A lightweight method for commercial vehicle cab based on fatigue durability

Abstract

The present invention discloses a lightweight method for a commercial vehicle cab based on fatigue durability, which mainly consists of data collection and processing, model establishment and verification, virtual iteration, fatigue simulation analysis, lightweight design and verification. The present invention combines three fatigue research methods, namely, prototype vehicle durability road test, finite element fatigue simulation and physical bench test, and applies them to engineering practice, providing an important reference for future automotive system-level durability and lightweight research.

Description

Technical Field

[0001] The invention relates to the field of vehicle fatigue durability analysis and lightweighting, and in particular to a commercial vehicle cab lightweighting method based on fatigue durability. Background Art

[0002] Lightweighting of automobiles is not simply about reducing the weight of the automobile. It is based on the premise that the overall quality, performance and cost of the automobile remain unchanged or even optimized. This means that the weight of the entire vehicle should be reduced to the maximum extent through advanced design methods, appearance and structural improvements, parts replacement or the use of new materials, so as to improve the power of the automobile, reduce fuel consumption and reduce exhaust pollution.

[0003] Durability and lightweight are the main topics in the field of automobile design. How to design a lightweight vehicle while ensuring its durability is a major issue in today's automobile design. At present, the lightweight design of the cab usually takes the mass, strength or stiffness of the white body as the optimization target, and there are few multi-target lightweight designs that comprehensively consider the performance indicators of the cab. Summary of the invention

[0004] In view of this, an object of the present invention is to provide a method for lightweighting a commercial vehicle cab based on fatigue durability.

[0005] The specific technical solution adopted by the present invention is:

[0006] A lightweight method for commercial vehicle cab based on fatigue durability mainly consists of data collection and processing, model establishment and verification, virtual iteration, fatigue simulation analysis, lightweight design and verification.

[0007] Data collection and processing. The purpose of road spectrum collection is to obtain the displacement and acceleration signals of the cab airbag bracket on the test site. Then use it as the expected signal for virtual iteration to obtain the fatigue analysis load spectrum. A domestic commercial vehicle was selected as the research object. According to the relevant test specifications of the enterprise user, the sensor was installed on the vehicle and the test reinforced road surface was selected. The road spectrum data was collected using a dynamic data collector to ensure that three sets of valid data were collected for each reinforced road surface. The test site reinforced road surfaces such as cobblestone roads, twisted roads, stone roads, and corrugated roads were selected for testing. The road load spectrum was collected using the HBM Somat EDAQ dynamic data collector. According to the test specifications, the sampling frequency was set to 512Hz. Three sets of valid data were collected for each road condition. The most stable set was selected as the expected signal for virtual iteration based on the Rossow sampling principle. Due to the impact vibration of the unevenness of the test site road surface and the electromagnetic interference of the vehicle-related wiring harness, the selected collected data cannot be directly used for virtual iteration. It is necessary to use nCode to perform Trend Remove on the data to eliminate data drift, so as to ensure the effectiveness of the fatigue test.

[0008] Establishment and verification of the model. The establishment of the model includes the finite element model and the rigid-flexible coupling multi-body dynamics model. The accuracy of the finite element model is a key factor affecting fatigue analysis. In order to accurately simulate the fatigue durability performance of the cab under full load conditions on the durable road surface of the test site, the weight of the occupants and luggage in the car during the road test on the test site is fully considered, and the model is loaded with counterweights. The establishment of the cab finite element model and its meshing are completed in HyperMesh, and the center of mass and moment of inertia of the model are compared with the measured values ​​of the moment of inertia test bench to verify the accuracy of the finite element model. The establishment of the cab finite element model is to use the inertia release method to solve its stress field distribution. When under the action of external force, if the structural damping is ignored, the equilibrium equation of the inertia release analysis is:

[0009]

[0010] Where: F is the external load matrix of the model nodes; M is the mass matrix of the model; is the velocity matrix of the model nodes.

[0011] X, Y and Z element forces and moments are applied to the connection between the cab mount and the frame, and the stress response is calculated by the Optistruct solver. The stress time history is calculated as follows:

[0012]

[0013] In the formula: ij , a(t) is the stress history; k is the load quantity; σ ij，a，k is the stress response of static load at point ij; P k，FEAis the magnitude of the static load force.

[0014] Specifically, X, Y, and Z unit forces and unit moments are applied to the four suspension points of the cab. Each unit load is set as a load case, for a total of 24 load cases. These unit load conditions established correspond to the fatigue load spectrum channels of the virtual iteration.

