Structure optimization method for speed reducer box body

Abstract

The invention relates to a structure optimization method of a speed reducer box body, which comprises the following steps of: carrying out static simulation and modal analysis on the speed reducer box body on the basis of stress of each part of a gearbox in a force transmission form of gear transmission without axial force under an actual working condition, and extracting a weak structure of the speed reducer box body under static and dynamic simulation calculation; response surface optimization is carried out for the weak structure of the speed reducer box body, and an optimized box body model is obtained; the design variables of response surface optimization are the thicknesses of two output end bearing seats of a main shell in a reducer box body, the lengths of all reinforcing ribs around an input shell, the thickness of a face gear bearing supporting plate in the main shell and the thickness of an input shaft bearing seat; the target variables are the maximum deformation amount of the box body, the maximum linear stress value of the bearing supporting seat and the mass of the box body; and multi-target topological optimization is carried out on the basis of the reducer box body after the response surface optimization, and design variables obtained by the installation and direct force bearing part and the response surface optimization are used as non-optimization areas in the boundary condition setting of the multi-target topological optimization.

Description

Technical Field

[0001] The present invention relates to the field of optimization design of a reducer housing, and in particular to an optimization design method for a reducer housing that uses a finite element method to integrate response surface optimization and topology optimization design. Background Art

[0002] As a widely used energy transmission device in mechanical equipment, the design level of the reducer housing reflects a country's comprehensive national strength and industrial technology level to a certain extent, especially in the fields of aerospace, navigation, ships and weapons. The complex working environment and strict design indicators put forward higher requirements on the noise level, vibration deformation and lightweight of the reducer. This undoubtedly forces the reducer housing to develop in the direction of reducing its own weight while improving the structural strength. The traditional design method of the reducer housing can only meet its most basic functionality such as transmission and lubrication, and cannot achieve precise strength optimization and lightweight design.

[0003] With the rapid development of science and technology, finite element methods have gradually matured. People have begun to use finite element methods to achieve structural optimization design of reducer housings. Reducer housings are complex structural parts. The weight of bearing seats, ribs and other structures is small, but they are the main load-bearing parts; the weight of non-load-bearing parts of the housing is large, which is the main consideration for lightweight design. However, the current housing optimization methods using finite element technology do not distinguish them. Generally, only a single optimization method is used to carry out structural design. For example, CN111259491B and CN111611732B patents only carry out topological optimization design for gearboxes. Although the former performs secondary topological optimization, it still ignores the strength design of the main load-bearing parts or deformation-sensitive parts of the gearbox. CN113609738A patents only carry out response surface optimization design for missile housings to achieve the purpose of lightweight housings, without considering the strength analysis and optimization of the main load-bearing parts of the housings. Similarly, patent CN114154371A only performs response surface optimization design on three structural parameters of the reinforced gearbox. From the results, it can be seen that the total mass of the box is not optimized much, and the purpose of both improving strength and reducing mass is not achieved.

[0004] Because the reducer housing has a complex structure, existing patents, when establishing a parametric model of the housing, mostly parameterize relatively independent dimensions such as the housing wall thickness and rib thickness, and often ignore the analysis of the dimensions of the bearing seat, which is the direct force-bearing part. Because the bearing seat is located inside the housing, it has many associated dimensions and is difficult to parameterize. Summary of the invention

[0005] The technical problem solved by the present invention is to solve the problem of neglecting one thing while focusing on another in the current optimization methods. The present invention proposes a reducer case structure optimization method that uses finite element methods, a comprehensive response surface optimization and topology optimization design based on the static simulation and modal analysis of the reducer case, which can achieve the purpose of improving the case structure strength while reducing the mass.

