%, the value range of the variable wave height is H0-H0*d%≤H≤H0+H0*d%, and a, b, c, d are all positive numbers less than 100; Step 3: Based on the RBF neural network model, construct the stiffness fitting function Y1=f(R, d, θ, δ), the mass fitting function Y2=g(R, d, θ, δ), and the life fitting function Y3=h(R, d, θ, δ); Step 4: Take Y1 as the constraint function, the weighted optimal value of Y2 and Y3 as the objective function, perform global optimization based on the multi-island genetic algorithm, and obtain the optimized structural parameters of the cryogenic hose.

2. The method for optimizing the structure of a cryogenic hose according to claim 1, characterized in that: a=30, b=30, c=20, d=20.

3. The method for optimizing the structure of a cryogenic hose according to claim 1, characterized in that: The confirmation method of the four control points P0, P1, P2, and P3 is as follows: first determine the horizontal center line of the cryogenic hose, and determine the upper parallel line and the lower parallel line parallel to the horizontal center line based on d, determine the vertical center line of the wave crest and the wave trough, and use the intersection of the vertical center line and the upper horizontal line and the lower horizontal line as the center of the circle and draw a circle with the wave diameter R, determine the control point P0 at the wave crest through the wave angle θ, draw the first extension line along P0 and the center of the circle, and draw the first tangent line at the control point P0; Draw a circle with a wave diameter R at the trough adjacent to the wave crest, determine a control point P3 at the trough by a wave angle θ, draw a second extension line along P3 and the center of the circle and draw a second tangent line at the control point P3, the second extension line is parallel to the first extension line, and the first tangent line is parallel to the second tangent line; The intersection point of the first extension line and the second tangent line is the control point P1, and the intersection point of the second extension line and the first tangent line is the control point P2.

4. An electronic device, characterized in that include: At least one processor, and a memory communicatively connected to the at least one processor, wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor performs the method according to any one of claims 1 to 3.

5. A storage medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, the method according to any one of claims 1 to 3 is implemented.

Images