[0015] A rigid-flexible coupling multi-body dynamics model is established. According to the load transfer path, the road excitation is transmitted to the cab through the frame and rubber bushing. Under the road excitation of the test site, the front half of the frame connected to the airbag suspension has a greater impact on the dynamic response of the cab, so the front half of the frame is taken. The intercepted frame is meshed and the frame MNF flexible body file is output based on the Craig-Bampton modal synthesis method. The generated MNF file is then imported into Adams / car to generate a rigid-flexible coupling multi-body dynamics model, and the input and output channels are set corresponding to the installation position of the test sensor in the test site to prepare for subsequent virtual iterations. The flexible body dynamics differential equation obtained in Adams based on the second-kind Lagrangian equation for the flexible processing of the frame can be expressed as:

[0016]

[0017] Where: M, K, D are the mass matrix, modal stiffness matrix and modal damping matrix of the flexible body respectively; ζ is the generalized coordinate vector; Q is the additional load vector; G is gravity; ψ is the constraint equation vector; λ is the Laplace multiplier vector corresponding to ψ.

[0018] Virtual iteration and verification. The essence of virtual iteration is to invert the unknown from the known, approximate the nonlinear system through the linear transfer function, and invert the external excitation load with the response signal collected from the test field. The process of solving the load spectrum is as follows: first, generate a random white noise signal to drive the multi-body dynamics model, and combine the response signal of the corresponding channel to derive the frequency response transfer function of the system.

[0019]

[0020]

[0021]

[0022] Where x(t) and x(ω) represent the time domain signal and frequency domain signal of white noise respectively; y(t) and y(ω) represent the time domain response signal and frequency domain response signal under white noise excitation respectively; H(ω) is the frequency response transfer function of multibody dynamics.

[0023] According to classical control theory, the external excitation can be calculated by the response signal of the test field and the frequency response transfer function of the system. However, the multi-body dynamic system is nonlinear and the transfer function is linear. There is an error between the external excitation obtained by inverse calculation and the actual load, so an iterative step is introduced to reduce the error.

[0024]

[0025]

[0026]

[0027] Where: k is the correction value. Using the initial excitation signal x 0 (t) Drive the multi-body dynamic system to obtain the response signal y 0 (t), and compare it with the test acquisition signal y Des (t) is compared. The excitation signal is corrected by the response deviation to obtain the excitation signal x 1 (t) is the next iteration.

[0028] e 0 (t) = y Des (t)-y 0 (t) (10)

[0029]

[0030]

[0031] x 1 (t) = x 0 (t)+a·△ 0 (t) (13)

[0032] where e 0 (t) is the response deviation value; Δ 0 is the correction amount of the excitation signal; a is the correction coefficient of the multi-body system. Through multiple iterations, the response signal is made infinitely close to the expected signal, and finally the external excitation of the system is obtained.

[0033] The accuracy of virtual iteration will directly affect the reliability of fatigue load spectrum, and its accuracy is judged from three aspects: relative damage value, time domain and frequency domain. Relative damage value refers to the approximate degree of fatigue damage caused by two different load spectra to the same structure. By calculating the pseudo-damage ratio between the iterative signal and the measured signal, if the relative damage of each channel is close to 1, it indicates that the iteration quality is high. It is believed that the relative damage value between 0.5 and 2 can basically meet the engineering requirements.

[0034] Fatigue simulation analysis. The cab of a commercial vehicle is mainly composed of thin metal plates, which are subjected to alternating loads from the road during transportation. Fatigue failure includes not only elastic strain but also plastic strain, so the stress-strain fatigue analysis method is used. The stress-strain method is based on the rain flow counting method, which is developed from the Coffin-Manson theory and is used to predict the fatigue life of the system under uniaxial load. The relationship between the total material strain and fatigue life is:

[0035]

[0036] Where: ε a is the total strain amplitude; σ f and b represent the strength coefficient and strength index of the material respectively; E is the elastic modulus of the material, N f is the fatigue life. f and c represent the durability coefficient and durability index of the material respectively. However, Coffin-Manson is only applicable to material strain under uniaxial load. Under multiaxial load conditions, the function needs to be equivalently processed according to the fourth strength theory. The total strain ε after processing qa for:

[0037]

[0038] Where: e1a, e2a and e3a represent the first, second and third principal strains respectively; μ represents the Poisson’s ratio of the material. Since the material has different Poisson’s ratios in the elastic and plastic deformation stages, μ is used respectively. e and μ p The equivalent total strain ε qa By elastic strain ε eqa and the plastic strain ε pqa It consists of two parts, and its expression is:

[0039]

[0040]

[0041] By integrating the above equation, the relationship between component strain and fatigue life can be modified to:

[0042]

[0043] When the stress amplitude of the structure is σ i When N i In the cycle, the damage caused by each cycle is D = 1 / N i If the number of cycles is n1, the structural damage is D1 = n1 / N i When the structure is subjected to σ1, σ2…, σn, the number of cycles corresponding to each stress amplitude is n1, n2…, nn , the damage caused by the above actions can be expressed as:

[0044]

[0045] According to the linear superposition criterion, it can be seen that the total damage under the above stress amplitude can be expressed as

[0046]

[0047] Once the total damage value D reaches 1, the structure will suffer fatigue failure. Combining the fatigue load spectrum, stress analysis results under unit load and material ε-N curve, a fatigue calculation simulation block was established in nCode for simulation. The simulation results were compared with the durability road test results to verify the authenticity. The results show that the simulated failure part is basically consistent with the road test.

[0048] Lightweight design and verification. Based on the sensitivity analysis theory, the design parameters are selected. In order to meet the lightweight requirements based on fatigue durability, the thickness of the body parts material is used as the design variable. The sensitivity of the body structure performance parameters to the component thickness variable is expressed as:

[0049]

[0050] Where S is the sensitivity, f(x) is the objective function, and x i is the sheet metal thickness variable.

[0051] The thickness of the body sheet metal is selected as a set of screening data variables, and the calculation results are the sensitivity of bending stiffness (BS), torsional stiffness (TS), first bending mode (BMS) and torsional mode (TMS) to the thickness variable. Through sensitivity calculation, the thickness of 20 components is determined as the optimization design variable. These components have high sensitivity to the objective function and low sensitivity to fatigue performance. They are selected as design variables for lightweight design because they meet the requirements of lightweight without sacrificing too much performance.

[0052] In order to improve the optimization efficiency, the Kriging approximation model is used to replace the original simulation model to express the relationship between the optimization variables and the response. According to the optimal Latin hypercube experimental design method, within the value range of each design variable, combined with the need for fitting accuracy, a total of 100 sample points are selected through uniform random sampling.

[0053] The accuracy of each approximation model is determined by the coefficient of 2 Perform the test, which is expressed as:

[0054]

[0055] Where n is the number of sampling points, y i and y represent the predicted value of the approximate model, the actual value of sampling point i, and y respectively. i The average value of R 2 The closer it is to 1, the more reliable the approximate model is.

[0056] The thickness of 20 sheet metal parts selected by sensitivity analysis is used as the design variable, and the mass, fatigue damage, stiffness and modal performance indicators are comprehensively considered. The particle swarm optimization algorithm is used to perform lightweight multi-objective optimization design of the cab. The optimization mathematical model can be described as follows:

[0057]

[0058] Where d(x) represents the fatigue damage value of the cab, m(x) represents the mass of the cab; k b (x) and k b0 is the bending stiffness value and its initial value; k t (x), k t0 is the torsional stiffness value and its initial value; f b (x), f b0 is the first-order bending mode and its initial value; f t (x), f t0 is the first-order torsional mode and its initial value; X L and X U are the upper and lower limits of the design variable value respectively.

[0059] The beneficial effect of the present invention is that the present invention integrates three fatigue research methods, namely, prototype vehicle durability road test, finite element fatigue simulation and physical bench test, and applies them to engineering practice, providing an important reference for future automotive system-level durability and lightweight research.

[0060] Compared with existing solutions, this method has the following advantages:

[0061] 1. In view of the poor correlation between the current fatigue analysis based on random road spectra and the durability road test on the test track, a fatigue durability analysis strategy combining the measured road spectrum with virtual simulation is proposed, which improves the speed of fatigue simulation solution and reproduces the road conditions of real users to the greatest extent. The fatigue simulation results are basically consistent with the results of the enhanced road durability test, which verifies that this method has good engineering practicality;

[0062] 2. Taking mass and fatigue damage value as optimization targets, bending-torsional stiffness and first-order bending-torsional vibration mode as constraints, the thickness is screened based on sensitivity analysis, and the cab is optimized by multi-objective design in combination with the Kriging approximation model and particle swarm algorithm to obtain the optimal solution set and determine the final lightweight solution;

[0063] 3. The durability performance was verified on a seven-axis road simulation test bench. The results show that the lightweight design of the cab can take into account both lightweight and durability. BRIEF DESCRIPTION OF THE DRAWINGS

[0064] Figure 1 The present invention provides a sensor arrangement of a test vehicle for a commercial vehicle cab lightweighting method based on fatigue durability.