[0006] The technical solution of the present invention is: a method for optimizing the structure of a reducer housing, comprising:

[0007] Based on the forces on various parts of the gearbox under the actual working conditions of the gear transmission without axial force, the static simulation and modal analysis of the reducer housing are carried out to extract the weak structure of the reducer housing under static and dynamic simulation calculations;

[0008] Response surface optimization was carried out for the weak structure of the reducer housing to obtain the optimized housing model; the design variables of the response surface optimization were the thickness of the bearing seats at the two output ends of the main housing in the reducer housing, the length of all reinforcing ribs around the input housing, the thickness of the face gear bearing support plate inside the main housing, and the thickness of the input shaft bearing seat; the target variables were the maximum deformation of the housing, the maximum linear stress of the bearing support seat, and the housing mass;

[0009] Multi-objective topology optimization is performed on the basis of the reducer housing after response surface optimization. In the boundary condition setting of the multi-objective topology optimization, the installation and direct load-bearing parts and the design variables obtained by response surface optimization are used as non-optimized areas.

[0010] Preferably, the forces on various parts of the gearbox are determined in the following manner: first, a dynamic model of the face gear transmission system inside the reducer housing is established, the speed and load under actual load conditions are used as model input conditions, the dynamic model is solved, and the radial support reaction force at the face gear bearing position and the axial and radial support reaction force data at the other bearings are extracted. The support reaction force data of the above-mentioned bearings are the force data of the gearbox.

[0011] Preferably, the target variable in the response surface optimization is determined by:

[0012] A three-dimensional model of the reducer housing is established and parameterized. The parameterized model is used for finite element statics simulation. The linear stress is drawn along the deformation direction on the bearing support plate to obtain the maximum deformation of the housing and the maximum linear stress value of the bearing support plate.

[0013] Preferably, before optimization, a modal analysis is performed on the reducer housing to extract the total deformation of the first ten modes and the first-order modal frequency; after multi-objective topology optimization, a modal analysis is performed again to extract the total deformation of the first ten modes and the first-order modal frequency, and based on the results of the two extractions, it is verified from the perspective of modal dynamics whether the optimized housing has improved dynamic stiffness.

[0014] Preferably, the response surface optimization method is an optimized space filling design method, which uniformly generates a preset number of sample point data, obtains a sensitivity bar graph of the design variable to the target variable through fitting calculation, adjusts the sample point data, and refines the design variable sample points with a sensitivity greater than 0.5 according to the bar graph; calculates the refined sample points to obtain the target variable value corresponding to each sample point; adopts a multi-objective genetic algorithm to select the sample point data that contributes the most to the target variable; reconstructs the reducer housing model according to the selected sample point data to obtain the housing model;

[0015] The sample point data are uniformly selected and generated within the required design variable variation range based on the initial values ​​of the reducer housing design variables before optimization.

[0016] Preferably, the preset number of groups is 80-120 groups.

[0017] Preferably, the design variable variation range is 10%-15%.

[0018] Preferably, the optimization objectives of the multi-objective topology optimization are maximizing the first-order modal frequency and minimizing the structural flexibility.

[0019] Preferably, the constraint condition of the multi-objective topology optimization is to retain 60-70% of the box mass, that is, the calculation is stopped when the variable density method is iterated to 60-70% of the original model.

[0020] Preferably, the maximum number of iterations is set to 500, the convergence accuracy is set to 0.1%, and the penalty factor is set to 3 in the analysis settings of the multi-objective topology optimization.

[0021] The beneficial effects of the present invention compared with the prior art are:

[0022] The present invention carries out structural optimization design for the reducer housing. The response surface optimization can accurately iterate and optimize the weak points and local dimensional parameters of the housing structure, which can effectively improve the structural strength of the weak points; the topological optimization can redistribute the material density of the entire model, effectively reducing the overall mass. The present invention couples the above two optimization methods to carry out comprehensive optimization, which not only effectively improves the strength of the weak structure of the housing, including reducing the maximum deformation, the maximum stress value and increasing the modal frequency, but also effectively reduces the overall mass while ensuring the overall strength of the housing.

[0023] The present invention can achieve the purpose of both improving the local strength of the housing and effectively reducing the overall mass, and has a significant improvement effect on the safety and reliability of the reducer housing under harsh working conditions. The basic technology is relatively mature and easy to implement.

[0024] The present invention abandons the method of selecting design variables based mainly on experience in traditional structural optimization design. The stress and deformation of the reducer housing under actual working conditions are intuitively obtained by means of static simulation, and the dimensional parameters of the structure with weaker rigidity are selected as design variables. This design variable selection method is more intuitive and accurate.

