A method for simulating the ride comfort of a vehicle under non-stationary conditions

Abstract

The application provides a simulation method for automobile ride comfort under non-stationary conditions, and relates to the technical field of automobile research and development. An automobile model is established by using a finite element method, boundary conditions are applied, and a bdf file is generated; a transient spatial frequency response function method is combined with a spatial domain virtual excitation method to solve; new delay times of front and rear wheels are calculated, corresponding statements in the bdf file are modified, new initial frequencies and frequency intervals are calculated, and corresponding statements in the bdf file are modified. The application can not only simulate the ride comfort of a simple model, but also simulate the ride comfort of a complex model, and the solving efficiency is high without the need of deducing complex mathematical formulas.

Description

Technical Field

[0001] This invention relates to the field of automotive R&D technology, and in particular to a method for simulating vehicle ride comfort under non-stationary conditions. Background Technology

[0002] Ride comfort, as one of the fundamental performance characteristics of a vehicle, not only affects the comfort and health of passengers and the integrity of cargo, but also has a significant impact on the vehicle's power, economy, handling stability, and its impact on road surface damage. Therefore, research on vehicle ride comfort is receiving increasing attention both domestically and internationally. Research on vehicle ride comfort mainly includes two methods: testing and simulation. Simulation can reduce the number of tests, improve the quality and efficiency of vehicle development, and reduce development costs and time. Therefore, simulation has become an important tool in modern automotive research and development.

[0003] This study investigates the uniform speed condition in a simulation of vehicle ride comfort under stationary conditions. When vehicles operate on urban roads, they often experience non-uniform speed conditions such as starting, accelerating, and braking, meaning the road surface excitation is non-stationary in the time domain. Therefore, studying vehicle ride comfort under non-stationary conditions is more realistic. However, the theory of non-stationary random vibration is far less mature and widely applied than that of stationary random vibration, and research on vehicle random vibration under non-stationary conditions is still limited.

[0004] Simulation of vehicle ride comfort under non-stationary conditions mainly includes two aspects: modeling and solving.

[0005] Currently, the lumped parameter method is mainly used to model vehicle ride comfort under non-stationary conditions. The drawback of this method is that it can only create simple vehicle models, not high-precision ones, and it requires complex formula derivations.

[0006] Currently, there are two main methods for solving vehicle ride comfort simulations under non-stationary conditions. One is the covariance equivalent method. This method transforms the non-stationary random process of a frequency-modulated structure into a stationary random process of amplitude-modulated structure. The drawback of this method is that it can only solve for the vehicle's vibration response, not for the power spectral density and root-mean-square value of the vehicle's non-stationary random vibration. The other method is the transient spatial power spectral density method. This method transforms the mathematical model of a linear time-invariant system under non-stationary excitation into a mathematical model of a linear non-stationary system under stationary excitation by transforming the time domain to the spatial domain. The advantage of this method is that it is essentially a natural generalization of stationary random processes, making it conceptually easier to understand. However, it has low solution efficiency, especially for complex models. Summary of the Invention

[0007] The technical problem to be solved by the present invention is to provide a method for simulating the ride comfort of a car under non-stationary conditions, which addresses the shortcomings of the prior art. This method can perform ride comfort simulations not only for simple models but also for complex models. It does not require the derivation of complex mathematical formulas and has high solution efficiency.

[0008] To solve the above-mentioned technical problems, the technical solution adopted by the present invention is as follows:

[0009] A simulation method for vehicle ride comfort under non-stationary conditions is proposed. The method involves establishing a vehicle model using the finite element method, applying boundary conditions, and generating a BDF file. The transient spatial frequency response function method combined with the spatial domain virtual excitation method is used for solving the problem. The new delay time of the front and rear wheels is calculated, and the corresponding statements in the BDF file are modified accordingly. The new starting frequency and frequency interval are also calculated, and the corresponding statements in the BDF file are modified accordingly.