[0065] Figure 2 is a safety airbag suspension displacement signal of a commercial vehicle cab lightweighting method based on fatigue durability provided by an embodiment of the present invention;

[0066] Figure 3 A finite element model of a cab of a commercial vehicle based on fatigue durability lightweighting method provided by an embodiment of the present invention;

[0067] Figure 4 A rigid-flexible coupling multi-body dynamics model of a commercial vehicle cab lightweight method based on fatigue durability provided by an embodiment of the present invention;

[0068] Figure 5 It is a virtual iterative relative damage value of each channel of a commercial vehicle cab lightweight method based on fatigue durability provided by an embodiment of the present invention;

[0069] Figure 6 It is a time domain and frequency domain comparison of a simulation value and a measured value of a commercial vehicle cab lightweight method based on fatigue durability provided by an embodiment of the present invention;

[0070] Figure 7 is a comparison between the test results and the simulation results of a commercial vehicle cab lightweighting method based on fatigue durability provided by an embodiment of the present invention;

[0071] Figure 8 A cab lightweight multi-objective optimization model for a commercial vehicle cab lightweight method based on fatigue durability provided by an embodiment of the present invention;

[0072] Fig. 9 A cab bench verification of a commercial vehicle cab lightweighting method based on fatigue durability provided by an embodiment of the present invention;

[0073] Fig.10 The present invention provides a method for lightweighting a commercial vehicle cab based on fatigue durability, and provides a frame test result after lightweight design. DETAILED DESCRIPTION

[0074] In order to explain the technical content, structural features, achieved objectives and effects of the technical solution in detail, the following is a detailed description in conjunction with specific embodiments and accompanying drawings.

[0075] A lightweight method for commercial vehicle cab based on fatigue durability mainly consists of data collection and processing, model establishment and verification, virtual iteration, fatigue simulation analysis, lightweight design and verification.

[0076] Data collection and processing. Figure 1 As shown in the figure, sensors are installed on the test vehicle, and HBM collectors are used to collect data on various reinforced roads. Due to the impact of road surface and electromagnetic interference of related wiring harnesses, the collected data must be preprocessed, such as Figure 2 is the processed displacement signal of the airbag suspension; the purpose of road spectrum acquisition is to obtain the displacement and acceleration signals of the cab airbag bracket on the test site. Then it is used as the expected signal of virtual iteration to obtain the fatigue analysis load spectrum. A domestic commercial vehicle was selected as the research object. According to the relevant test specifications of the enterprise user, the vehicle was equipped with sensors and the test reinforced road surface was selected. The road spectrum data was collected using a dynamic data collector to ensure that three sets of valid data were collected for each reinforced road surface. The test site reinforced road surfaces such as cobblestone roads, twisted roads, stone roads, and corrugated roads were selected for testing. The road load spectrum was collected using the HBM Somat EDAQ dynamic data collector. According to the test specifications, the sampling frequency was set to 512Hz. Three sets of valid data were collected for each road condition. The most stable set was selected as the expected signal for virtual iteration based on the Rossow sampling principle. Due to the impact vibration of the unevenness of the test site road surface and the electromagnetic interference of the vehicle-related wiring harness, the selected collected data cannot be directly used for virtual iteration. It is necessary to use nCode to perform Trend Remove on the data to eliminate data drift, so as to ensure the effectiveness of the fatigue test.

[0077] Model establishment and verification. The establishment of the model includes the finite element model and the rigid-flexible coupling multi-body dynamics model. The accuracy of the finite element model is a key factor affecting fatigue analysis. In order to accurately simulate the fatigue durability performance of the cab under full load conditions on the durable road surface of the test site, the weight of the occupants and luggage in the vehicle during the road test on the test site is fully considered, and the model is loaded with counterweights. The establishment of the finite element model of the cab and its meshing are completed in HyperMesh, and the calculated values ​​of the center of mass and moment of inertia of the model are compared with the measured values ​​of the moment of inertia test bench to verify the accuracy of the finite element model. Use HyperMesh to establish a commercial vehicle cab model and perform meshing, and then verify the accuracy of the model by comparing it with the center of mass and moment of inertia values ​​of the actual vehicle test. Figure 3 For the finite element model of the cab, in order to simplify the stress analysis of constraints and loading conditions, the inertia release method is used to balance the external forces and eliminate the errors caused by the constraint reaction forces in the static analysis.