[0025] The present invention aims at the structural complexity of the face gear reducer housing, and the setting is made on the basis of fully considering the convergence of the calculation and the amount of calculation. Meanwhile, in the parameter setting of the topology optimization calculation, the distortion of the topology optimization model obtained based on the variable density method is taken into account. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] Figure 1 Flow chart of the comprehensive optimization design method for the reducer housing;

[0027] Figure 2 Simplified model diagram of reducer housing;

[0028] Figure 3 Histogram of sensitivity of design variables to target variables;

[0029] Figure 4 Reconstructed reducer housing model diagram after response surface optimization;

[0030] Figure 5 Finite element model material density distribution cloud map;

[0031] Figure 6 The reducer housing model is reconstructed after topology optimization. DETAILED DESCRIPTION

[0032] The present invention will be further described below in conjunction with the embodiments.

[0033] like Figure 1 As shown, the comprehensive optimization design method for the reducer housing in this embodiment includes the following steps:

[0034] 1. Based on the force on each part of the reducer housing under the unique force transmission form of gear transmission without axial force under actual working conditions, the optimized design is carried out for the weak links of the housing in the radial force direction;

[0035] 2. Use static simulation and modal analysis of the reducer housing as pre-processing for optimization;

[0036] (1) Simplification of the 3D model of the reducer housing: Simplify the structures that have little impact on the load-bearing path and dynamic characteristics of the housing, such as removing the oil pipes and observation holes on the housing; remove all bolt connections on the housing and replace them with part merging. The simplified model is as follows: Figure 2 shown.

[0037] (2) Finite element model establishment: Import the simplified box model into ANSYS Workbench, create a "staticstructural", define material properties, set boundary conditions, apply constraints such as fixed pairs based on the actual installation of the reducer box, and divide the tetrahedral mesh. The mesh size of this example is 0.005m, and the total number of meshes is 942797.

[0038] (3) Calculation and application of bearing support forces at various parts of the reducer housing: Build a dynamic model of the face gear transmission system inside the reducer housing in Adams, and complete the corresponding material property settings and motion constraint settings. Set a rotating pair for the face gear and the spur gear that matches it, and set a rotating pair for the planetary differential gear at the output end with the face gear as a reference; complete the collision parameter settings, build a rigid-flexible coupling relationship, and set a rigid-flexible coupling bearing support reaction force model at the bearing position of the above-mentioned transmission system model; apply the rated load and speed, solve the dynamic model, extract the radial support reaction force at the face gear bearing position and the axial and radial support reaction force data at the other bearings, and do not consider the change in phase angle. The amplitude of the above support reaction force is used as the bearing support reaction force and applied to the corresponding bearing installation position of the housing finite element model.

[0039] (4) Static simulation analysis: The above finite element model is solved to obtain the stress cloud map and deformation cloud map of the reducer housing under the reaction force of the above bearing support.

[0040] (5) Modal analysis: The above finite element model is inherited into the modal analysis module "modal". According to the actual installation situation, constraints are imposed to calculate the first ten natural vibration modes and corresponding natural frequencies of the reducer housing. The total deformation of the first ten modes is extracted in the solution, and the first-order modal frequency is extracted.

[0041] 3. Carry out response surface optimization design for the weak structure of the reducer housing. The specific steps are as follows:

[0042] (1) Analysis of weak structure of housing and extraction of design variables: Analyze the finite element model calculation results in step 2 (4) and (5), and extract the structural dimensions of the main load-bearing and deformation parts such as the bearing seat and rib plate as design variables, as shown in Table 1.

[0043] Table 1 Description of design structure parameters

[0044]

[0045]

[0046] (2) Model parameterization: The parameterization of the weak structural dimensions of the above-mentioned box body is implemented in the three-dimensional software Ug, and a model connection is established with the ANSYS workbench. The parameterized dimensions are used as design variables for response surface optimization, and the variation range of the design variables is 10%.