[0010] Furthermore, the specific steps of the method are as follows:

[0011] Step 1: Create a finite element model of the car, apply boundary conditions, and generate a BDF file;

[0012] Step 2: Loop over time t;

[0013] Step 3: Automatically call Nastran software, run the bdf file, and perform simple harmonic vibration analysis;

[0014] Step 4: Extract the real and imaginary parts of the relevant node displacements, calculate the amplitude of the response, and calculate the power spectral density of the response based on the transient spatial frequency response function method combined with the spatial domain virtual excitation method;

[0015] Step 5: According to Calculate the new delay time for the front and rear wheels, and modify the corresponding statements in the BDF file, where L is the wheelbase and τ is the distance between the front and rear wheels. i For the new delay time of the front and rear wheels, u i For velocity; according to f i =n i u i Calculate the new starting frequency and frequency interval, and modify the corresponding statements in the BDF file, where f i n i u i These are time frequency, spatial frequency, and velocity, respectively.

[0016] Step 6: When t = t k At that time, t k The runtime is [time], and the loop ends with the output graph.

[0017] Furthermore, the process of establishing the simplified finite element model of the automobile in step 1 is as follows:

[0018] First, establish the key nodes, then establish the elements; use spring elements for the suspension and wheels, beam elements for the body, and mass elements for each mass; add element properties and material properties, and degrees of freedom for fully constrained ground nodes; apply road surface excitation.

[0019] Furthermore, the process of establishing the complex finite element model of the automobile in step 1 is as follows:

[0020] The 3D model of the car was imported into HyperMesh software. Neutral surfaces were extracted, geometry was cleaned and repaired, and the mesh was divided into 2D and 3D elements. The element quality was checked and repaired. Element properties and material properties were added, the degrees of freedom of the fully constrained ground nodes were determined, and road surface excitation was applied. After removing the four tires, they were simulated using spring elements.

[0021] Furthermore, the specific method for applying road surface excitation is as follows:

[0022] Power spectral density matrix G of four-wheel vehicle road excitation in spatial domain * q (n) is:

[0023]

[0024]

[0025] coh(n)=e 2πnB (18)

[0026] In the formula, L is the wheelbase; B is the track width; and coh(n) is the coherence function.

[0027] According to equation (7), G can be solved. * q The two non-zero eigenvalues ​​λ1 and λ2 of (n) are:

[0028] λ1=2[1-coh(n)]G q (n) (19)

[0029] λ² = 2[1 + coh(n)]G q (n) (20)

[0030] Eigenvectors corresponding to two eigenvalues and for:

[0031]

[0032]

[0033] The corresponding virtual road surface excitation is then constructed as follows:

[0034]

[0035] According to equations (21) and (22), the tire grounding nodes are loaded in two load steps, with an excitation amplitude of 1, and a delay time is added at the corresponding wheel.

[0036] Furthermore, in step 4, the specific method of the transient spatial frequency response function method is as follows:

[0037] By utilizing the relationship between time frequency f and spatial frequency n, the transient spatial frequency response function H(s,n) is obtained.

[0038] The expressions for road surface roughness h in the time domain and spatial domain are as follows:

[0039] h = h0e 2πft (1)

[0040] h = h0e 2πns (2)

[0041] In the formula, h0 is the road surface unevenness height coefficient, f is the time frequency, t is time, n is the spatial frequency, and s is the distance;

[0042] Through comparison, we can conclude that:

[0043] ft=ns (3)

[0044] f = nu (4)

[0045] In the formula, u is the vehicle speed;

[0046] When the car accelerates at a constant speed, there is

[0047]

[0048] In the formula, u0 is the initial vehicle speed, and a is the acceleration;

[0049] but

[0050]

[0051] Substituting equation (6) into the frequency response function H(f) of the vehicle vibration system in the time domain, we obtain the transient spatial frequency response function H(s,n).