[0078] The finite element model of the cab is established to solve its stress field distribution using the inertia release method. When the external force is applied, if the structural damping is ignored, the equilibrium equation of the inertia release analysis is:

[0079]

[0080] Where: F is the external load matrix of the model nodes; M is the mass matrix of the model; is the velocity matrix of the model nodes.

[0081] X, Y and Z element forces and moments are applied to the connection between the cab mount and the frame, and the stress response is calculated by the Optistruct solver. The stress time history is calculated as follows:

[0082]

[0083] In the formula: ij , a(t) is the stress history; k is the load quantity; σ ij，a，k is the stress response of static load at point ij; P k，FEA is the magnitude of the static load force.

[0084] Specifically, X, Y, and Z unit forces and unit moments are applied to the four suspension points of the cab. Each unit load is set as a load case, for a total of 24 load cases. These unit load conditions established correspond to the fatigue load spectrum channels of the virtual iteration.

[0085] A rigid-flexible coupling multi-body dynamic model is established. According to the load transfer path, the road excitation is transmitted to the cab via the frame and rubber bushing. Under the road excitation of the test field, the front half of the frame connected to the airbag suspension has a greater influence on the dynamic response of the cab. Therefore, the front half of the frame is taken, such as Figure 4 As shown in the figure, the intercepted frame is meshed and the frame MNF flexible body file is output based on the Craig-Bampton modal synthesis method. The generated MNF file is then imported into Adams / car to generate a rigid-flexible coupling multi-body dynamics model, and the input and output channels are set corresponding to the installation position of the test sensor in the test field to prepare for subsequent virtual iterations. The flexible body dynamics differential equation obtained in Adams based on the second-kind Lagrangian equation for the flexible processing of the frame can be expressed as:

[0086]

[0087] Where: M, K, D are the mass matrix, modal stiffness matrix and modal damping matrix of the flexible body respectively; ζ is the generalized coordinate vector; Q is the additional load vector; G is gravity; ψ is the constraint equation vector; λ is the Laplace multiplier vector corresponding to ψ.

[0088] Virtual iteration and verification. The essence of virtual iteration is to invert the unknown from the known, approximate the nonlinear system through the linear transfer function, and invert the external excitation load with the response signal collected from the test field. The process of solving the load spectrum is as follows: first, generate a random white noise signal to drive the multi-body dynamics model, and combine the response signal of the corresponding channel to derive the frequency response transfer function of the system.

[0089]

[0090]

[0091]

[0092] Where x(t) and x(ω) represent the time domain signal and frequency domain signal of white noise respectively; y(t) and y(ω) represent the time domain response signal and frequency domain response signal under white noise excitation respectively; H(ω) is the frequency response transfer function of multibody dynamics.

[0093] According to classical control theory, external excitation can be calculated by the response signal of the test field and the frequency response transfer function of the system. However, the multi-body dynamic system is nonlinear and the transfer function is linear. There is an error between the external excitation obtained by inverse calculation and the actual load, so an iterative step is introduced to reduce the error.

[0094]

[0095]

[0096]

[0097] Where: k is the correction value. Using the initial excitation signal x 0 (t) Drive the multi-body dynamic system to obtain the response signal y 0 (t), and compare it with the test acquisition signal y Des (t) is compared. The excitation signal is corrected by the response deviation to obtain the excitation signal x 1 (t) is the next iteration.

[0098] e 0 (t) = y Des (t)-y 0 (t) (10)

[0099]

[0100]

[0101] x 1 (t) = x 0 (t)+a·△0 (t) (13)

[0102] where e 0 (t) is the response deviation value; Δ 0 is the correction amount of the excitation signal; a is the correction coefficient of the multi-body system. Through multiple iterations, the response signal is infinitely close to the expected signal, and finally the external excitation of the system is obtained.

[0103] The accuracy of virtual iteration will directly affect the reliability of fatigue load spectrum, and its accuracy is judged from three aspects: relative damage value, time domain and frequency domain. Relative damage value refers to the approximate degree of fatigue damage caused by two different load spectra to the same structure. By calculating the pseudo-damage ratio between the iterative signal and the measured signal, if the relative damage of each channel is close to 1, it indicates that the iteration quality is high. It is believed that the relative damage value between 0.5 and 2 can basically meet the engineering requirements. Figure 5 is the relative damage value between the test signal and the target signal. It can be seen that the relative damage values ​​are between 0.5 and 2, which meets the actual needs of the project. Figure 6 This is the time domain comparison between the test value and the simulation value, and the overall trend is consistent;