[0047] (3) Selecting target variables and establishing a response surface optimization mathematical model: In order to improve the stiffness level of the weak structure, in the result processing of step 2 (4), the linear stress is drawn along the deformation direction on the bearing support plate, and the maximum linear stress value, the maximum deformation of the box and the box mass are extracted as the response surface optimization target variables. The response surface optimization mathematical model established is:

[0048] Find X=(x1,x2,x3,x4,x5)

[0049] F Min {m1(x),m2(x),Mass)

[0050] 28.35≤x1≤34.65

[0051] 63≤x2≤77

[0052] 20.7≤x3≤25.3

[0053] 24.3≤x4≤29.7

[0054] 26.55≤x6≤32.45

[0055] x1, x2, x3, x4, and x5 correspond to the design variable parameters DS_FLHD, DS_LBCD, DS_LSHD, DS_ZCHD, and DS_ZDZC respectively; m1(x) represents the maximum deformation of the box, m2(x) represents the maximum stress value, and Mass represents the mass of the box.

[0056] (4) Response surface calculation and sensitivity analysis: The response surface optimization experimental design method is an optimized space filling design method. 100 sets of sample point data are uniformly generated within the variation range of the above design variables. The sensitivity bar graph of the design variables to the target variables is obtained through fitting calculation, such as Figure 3 As shown in the figure, the design variables DS_LSHD and DS_ZCHD have a greater impact on the three target variables, both greater than 0.8. The two groups of design variable sample points are refined, and the refined sample points are calculated to obtain the target variable values ​​corresponding to each group of sample points.

[0057] (5) Optimal sample point screening and model reconstruction: The above sample point screening method is a multi-objective genetic algorithm, which selects the sample point data that contributes the most to the target variable, as shown in Table 2, and uses this set of parameters to reconstruct the model in Ug.

[0058] Table 2 Optimization results of design variables

[0059]

[0060] (6) Comparative analysis of target variables after optimization: The above-mentioned reconstructed model is subjected to the same static simulation analysis and modal analysis as in step 2, and the comparison of the target variable results is shown in Table 3. In the response surface optimization for the weak structure, the maximum deformation of the box is reduced by 14.1%, the maximum stress is reduced by 13%, the first-order modal frequency is increased by 3.5%, and the mass of the box remains basically unchanged.

[0061] Table 2 Optimization results of target variables

[0062]

[0063] 4. Based on the above reconstructed model after response surface optimization, multi-objective topology optimization design of the reducer housing is carried out.

[0064] (1) Topology optimization preprocessing: The reconstructed box model obtained by response surface optimization in step 3(5) is imported into the Topology Optimization module of ANSYS. The preprocessing operations such as setting material properties, meshing and boundary conditions are the same as step 1.

[0065] (2) Topology analysis settings: In the topology optimization analysis settings, set the maximum number of iterations to 500, the convergence accuracy to 0.1%, and the penalty factor to 3 (a too small value will affect the intermediate density filtering effect, and a too large value will easily lead to non-convergence).

[0066] (3) Optimization area selection: In the optimization area setting, the structural parts selected as design variables during response surface optimization and the housing installation location are set as non-optimization areas, and the rest of the reducer housing is set as the optimization area.

[0067] (4) Constraint setting: In order to more clearly display the optimized area, the constraint condition in the topology optimization is set to retain 60% of the box mass, that is, the calculation is stopped when the variable density method is iterated to 60% of the original model.

[0068] (5) Optimization target setting: The optimization target is set to maximize the first-order modal frequency and minimize the structural flexibility.

[0069] (6) Generation of optimized model: After topology optimization iterative calculation, the model material density distribution cloud map is obtained as follows: Figure 5 As shown in the figure, the reducer housing is reconstructed twice, the functionality and sealing of the housing are considered, and the over-optimized parts are discarded. The secondary reconstruction model is as follows Figure 6 shown.

[0070] (7) Comparative analysis after optimization: Finite element analysis and modal analysis were performed on the secondary reconstruction model. The results showed that the mass of the reducer housing was reduced by 5.41%, the maximum deformation under the same working conditions was reduced by 13.85%, the maximum stress value was reduced by 7.65%, and the first-order modal frequency was increased by 4.62%.

[0071] The parts not elaborated in detail in the present invention belong to the known technologies in the art. Any technical solutions formed by equivalent transformation or equivalent replacement shall fall within the protection scope of the present invention.