[0052] Furthermore, the spatial domain virtual excitation method is the implementation of the multi-point excitation spatial domain virtual excitation method in finite element software. The specific method is as follows:

[0053] Substituting Euler's formula (11) into the formula (8) for the independent spatial virtual excitation vector, we obtain formula (12);

[0054]

[0055] e j2πns =cos(2πns)+jsin(2πns) (11)

[0056]

[0057] In the formula, Let λ be the i-th spatial virtual activation vector. i For eigenvalues, For unit eigenvectors, for The conjugate vector; r is the number of independent virtual excitation vectors;

[0058] After simple harmonic motion, the virtual displacement response at point k can be obtained as follows:

[0059]

[0060] In the formula, B i (s,n) represents the amplitude of the response under simple harmonic vibration, ψ i Phase angle;

[0061] From equation (11), we get

[0062]

[0063] According to equation (10), the self-power spectral density of the displacement response at point k is:

[0064]

[0065] Therefore, it is only necessary to find the amplitude B of the response under simple harmonic vibration. i (s,n), the spatial power spectral density of the response can be obtained by equation (15).

[0066] The beneficial effects of adopting the above technical solution are as follows: The vehicle ride comfort simulation method under non-stationary conditions provided by this invention addresses the problem that existing solutions can only simulate simple models and not complex models. It uses the finite element method to establish vehicle models, enabling the creation of both simple and complex high-precision models. To address the low solution efficiency of existing solutions, especially for complex models, this technology employs a transient spatial frequency response function method combined with a spatial domain virtual excitation method. The transient spatial power spectral density method utilizes the frequency response function in the frequency domain, which is a natural generalization of stationary random processes, making it easier to understand and primarily solving the solution problem without requiring the derivation of complex mathematical formulas. Since the virtual excitation method transforms the stationary random vibration analysis of the structural system into a deterministic harmonic response analysis, it has the advantages of simplicity, accuracy, and high efficiency, primarily addressing the solution efficiency problem, and achieving high solution efficiency. Attached Figure Description

[0067] Figure 1 This is a flowchart of a vehicle ride comfort simulation method under non-stationary conditions provided in an embodiment of the present invention;

[0068] Figure 2 This invention provides a mechanical model of a 7-DOF spatial vibration system for an automobile.

[0069] Figure 3 This is a 7-DOF spatial finite element model of an automobile provided in an embodiment of the present invention;

[0070] Figure 4 The simulation results of the ride comfort of the 7-DOF spatial finite element model of a car provided in the embodiments of the present invention are shown in Figure (4a), which is the simulation result of the vehicle body acceleration power spectral density, Figure (4b) is the simulation result of the front left dynamic deflection, Figure (4c) is the simulation result of the front right dynamic deflection, Figure (4d) is the simulation result of the rear left dynamic deflection, Figure (4e) is the simulation result of the rear right dynamic deflection, Figure (4f) is the simulation result of the relative dynamic load of the front left wheel, Figure (4g) is the simulation result of the relative dynamic load of the rear left wheel, Figure (4h) is the simulation result of the relative dynamic load of the front right wheel, and Figure (4i) is the simulation result of the relative dynamic load of the rear wheels.

[0071] Figure 5 This is a finite element model of a complete passenger car provided in an embodiment of the present invention;

[0072] Figure 6 The simulation results of the ride comfort of the finite element model of the whole car provided in the embodiment of the present invention are shown in Figure (6a), which is the simulation result of the front seat acceleration power spectral density, and Figure (6b) is the simulation result of the rear seat acceleration power spectral density. Detailed Implementation

[0073] The specific embodiments of the present invention will be described in further detail below with reference to the accompanying drawings and examples. The following examples are for illustrative purposes only and are not intended to limit the scope of the invention.

[0074] like Figure 1 As shown, the method of this embodiment is described below.

[0075] Step 1: Create a finite element model of the car, apply boundary conditions, and generate a BDF file;

[0076] Step 2: Loop over time t;

[0077] Step 3: Automatically call Nastran software, run the bdf file, and perform simple harmonic vibration analysis;

[0078] Step 4: Extract the real and imaginary parts of the relevant node displacements, calculate the amplitude of the response, and calculate the power spectral density of the response based on the transient spatial frequency response function method combined with the spatial domain virtual excitation method.