[0104] Fatigue simulation analysis. The cab of a commercial vehicle is mainly composed of thin metal plates, which are subjected to alternating loads from the road during transportation. Fatigue failure includes not only elastic strain but also plastic strain, so the stress-strain fatigue analysis method is used. The stress-strain method is based on the rain flow counting method, which is developed from the Coffin-Manson theory and is used to predict the fatigue life of the system under uniaxial load. The relationship between the total material strain and fatigue life is:

[0105]

[0106] Where: ε a is the total strain amplitude; σ f and b represent the strength coefficient and strength index of the material respectively; E is the elastic modulus of the material, N f is the fatigue life. f and c represent the durability coefficient and durability index of the material respectively. However, Coffin-Manson is only applicable to material strain under uniaxial load. Under multiaxial load conditions, the function needs to be equivalently processed according to the fourth strength theory. The total strain ε after processing qa for:

[0107]

[0108] Where: e1a, e2a and e3a represent the first, second and third principal strains respectively; μ represents the Poisson’s ratio of the material. Since the material has different Poisson’s ratios in the elastic and plastic deformation stages, μ is used respectively. e and μp The equivalent total strain ε qa By elastic strain ε eqa and the plastic strain ε pqa It consists of two parts, and its expression is:

[0109]

[0110]

[0111] By integrating the above equation, the relationship between component strain and fatigue life can be modified to:

[0112]

[0113] When the stress amplitude of the structure is σ i When N i In the cycle, the damage caused by each cycle is D = 1 / N i If the number of cycles is n1, the structural damage is D1 = n1 / N i When the structure is subjected to σ1, σ2…, σn, the number of cycles corresponding to each stress amplitude is n1, n2…, n n , the damage caused by the above actions can be expressed as:

[0114]

[0115] According to the linear superposition criterion, it can be seen that the total damage under the above stress amplitude can be expressed as

[0116]

[0117] Once the total damage value D reaches 1, the structure will suffer fatigue damage. Combining the fatigue load spectrum, stress analysis results under unit load and material ε-N curve, a fatigue calculation simulation block was established in nCode for simulation. The simulation results were compared with the durability road test results to verify the authenticity. Figure 7 The fatigue failure diagram of the actual vehicle test components and their fatigue simulation cloud diagram were compared for the actual vehicle. The results showed that the failure part of the simulation was basically consistent with the road test.

[0118] Lightweight design and verification. Based on the sensitivity analysis theory, the design parameters are selected. In order to meet the lightweight requirements based on fatigue durability, the thickness of the body parts material is used as the design variable. The sensitivity of the body structure performance parameters to the component thickness variable is expressed as:

[0119]

[0120] Where S is the sensitivity, f(x) is the objective function, and x iis the sheet metal thickness variable.

[0121] The thickness of the body sheet metal is selected as a set of screening data variables, and the calculation results are the sensitivity of bending stiffness (BS), torsional stiffness (TS), first bending mode (BMS) and torsional mode (TMS) to the thickness variable. Through sensitivity calculation, the thickness of 20 components is determined as the optimization design variable. These components have high sensitivity to the objective function and low sensitivity to fatigue performance. They are selected as design variables for lightweight design because they meet the requirements of lightweight without sacrificing too much performance.

[0122] In order to improve the optimization efficiency, the Kriging approximation model is used to replace the original simulation model to express the relationship between the optimization variables and the response. According to the optimal Latin hypercube experimental design method, within the value range of each design variable, combined with the need for fitting accuracy, a total of 100 sample points are selected through uniform random sampling.

[0123] The accuracy of each approximation model is determined by the coefficient of 2 Perform the test, which is expressed as:

[0124]

[0125] Where n is the number of sampling points, y i and y represent the predicted value of the approximate model, the actual value of sampling point i, and y respectively. i The average value of R 2 The closer it is to 1, the more reliable the approximate model is. The determination coefficient R corresponding to the bending stiffness, torsional stiffness, first-order bending modal frequency and first-order torsional modal frequency 2 They are 0.957, 0.968, 0.949 and 0.963 respectively. The test results show that all approximate models meet the accuracy requirements.

[0126] The thickness of 20 sheet metal parts selected by sensitivity analysis is used as the design variable, and the mass, fatigue damage, stiffness and modal performance indicators are comprehensively considered. The particle swarm optimization algorithm is used to perform lightweight multi-objective optimization design of the cab. The optimization mathematical model can be described as follows:

[0127]

[0128] Where d(x) represents the fatigue damage value of the cab, m(x) represents the mass of the cab; k b (x) and k b0 is the bending stiffness value and its initial value; k t (x), k t0 is the torsional stiffness value and its initial value; f b (x), f b0is the first-order bending mode and its initial value; f t (x), f t0 is the first-order torsional mode and its initial value; X L and X U are the upper and lower limits of the design variable value respectively.