[0072] Although the present invention has been disclosed as above in the form of a preferred embodiment, it is not intended to limit the present invention. Any person skilled in the art may make possible changes and modifications to the technical solution of the present invention by using the methods and technical contents disclosed above without departing from the spirit and scope of the present invention. Therefore, any simple modifications, equivalent changes and modifications made to the above embodiments based on the technical essence of the present invention without departing from the content of the technical solution of the present invention shall fall within the protection scope of the technical solution of the present invention.

[0073] Parts not described in detail in the present invention belong to common knowledge of those skilled in the art.

Claims

1. A method for optimizing the structure of a reducer housing, characterized in that include: Based on the forces on various parts of the gearbox under the actual working conditions of the gear transmission without axial force, the static simulation and modal analysis of the reducer housing are carried out to extract the weak structure of the reducer housing under static and dynamic simulation calculations; Response surface optimization was carried out for the weak structure of the reducer housing to obtain the optimized housing model; the design variables of the response surface optimization were the thickness of the bearing seats at the two output ends of the main housing in the reducer housing, the length of all reinforcing ribs around the input housing, the thickness of the face gear bearing support plate inside the main housing, and the thickness of the input shaft bearing seat; the target variables were the maximum deformation of the housing, the maximum linear stress of the bearing support seat, and the housing mass; Multi-objective topology optimization is performed on the basis of the reducer housing after response surface optimization. In the boundary condition setting of the multi-objective topology optimization, the installation and direct load-bearing parts and the design variables obtained by response surface optimization are used as non-optimized areas.

2. The method according to claim 1, characterized in that: The forces on various parts of the gearbox are determined in the following way: first, a dynamic model of the face gear transmission system inside the reducer housing is established, and the speed and load under actual load conditions are used as model input conditions. The dynamic model is solved to extract the radial support reaction force at the face gear bearing position and the axial and radial support reaction force data at the other bearings. The support reaction force data of the above-mentioned bearings are the force data of the gearbox.

3. The method according to claim 1, characterized in that: The target variable in response surface optimization is determined in the following way: A three-dimensional model of the reducer housing is established and parameterized. The parameterized model is used for finite element statics simulation. The linear stress is drawn along the deformation direction on the bearing support plate to obtain the maximum deformation of the housing and the maximum linear stress value of the bearing support plate.

4. The method according to claim 1, characterized in that: Before optimization, the modal analysis of the reducer housing was performed to extract the total deformation of the first ten modes and the first-order modal frequency. After multi-objective topology optimization, the modal analysis was performed again to extract the total deformation of the first ten modes and the first-order modal frequency. Based on the results of the two extractions, it was verified from the perspective of modal dynamics whether the optimized housing had improved dynamic stiffness.

5. The method according to claim 1, characterized in that: The response surface optimization method is an optimized space filling design method, which uniformly generates a preset number of sample point data, obtains a sensitivity bar graph of the design variable to the target variable through fitting calculation, adjusts the sample point data, and refines the design variable sample points with a sensitivity greater than 0.5 according to the bar graph; calculates the refined sample points to obtain the target variable value corresponding to each sample point; adopts a multi-objective genetic algorithm to select the sample point data that contributes the most to the target variable; reconstructs the reducer housing model according to the selected sample point data to obtain the housing model; The sample point data are uniformly selected and generated within the required design variable variation range based on the initial values ​​of the reducer housing design variables before optimization.

6. The method according to claim 1, characterized in that: The preset number of groups is 80-120 groups.

7. The method according to claim 1, characterized in that: The design variable variation range is 10%-15%.

8. The method according to claim 1, characterized in that: The optimization objectives of the multi-objective topology optimization are maximizing the first-order modal frequency and minimizing the structural flexibility.

9. The method according to claim 1, characterized in that: The constraint condition of the multi-objective topology optimization is to retain 60-70% of the box mass, that is, the calculation is stopped when the variable density method is iterated to 60-70% of the original model.

10. The method according to claim 1, characterized in that: In the analysis settings of multi-objective topology optimization, the maximum number of iterations is set to 500, the convergence accuracy is 0.1%, and the penalty factor is 3.

Images