[0079] The specific method of the transient spatial frequency response function method is as follows:

[0080] By utilizing the relationship between the time frequency f and the spatial frequency n, the transient spatial frequency response function H(s,n) is obtained.

[0081] The expressions for road surface roughness in the time domain and spatial domain are as follows:

[0082] h = h0e 2πft (1)

[0083] h = h0e 2πns (2)

[0084] In the formula, h0 is the road surface unevenness height coefficient, f is the time frequency, t is time, n is the spatial frequency, and s is the distance.

[0085] Through comparison, we can conclude that:

[0086] ft=ns (3)

[0087] f = nu (4)

[0088] In the formula, u is the vehicle speed.

[0089] When the car accelerates at a constant speed, there is

[0090]

[0091] In the formula, u0 is the initial vehicle speed, and a is the acceleration.

[0092] but

[0093]

[0094] Substituting equation (6) into the frequency response function H(f) of the vehicle vibration system in the time domain, we obtain the transient spatial frequency response function H(s,n).

[0095] Based on the concept of time-domain virtual excitation, a spatial-domain virtual excitation method for multi-point excitation is derived. Suppose a linear structural system is subjected to multi-point partially coherent stationary random excitation, and its power spectral density matrix is ​​S. x (n), since the power spectral density matrix is ​​a Hermitian matrix, it can be expressed as

[0096]

[0097] In the formula, λ i For eigenvalues, For unit eigenvectors, for The conjugate transpose of .

[0098] Construct r independent spatial virtual activation vectors

[0099]

[0100] In the formula: Let be the i-th spatial virtual activation vector.

[0101] achievable

[0102]

[0103]

[0104] The spatial domain virtual excitation method in this embodiment is the implementation of the multi-point excitation spatial domain virtual excitation method in finite element software.

[0105] Substituting Euler's formula (11) into the formula (8) for the independent spatial virtual excitation vector, we obtain formula (12);

[0106] e j2πns =cos(2πns)+jsin(2πns) (11)

[0107]

[0108] In the formula, for The conjugate vector.

[0109] After simple harmonic motion, the virtual displacement response at point k can be obtained as follows:

[0110]

[0111] In the formula, B i(s,n) represents the amplitude of the response under simple harmonic vibration, ψ i This is the phase angle.

[0112] From equation (11), we get

[0113]

[0114] According to equation (10), the self-power spectral density of the displacement response at point k is:

[0115]

[0116] Therefore, it is only necessary to find the amplitude B of the response under simple harmonic vibration. i (s,n), the transient spatial power spectral density of the response can be obtained by equation (15).

[0117] Step 5: According to Calculate the new delay time for the front and rear wheels, and modify the corresponding statements in the BDF file, where L is the wheelbase and τ is the distance between the front and rear wheels. i For the new delay time of the front and rear wheels, u i For velocity; according to f i =n i u i Calculate the new starting frequency and frequency interval, and modify the corresponding statements in the BDF file, where f i n i u i These are time frequency, spatial frequency, and velocity, respectively.

[0118] Step 6: When t = t k At that time, t k The runtime is [time], and the loop ends with the output graph.

[0119] Example 1

[0120] Simulations are performed using a typical 7-DOF spatial model as an example. Figure 2 As shown in Table 1, the parameters of the car model are shown in Table 1. The simulation condition is a Class B road surface, with an initial speed of 0 m / s and an acceleration of 2 m / s². 2 The runtime is 15 seconds.