[0129] The optimization results show that the first-order bending and torsional frequencies increased by 1.82% and 1.13% respectively, the cab bending and torsional stiffness decreased by 0.11% and 0.23% respectively, and the mass was reduced by 53kg. Through optimization iterations, while reducing the mass, the performance changes were within an acceptable range, indicating that this method is effective in the lightweight optimization design of the car body. Figure 8 Finally, the lightweight cab is placed on the designed test bench for testing. Fig. 9 The bench test was used to reproduce the state of the cab on the test road, and the test was carried out according to the equivalent conversion between the road test and the bench test. The results are shown in Fig.10 It is shown that the fatigue durability of cab components can be guaranteed while meeting the requirements of lightweight design of the cab.

[0130] Although the above embodiments have been described, once those skilled in the art know the basic creative concepts, they can make additional changes and modifications to these embodiments. Therefore, the above description is only an embodiment of the present invention and does not limit the patent protection scope of the present invention. Any equivalent structure or equivalent process transformation made by using the contents of the specification and drawings of the present invention, or directly or indirectly used in other related technical fields, are also included in the patent protection scope of the present invention.

Claims

1. A method for lightweighting a commercial vehicle cab based on fatigue durability, characterized in that: Including data collection and processing, model establishment and verification, virtual iteration, fatigue simulation analysis and lightweight design and verification; The purpose of the virtual iteration is to inversely obtain the fatigue load spectrum of the cab component position based on the measured road spectrum. Whether it converges will directly affect the accuracy of the fatigue load spectrum. The accuracy of the virtual iteration is determined by the time domain and frequency domain. The process of solving the fatigue load spectrum is as follows: first, generate a random white noise signal to drive the multi-body dynamics model, and combine the response signal of the corresponding channel to derive the frequency response transfer function of the system: Where x(t) and x(ω) represent the time domain signal and frequency domain signal of white noise respectively; y(t) and y(ω) represent the time domain response signal and frequency domain response signal under white noise excitation respectively; H(ω) is the frequency response transfer function of multibody dynamics; Introduce an iterative step to reduce the error: Where: k is the correction value, using the initial excitation signal x 0 (t) Drive the multi-body dynamic system to obtain the response signal y 0 (t), and compare it with the test acquisition signal y Des (t) is compared and the excitation signal is corrected by the response deviation to obtain the excitation signal x 1 The next iteration of (t): e 0 (t)=y Des (t)-y 0 (t) x 1 (t)=x 0 (t)+a·Δ 0 (t) where e 0 (t) is the response deviation value; Δ 0 is the correction amount of the excitation signal; a is the correction coefficient of the multi-body system; through multiple iterations, the response signal is made infinitely close to the expected signal, and finally the external excitation of the system is obtained; The cab is designed to be lightweight based on the Kriging approximation model. First, the thickness of 20 components of the cab is selected as the optimization design variable by using sensitivity analysis. Then, the particle swarm optimization algorithm is used to perform multi-objective optimization design on the established Kriging cab model to achieve lightweight effect while meeting fatigue durability requirements. Taking the thickness of the body parts material as the design variable, the sensitivity of the body structure performance parameters to the component thickness variable is expressed as: Where S is the sensitivity, f(x) is the objective function, and x i is the sheet metal thickness variable; The thickness of the body sheet metal is selected as a set of screening data variables, and the calculated results are the sensitivity of bending stiffness (BS), torsional stiffness (TS), first bending mode (BMS) and torsional mode (TMS) to the thickness variable; The Kriging approximation model is used to replace the original simulation model to express the relationship between the optimization variables and the response. According to the optimal Latin hypercube experimental design method, within the value range of each design variable, combined with the need for fitting accuracy, a total of 100 sample points are selected through uniform random sampling; The accuracy of each approximation model is determined by the coefficient of 2 Perform the test, which is expressed as: Where n is the number of sampling points, y i and y represent the predicted value of the approximate model, the actual value of sampling point i, and y respectively. i The average value of Taking the thickness of 20 sheet metal parts selected by sensitivity analysis as design variables, the particle swarm optimization algorithm is used to perform lightweight multi-objective optimization design of the cab by comprehensively considering the mass, fatigue damage, stiffness and modal performance indicators. The optimization mathematical model can be described as follows: Where d(x) represents the fatigue damage value of the cab, m(x) represents the mass of the cab; k b (x) and k b0 is the bending stiffness value and its initial value; k t (x), k t0 is the torsional stiffness value and its initial value; f b (x), f b0 is the first-order bending mode and its initial value; f t (x), f t0 is the first-order torsional mode and its initial value; X L and X U are the upper and lower limits of the design variable value respectively.