[0121] Table 1 Parameters of the Car Space Model

[0122]

[0123]

[0124] Create a finite element model of the car. Create the finite element model in HyperMesh. For example... Figure 3As shown, the process of establishing a simple model is as follows: First, 19 key nodes are established. The body uses CBEAM beam elements, the suspension uses PBUSH1D spring-damping elements, the tires use PELAS spring elements, and each lumped mass uses CONM2 mass elements. Element properties and material properties are added, along with the degrees of freedom of the fully constrained ground nodes. Road surface excitation is then applied, specifically as follows:

[0125] Power spectral density matrix G of four-wheel vehicle road excitation in spatial domain * q (n) is:

[0126]

[0127]

[0128] coh(n)=e 2πnB (18)

[0129] In the formula, L is the wheelbase; B is the track width; coh(n) is the coherence function; n0 is the reference space frequency, n0=0.1m / s.

[0130] According to equation (7), G can be solved. * q The two non-zero eigenvalues ​​λ1 and λ2 of (n) are:

[0131] λ1=2[1-coh(n)]G q (n) (19)

[0132] λ² = 2[1 + coh(n)]G q (n) (20)

[0133] Eigenvectors corresponding to two eigenvalues and for:

[0134]

[0135]

[0136] The corresponding virtual road surface excitation is then constructed as follows:

[0137]

[0138] According to equations (21) and (22), the tire grounding nodes are loaded in two load steps, with an excitation amplitude of 1, and a delay time is added at the corresponding wheel.

[0139] After the finite element model is established, according to Figure 1The program flowchart was created in Matlab software. The response quantities are the vehicle body center of gravity acceleration, dynamic deflection, and relative dynamic load of the wheels.

[0140] The virtual responses for each response quantity are:

[0141] The virtual response of acceleration at the center of gravity of the vehicle body is

[0142]

[0143] In the formula, 13 is the node number.

[0144] The virtual response of the front and rear suspension dynamic deflection is

[0145]

[0146]

[0147] The virtual response of the relative load between the front and rear wheels is

[0148]

[0149]

[0150] Calculate the amplitude of the response under harmonic excitation according to equations (24) to (28). Calculate the total power spectral density of the response according to equation (15).

[0151] Simulation results of vehicle ride comfort under non-stationary conditions are as follows: Figure 4 As shown.

[0152] Example 2

[0153] A complex model of a car is used as an example for simulation. The simulation condition is a Class B road surface, with an initial speed of 0 m / s and an acceleration of 2 m / s². 2 The runtime is 15 seconds.

[0154] The process of building the complex model is as follows: The 3D model of the car is imported into HyperMesh software; neutral surfaces are extracted; geometry is cleaned and repaired; the mesh is divided into 2D and 3D elements; element quality is checked and repaired; element and material properties are added; degrees of freedom of the fully constrained ground nodes are defined; and road surface excitation is applied; a BDF file is generated. For ease of loading, the four tires are removed and simulated using spring elements. Ultimately, the finite element model consists of 1,394,272 elements and 1,490,235 nodes. Figure 5 As shown. The method for applying road surface excitation is the same as that in Example 1.

[0155] When performing harmonic response analysis in Nastran, the modal method is used. Since the modal analysis is completely consistent under different time conditions, a restart method is adopted. This avoids repeated calculations of the modes, thus saving a significant amount of time and computational costs.

[0156] according to Figure 1 The program flowchart was created in Matlab software. The acceleration at the driver's seat was used as a separate factor. And the corresponding acceleration on the rear seats In response, the simulation results of vehicle ride comfort under non-stationary conditions are as follows: Figure 6 As shown.

[0157] Compared with other methods, this invention can simulate not only simple models, but also high-precision complex models; at the same time, it does not require the derivation of complex mathematical formulas and has high solution efficiency.

[0158] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features therein; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope defined by the claims of the present invention.

Claims

1. A method for simulating vehicle ride comfort under non-stationary conditions, characterized in that: The method described above uses the finite element method to establish a car model, applies boundary conditions, and generates a BDF file. The transient spatial frequency response function method combined with the spatial domain virtual excitation method is used to solve the problem; Calculate the new delay time for the front and rear wheels, modify the corresponding statements in the BDF file, calculate the new starting frequency and frequency interval, and modify the corresponding statements in the BDF file. The specific steps are as follows: Step 1: Establish a finite element model of the vehicle, apply boundary conditions, and generate a BDF file; the process of establishing a simple finite element model of the vehicle is as follows: First, establish the key nodes, then establish the elements; use spring elements for the suspension and wheels, beam elements for the body, and mass elements for each mass; add element properties and material properties, and degrees of freedom for fully constrained ground nodes; apply road surface excitation. The specific method for applying road surface excitation is as follows: Power spectral density matrix of four-wheel vehicle road excitation in spatial domain for: (16) (17) (18) In the formula, L This refers to the wheelbase; B Wheelbase; It is a coherence function; For spatial frequency, n 0 is the reference spatial frequency; According to equation (7), the solution is obtained. Two non-zero eigenvalues and for: (19) (20) Eigenvectors corresponding to two eigenvalues and for: (21) (22) The corresponding virtual road surface excitation is then constructed as follows: (23) in, The distance is given; according to equations (21) and (22), the tire grounding nodes are loaded in two load steps respectively, with an excitation of amplitude of 1, and a delay time is added at the corresponding wheel. Step 2: By time t Perform a loop; Step 3: Automatically call Nastran software, run the bdf file, and perform simple harmonic vibration analysis; Step 4: Extract the real and imaginary parts of the relevant node displacements, calculate the amplitude of the response, and calculate the power spectral density of the response based on the transient spatial frequency response function method combined with the spatial domain virtual excitation method; The spatial domain virtual excitation method is the implementation of the multi-point excitation spatial domain virtual excitation method in finite element software. The specific method is as follows: Substituting Euler's formula (11) into the formula (8) for the independent spatial virtual excitation vector, we get formula (12). (8) (11) (12) In the formula, For the first i A spatial virtual activation vector, For eigenvalues, For unit eigenvectors, for The conjugate vector; r is the number of independent virtual excitation vectors; After simple harmonic motion, the virtual displacement response at point k can be obtained as follows: (13) In the formula, Let be the amplitude of the response quantity under simple harmonic vibration. Phase angle; From equation (11), we get: (14) According to equation (10), the self-power spectral density of the displacement response at point k is: (15) Therefore, it is only necessary to determine the amplitude of the response under simple harmonic vibration. The spatial power spectral density of the response can be obtained by equation (15). Step 5: According to Calculate the new delay time for the front and rear wheels, and modify the corresponding statements in the BDF file, where... L Wheelbase For the new delay time of the front and rear wheels, For speed; according to Calculate the new starting frequency and frequency interval, and modify the corresponding statements in the BDF file, including... , , These are time frequency, spatial frequency, and velocity, respectively. Step 6: When t = t k hour, t k The runtime is [time], and the loop ends with the output graph.

2. The vehicle ride comfort simulation method under non-stationary conditions according to claim 1, characterized in that: The process of establishing the complex finite element model of the automobile in step 1 is as follows: The 3D model of the car was imported into HyperMesh software. Neutral surfaces were extracted, geometry was cleaned and repaired, and the mesh was divided into 2D and 3D elements. The element quality was checked and repaired. Element properties and material properties were added, the degrees of freedom of the fully constrained ground nodes were determined, and road surface excitation was applied. After removing the four tires, they were simulated using spring elements.

3. The vehicle ride comfort simulation method under non-stationary conditions according to claim 1, characterized in that: In step 4, the specific method of the transient spatial frequency response function method is as follows: The transient spatial frequency response function is obtained by utilizing the relationship between the time frequency f and the spatial frequency n. ; Road surface unevenness h The expressions in the time domain and the spatial domain are as follows: (1) (2) In the formula, This is the road surface unevenness height coefficient. For time frequency, For time, For spatial frequency, For distance; Through comparison, we can conclude that: (3) (4) In the formula, u is the vehicle speed; When the car accelerates at a constant speed, there is (5) In the formula, The initial vehicle speed, For acceleration; but (6) Substituting equation (6) into the frequency response function of the vehicle vibration system in the time domain... The transient spatial frequency response function is obtained. .

Images