2. A method for lightweighting a commercial vehicle cab based on fatigue durability according to claim 1, characterized in that: The purpose of the data acquisition is to obtain the displacement and acceleration signals of the cab airbag bracket on the test site, and then use them as the expected signals for virtual iteration; the acquired data cannot be directly used for virtual iteration, and nCode needs to be used to perform Trend Remove on the data to eliminate data drift, so as to ensure the effectiveness of the fatigue test.

3. A method for lightweighting a commercial vehicle cab based on fatigue durability according to claim 2, characterized in that: The establishment of the model includes the establishment of a cab finite element model and the establishment of a rigid-flexible coupling multi-body dynamics model; The finite element model of the cab is established to solve its stress field distribution using the inertia release method. When the external force is applied, if the structural damping is ignored, the equilibrium equation of the inertia release analysis is: Where: F is the external load matrix of the model node; M is the mass matrix of the model; is the velocity matrix of the model nodes; X, Y and Z element forces and moments are applied to the connection between the cab mount and the frame. The stress response is calculated by the Optistruct solver. The stress time history is calculated as follows: In the formula: ij , a(t) is the stress history; k is the load quantity; σ ij，a，k is the stress response of static load at point ij; P k，FEA is the magnitude of the static load force; Specifically, X, Y, and Z unit forces and unit moments are applied to the four suspension points of the cab, and each unit load is set as a load case, for a total of 24 load cases. These unit load conditions established correspond to the fatigue load spectrum channels of the virtual iteration; A rigid-flexible coupling multi-body dynamics model is established. The front half of the frame is taken, the intercepted frame is meshed, and the frame MNF flexible body file is output based on the Craig-Bampton modal synthesis method. The generated MNF file is then imported into Adams / car to generate a rigid-flexible coupling multi-body dynamics model. The input and output channels are set corresponding to the installation positions of the test sensors in the test field to prepare for subsequent virtual iterations. The flexible body dynamics differential equation obtained by the flexible processing of the frame in Adams based on the second-kind Lagrangian equation can be expressed as: Where: M, K, D are the mass matrix, modal stiffness matrix and modal damping matrix of the flexible body respectively; ζ is the generalized coordinate vector; Q is the additional load vector; G is gravity; ψ is the constraint equation vector; λ is the Lagrangian multiplier vector corresponding to ψ.

4. A method for lightweighting a commercial vehicle cab based on fatigue durability according to claim 3, characterized in that: The fatigue simulation analysis is based on nCode fatigue analysis software, adopts stress-strain fatigue analysis method, establishes fatigue calculation simulation block in nCode software for simulation, and compares the simulation results with the actual vehicle test results to verify the accuracy of the simulation results; the relationship between the total strain of the material and the fatigue life is: Where: ε a is the total strain amplitude; σ f and b represent the strength coefficient and strength index of the material respectively; E is the elastic modulus of the material, N f is fatigue life; f and c represent the durability coefficient and durability index of the material respectively; Coffin-Manson is only applicable to material strain under uniaxial load. Under multiaxial load conditions, the function needs to be equivalently processed according to the fourth strength theory. The total strain ε after processing qa for: Where: e1a, e2a and e3a represent the first, second and third principal strains respectively; μ represents the Poisson’s ratio of the material; Since the material has different Poisson’s ratios in the elastic and plastic deformation stages, μ is used respectively. e and μ p To express; equivalent total strain ε qa By elastic strain ε eqa and the plastic strain ε pqa It consists of two parts, and its expression is: By integrating the above equation, the relationship between component strain and fatigue life can be modified to: When the stress amplitude of the structure is σ i When N i In the cycle, the damage caused by each cycle is D = 1 / N i； If the number of cycles is n1, the structural damage is D1 = n1 / N i ; When the structure is subjected to σ1, σ2…, σn, the number of cycles corresponding to each stress amplitude is n1, n2…, n n , the damage caused by the above actions can be expressed as: According to the linear superposition criterion, the total damage under the above stress amplitude is expressed as Once the total damage value D reaches 1, the structure will suffer fatigue failure. Combining the fatigue load spectrum, stress analysis results under unit load and material ε-N curve, a fatigue calculation simulation block was established in nCode for simulation, and compared with the durability road test results to verify the authenticity of the simulation results.

Images